The theory of practice architectures and its discontents: disturbing received wisdom to make it dangerous again

ABSTRACT

This paper critically examines the theory of practice architectures in response to a call for its reinvigoration towards more politically charged scholarship. While such goals seem acceptable, necessary even, it raises questions about the consequences of strengthening the moral imperatives that already undergird the theory. Drawing on a Foucauldian analytical ‘toolkit’, we explore why Schatzkian philosophy, which is conspicuous in its ongoing usage, was adopted in the development of the theory and critically consider why it remains valuable to this day. To do this work, we make a genealogical incision into the corpus of the theory of practice architectures beginning with Changing Practices, Changing Education, which is by far the most influential text in the field. By doing so, we bring to light the political imperatives shaping the authorial choices made in the text and think through the consequences for a reflexive criticality in the development of the theory.

KEYWORDS:

• Practice
• practice architectures
• changing practices
• praxis
• education

Introduction

The impetus for this paper stems from our reaction to Hopwood’s (2021) provocation that ‘we need to make the [theory of practice architectures] dangerous (again)’ (p. 88). The theory of practice architectures (hereafter TPA) resides within the broad family of practice theories which typically understand social life through an emphasis on how it is practised (see, e.g. Nicolini, 2013, Hui, et al., 2017). The notion of ‘architectures’ gestures towards the idea of seeing social formations as ‘made by people, and thus open to be re-made by people’ (Kemmis et al., 2014b, p. 17). TPA itself aims to articulate a theory of practice that is practical and critical in orientation, so that transformation of unjust conditions that delimit human flourishing becomes possible and achievable (Grootenboer & Edwards-Groves, 2024). As such, ambitions towards the ‘dangerous’ work of social transformation have long-been foundational elements. However, for Hopwood, the road to a more ‘dangerous’ version of TPA might be achieved by recommitting to its teleology of social and educational change undergirded by moral imperatives. To our thinking Hopwood’s claim warrants a more critical examination, which we hope will generate reflective conversations, perhaps even further critique. In doing so, this paper raises important questions more broadly about how critically oriented theories are inherited, built on and practised by researchers interested in the nature and influence of practices on people and the worlds in which they inhabit.

Hopwood (2021) is guided by Stetsenko’s (2016) transformative activist stance, which emphasises the role of human agency in social change, and tasks us with moving beyond describing the social worlds of our subjects to get on with the task of transforming them through our scholarly intuitions and interventions. While we applaud Hopwood’s attempt to renew critical scholarship using TPA for social transformation, we began to consider if such calls for ‘a reinvigoration of politically charged theory’ (Hopwood, 2021, p. 78) could do more harm than good. His ‘call to arms’ is arguably a seductive ethico-politics, where every reader becomes ‘a potential convert, vulnerable to persuasion’ (Rooney, 1989, p. 2). So, we started to ask if a ‘resolutely committed and partisan stance’ (Hopwood, 2021, p. 90) would have unforeseen effects on the future development and use of TPA. For instance, we wondered if strengthening the moral convictions undergirding TPA may discursively lead to self-censoring rather than self-critique – more ‘doing’ and less discretion in relation to the theory itself.

In light of such ruminations, we began to wonder if scholars involved in the development of TPA had been sufficiently cognisant of the need to hold open a foyer to ‘question ruthlessly those instruments [being used]’ (Bourdieu & Wacquant, 1992, p. 249). This ‘ruthless’ reflexivity is crucial to avoid the pitfalls of adopting ‘unthought categories of thought which delimit the thinkable and predetermine thought’ (Bourdieu & Wacquant, 1992, p. 249). However, we do not want to suggest that TPA and its practitioners and advocates fail to question and improve upon preceding iterations of TPA (see, for example, Kemmis 2019, 2022). Yet, we wonder if radical reimagining of TPA is indeed possible and open to all, and if a re-emphasis of the political and social goals of TPA, as Hopwood (2021) has argued, increases the ‘dangers’ of stifling such work.

As TPA finds itself being adopted in research approaches and contexts beyond the theory’s historic and foundational commitments, i.e. to action research (e.g. Carr & Kemmis, 1986, Mattsson & Kemmis, 2007, Kemmis, 2010, Kemmis, et al, 2014b) and work in educational institutions like schools (e.g. Kemmis & Smith, 2008, Kemmis et al., 2014c), it is arguably an opportune time to ask such questions. As Bourdieu has argued, ‘sometimes we must refurbish concepts – first, to be more precise, and second, to make them more alive’ (Bourdieu & Eagleton, 2012, p. 220). Thus, in this paper, we examine the ‘received wisdom’ in TPA to consider if these inheritances are potentially constraining more radical reimaginings of the theory or if the theory remains sufficient, or sufficiently ‘alive’, as it continues to ‘travel’ beyond education.

While Marx (e.g. Theses on Feuerbach, 1888/1995) is often invoked in TPA to argue that one can never be free of one’s history, we think this is precisely the task at hand. As scholars who ourselves have been ‘stirred in’ (Kemmis et al., 2014c, p. 59) to the practices and communities of practitioners engaging with TPA, we think it is now time to ‘stir things up’. We think that such a worrying of TPA ‘canon’ can make room for the possibility of a more heterodox conceptualisation of TPA, which would then mitigate against the risk of orthodoxy and embrittlement. In this surmising, instead of petitioning Marx, we are reminded of Engels’ depiction of a meeting of the young Hegelians in his sketch Die Freien, which translates to ‘the free ones’. Engels’ sketch makes vivid the kind of intellectual work that involves risk, which is work that can be gritty and contrarian, which also might very well mean a few ‘upended chairs’ along the way. However, other than the feistiness of the intellectual encounter, Die Freien also suggests that we need to have an ‘understanding the social conditions of the production of knowledge’ (Ball, 2006, p. 9). Taken together with Marx’s above-mentioned injunction that history ‘matters’, Engels’ reflexive perspective of late 19^th-century political philosophers in Germany, of which he himself was one, suggested to us that using a socio-historical approach in relation to the development of TPA could be fruitful, even necessary. Such a critique, we hope, may help ‘open up a space of research’ (Foucault, 1991, p. 74) and augur in a more pluralistic and creative engagement with TPA.

Our focal point for this interrogation is Kemmis et et al.’ s (2014c) Changing Practices, Changing Education (hereafter Changing Practices). In relation to TPA, Changing Practices has great ‘inaugurative value’ (Foucault, 1977b, p. 133), because it is the earliest publication that shows how TPA can be put to work analytically and empirically. Moreover, the visual devices used in Changing Practices to elaborate the theory have continued to inform both subsequent theorisation and empirical work. As Kemmis (2022) has recently argued, ‘in our research on practices, both in work in the field and in subsequent analysis, my colleagues and I frequently use [the visual rendering of TPA seen in Changing Practices] as a guide’ (p. 19). With almost 1299 citations at the time of writing, Changing Practices is the most highly cited text in TPA corpus, which then marks other efforts at theorisation as derivative or subordinate, even those that are historically antecedent (e.g. Kemmis & Grootenboer, 2008).

Thus, while Changing Practices may only be partially synoptic of the broader efforts in the development of TPA, an analytical incision into the TPA corpus starting with Changing Practices is arguably a crucial vantage point to examine the presuppositions, assumptions and silences involved in ongoing development of TPA itself. However, even in our ambition to question and push past the limits of our own taken-for-granted understandings of TPA, we

‘don’t yet know whether [we are] going to get anywhere. What [we] say ought to be taken as “propositions”, “game openings”, where those who may be interested are invited to join in - they are not meant to be dogmatic assertions that have to be taken or left on bloc’. (Foucault, 1991, p. 74)

Hence, we do not aim to offer a ‘proof’, but a series of propositions. It is our hope that these provocations will encourage but also authorise others to work ‘against the grain’ of orthodoxy, to ‘think something other than what one has thought before’ (Foucault, 1989, p. 293, emphasis added).

The paper proceeds as follows. We begin with a discussion of the Foucauldian analytical ‘tools’ we use. We then turn our attention to Changing Practices. We examine the text’s correlations to other works, the assumptions and limits acknowledged but also those unspoken, and explore how power is mobilised reciprocally and interdiscursively. While there are a range of inter-textual imbrications and philosophical encounters within Changing Practices, we focus on the ‘conversation’ with Schatzki’s (e.g. 2002, 2003, 2010, 2012, 2016) practice-theoretic philosophy, which has been highly influential in the development of TPA. The tidiness and tensions between Kemmis et al.’s (2014c) and Schatzki’s thinking have much to say about the authorial choices and arguments made in Changing Practices. We conclude the article by discussing the implications of our critique and speculate about ways in which TPA may be reconsidered.

A Foucauldian encounter

In this paper, we bring to bear Foucault’s (1977a, 1977b) genealogical method, which ‘is first and foremost a mode of critique’ (Hook, 2005, p. 4). As Hook (2005) surmises, the objective of a genealogical analysis is to “interrogate the present, to examine its values, discourses and understandings with recourse to the past as a resource of destabilizing critical knowledge’ (Hook, 2005, p. 121–122). As Foucault’s (1977a) argues, a genealogical analysis

permits the discovery, under the unique aspect of a trait or a concept, of the myriad events - thanks to which, against which - they were formed. Genealogy does not pretend to go back in time to restore an unbroken continuity … and does not map the destiny of a people. On the contrary, to follow the complex course of descent is to maintain passing events in their proper dispersion; it is to identify the accidents, the minute deviations—or conversely, the complete reversals—the errors, the false appraisals, and the faulty calculations that gave birth to those things that continue to exist and have value for us. (p. 146)

In above passage, Foucault is explicit in highlighting that a genealogy is a is not a re-historicisation. Instead, a genealogy searches for the strategies, events and mechanisms involved in the ‘production of truth’ (Foucault, 1977b, p. 14), which reveals how words do more than convey meaning. It is this ‘more’ that we must reveal and describe” (Foucault, 1970, p. 49). Hence, we do not seek to establish a ‘lineage’, but instead to identify the repetitions and orchestrations, and ‘the different scenes where [discourse is] engaged in different roles’ (Foucault, 1977a, p. 140). In addition, as Foucault (1991) suggests, our genealogical descriptions will alternate with critical interrogations, each aspect complementing the other.

Throughout our critical analysis we reflexively ask, “what is this reason that we use? What are its historical effects? What are its limits, and what are its dangers? (Foucault, 1984, p. 249). Our critique in this article is thus not a search for the truth of TPA, but an effort to interrogate the mechanisms by which an authoritative version of TPA has been constructed and subsequently defended.

Turning to Schatzki

In Changing Practices there is a complex interplay of ‘borrowings’, but also ‘forgettings’, silences and elisions of arguments from various historically prior texts (e.g. notably, Kemmis & Grootenboer, 2008 and their theorisation of the notion of practice architectures). While Changing Practices (Kemmis et al., 2014c) was published in the same year as other publications that also featured the theory of practice architectures (e.g. Kemmis et al., 2014a, 2014b), none of these texts are as influential (at least if measured by citation metrics).

Occupying a prominent position in this interplay is the various dialogues with the philosophical perspective of Theodore Schatzki (e.g. 2002, 2003, 2010, 2012). The development of the TPA is arguably much indebted to Schatzki’s practice-theoretic perspective. Schatzki’s work, e.g. The Site of the Social (hereafter Sites) first published in 2002 and Timespace of Human Activity (hereafter Timespace) published in 2010, has long been ‘in conversation’ with Kemmis and colleagues (see, e.g. Kemmis, 2009, 2010, 2012, Kemmis & Grootenboer, 2008, Kemmis & Mutton, 2012, Kemmis et al. 2012). Sites follows from and builds on earlier works by Schatzki, for example, Social Practices (2002), where his Wittgensteinian influence is already visible. Our focus on Schatzki is not to deny the fact that there are many other influences involved in the development of TPA. However, TPA is conspicuous in its use of Schatzki’s language and his ideas on ontology.

Ontologically oriented theories such as espoused in Sites and Timespace are examples of philosophies of the so-called ‘ontological turn’, which has precipitated a move in the social sciences from thinking about differences of worldviews (e.g. cultural explanations) to thinking about different worlds (ontologically dependent explanations). Social research, having previously given thought to epistemological (e.g. positivism, constructivism) and methodological (e.g. quantitative, qualitative, and mixed methods) concerns, began to consider the question of ‘what is there?’ (Quine, 1951, p. 21, original emphasis). If one accepts that their choices of epistemology and methodology are based on deeper assumptions about the nature of reality, i.e. that ‘the possibilities for knowledge are given in the ontology’ (Brown, 2009, p. 5), then a theory of social action necessitates a deeper consideration of ontology. Thus ‘ontological self-consciousness’ (p. 189) should be ‘part of the conceptual armature of social investigation’ (p. 189) which ultimately informs investigative methods. While Schatzki (2003) admits that critics would argue that ‘ontologies are nothing but unnecessary and empirically unconfirmable presumptions’ (p. 188), he contends that one is doing ontology whether one ignores it or not.

Structuring an otherwise flat ontology

For Schatzki (2002), the ‘mesh’ of practice is a contingent and metamorphosing ‘horizontal structure’ (p. 154), but not ‘vertical’ structures that tend to be reified. As he argues,

whatever patterns are discernible from the bird’s-eye perspectives of general historians emerge as by-products of the rich and intricate congeries of processes that contingently join and fall apart with (and partly as) the passage of the social site. (Schatzki, 2002, pp. 261–262)

In the excerpt above, Schatzki is alerting us to the fact that sites of human coexistence do not ‘embrace multiple levels. It is better thought of as a single plane’ (Timespace, p. 91). As such, for Schatzki (2002), following a Wittgenstenian sensibility around thinking about disjuncture rather than ‘meta’ patterns of reality, there is no ‘level’ that is beyond the social, no ‘meta’ theory that expresses a rationality of an agent over-against a structured social world. Schatzki’s (2016) ontological conception is ‘flat’, that is, it ‘grants no priority to (face-to-face) interactions or the local situation’ (p. 32).

For Schatzki (2003) problems arise in social theorisation that proposes a variety of extra-individual features or influences in human co-existence (what Schatzki calls ‘societist’ theories), tend to ‘suffer a penchant towards reification’ (p. 188), because ‘distinctiveness of the social can be upheld only so long as these [extra-individual] facts are treated as a distinct level of reality, that is to say, only so long as they are reified’ (p. 187). As, Schatzki (2003) argues, any rendering of extra-individual conditions, what Foucault (1980) might term ‘the conditions of possibility’ (p. 162), suggests an ‘externality’ to those things we want to consider as ‘facts about individuals’, e.g. individual actions, cognitions, and intentions.¹

We read this danger in Changing Practices in how extra-individual properties of social inter-actions have a power over individuals, i.e. that such social features ‘enable’ and/or ‘constrain’ individuals. Our reading of Changing Practices suggests that Schatzki’s (2016) flat ontology is assimilated into TPA and put to work to justify precisely what Schatzki in fact seems to oppose. Kemmis and Grootenboer’s (2008) earlier theorisations about extra-individual characteristics also describe how they ‘shape’ and ‘form’ individual dispositions. Schatzki (2002) would likely reject such a formulation, because it suggests that ‘an actor self-propellingly follows an action trajectory of her or his own determination, [and] the state of the world forces her or him to take particular twists and turns and to leave off particular routes’ (p. 211).

For Schatzki (2002), ‘the mesh of practices and orders does not simply clear some paths and obliterate others’ (p. 226). However, in Changing Practices, constraints and affordances, i.e. extra-individual features ‘excluding certain actions … [and/or those] making actions available’ (p. 216), features this precise tendency. Now, a closer reading of Changing Practices may alert the reader to the fact that Kemmis et al. do at times suggest that agency is only ‘partly prefigured’ (p. 32), and that ‘projects’, i.e. goals and ‘projected activities’ (p. 8) that relate to them, can reshape and/or transform the ‘other, not so invisible “players” in our social world apart from the people’ (p. 7). The notion of prefiguration is Schatzkian, which he describes as ‘enablement/constraint’ (2016, p. 98). Schatzki argues that ‘the site of social life prefigures the paths taken by the human activity that perpetuates and alters it’ (Sites, p. xxii, emphasis added). However, Schatzki also cautions that prefiguration is a ‘variegated phenomenon’ (Sites, p. xxii) and ‘dispute[s] the widespread practice of analyzing the prefiguration of activity as the delimitation of fields of possibility’ (Sites, p. xxii). This is precisely the mistake that Changing Practices appears to make. As is argued in Changing Practices, an analysis of practice requires evidencing

the kinds of cultural-discursive, material-economic and social-political arrangements that make those practices possible. These arrangements, found in or brought to the site, create the conditions of possibility that do or do not support the sayings, doings and relatings of the practice and the project in which they hang together (p. 224–225, emphasis added)

Kemmis et al. (2014c) seem to misread the caution in Sites around reifying a priori structuring of the social, which then establishes ‘fields of possibility’. This then leads Kemmis et al. to a ‘working definition of practice’ (p. 31) wherein ‘practice architectures that enable and constrain practices exist in three dimensions [i.e. semantic space, physical-space-time and social space “on the side of the social”] parallel to the activities of saying, doing and relating [“on the side of the individual”]’ (p. 31, emphasis added). Such an individual versus extra-individual dichotomy implies ‘two distinct levels of reality’ (Schatzki, 2003, p. 187), and conceives agency, at best, in a reciprocal struggle with extra-individual and ‘not so invisible’ features.

Staying alive to the happeningness of practice

For Schatzki (2016), practices are not inherently individual and extra-individual nor micro or macro in character. While theorists working with TPA since Changing Practices may have explored the ways in which agency can be emphasised as a corrective to a deterministic reading of TPA, for example, as self-expression, self-development and self-determination (Kemmis & Edwards-Groves, 2018), it stands to reason that a two-tiered ontology can leave the impression that there is a ‘kind of a background condition that is always already given to unidirectionally shape and determine human development from the outside’ (Stetsenko, 2016). As Schatzki (2016) argues, a distinction between the micro and macro, depending on emphasis, can tend to grant an ontological priority to either level of reality. A perpetual balancing act is necessitated by such a conceptual framework. We can see this effort by how theorists have strived since the publication of Changing Practices to reemphasise agency (e.g. Kemmis & Edwards-Groves, 2018; Kemmis, 2022) and explore the lacuna around emotions (Wilkinson, 2021) as a corrective to earlier works, we think the bifurcation of self and the social persists in the TPA framing. Such a dyad between individual and extra-individual features necessitates a post-facto analytical pincer movement from above (TPA’s tripartite theory) and below (practitioner observation) to capture the ‘happeningness’ (Changing Practices, p. 29) of practice. In our understanding, happeningness is the aliveness of practices being practised, which the bifurcation of self and social annihilates as a first theoretical and methodological step.

Schatzki’s idea of happening is taken up as happeningness in Changing Practices, which does not simply denote the real-time events of practice but their comportment towards present, future and past events. Yet, Schatzki goes on to suggest that this happening (Geschehen), does not ‘entail any single conception of the temporal structure of action’ (p. 166). In other words, Schatzki suggests that people may take action (or not) but happeningness points to how practice ‘befalls’ them, i.e. ‘human activity combines something done and something that just happens’ (p. 168). This perspective allows Schatzki the latitude to avoid ‘hypostasizing teleology and motivation’ (p. 169) and admitting in the indeterminacy of action. In this way what might seem like rule-following or ‘like projects’ (Changing Practices, p. 95) are post-facto judgements of practices that are more variegated phenomena, linked together by ‘causation, use, constitution, intentionality, constraint and prefiguration’ (Schatzki, 2016, p. 32) but also involving ‘spontaneous, on the spot interactional improvisations’ (Schatzki, 2010, p. 181). In other words, the visual representations of TPA in Changing Practices, suggest a ‘generic ontological-compositional feature of the layout of social life that can characterize, in principle, any component or space-time swath of social existence’ (Sites, p. 5, emphasis added).

Despite the suggestion that a view from within (individual aspects) and without (extra-individual aspects) can give one post-facto access to ‘collective praxis from within’ (Kemmis, 2010, p. 25), we are not so sure. Good research relies on ‘getting a close-enough interpretation of the practisescape’ (Grootenboer & Edwards-Groves, 2024, p. 59), without falling prey to the distortions and distillations that invariably occur through our scholarly gaze (Freebody, 2003, p. 28). Moreover, empirical methods (e.g. interviews, surveys and policy analysis) that places the scholar outside the happeningness of practice can risk replacing actuality with the virtuality of theoretical idealisations (Grootenboer & Edwards-Groves, 2024).

Visual meta-rationalities that mystify and reify

Changing Practices may have started with the consideration of a Schatzkian flat ontology, but the rationalities that become foregrounded, especially in the visual representations of TPA in Changing Practices, belies a drift to a cartesian-like and compositional ontological perspective. For example, Kemmis et al’.s (2014c) use of ‘table[s] of invention’ (p. 224; hereafter TOI), described as ‘viewing platforms’ (p. 224, emphasis added) for observing (and interpreting) the action in classrooms, which are visual devices that are autopsy-like in its function and rationality. To our thinking, the TOIs in Changing Practices, which is offered as ‘merely a prompt for a certain way of making a reading’ (p. 227), has over time become a ‘specific set of thinking tools’ (Wilkinson, 2021, p. 43), which we think is discursively stereomorphic. That is, it suggests a ‘way’ of perceiving the social that analytically and a priori structures it into a duality as a logical precondition for analyst to then ‘think about relationships across the columns and between the rows, to find connections and interdependencies’ (p. 227, emphasis added). Yet, analytically and visually renting apart agency and structure has a tendency to resemiotise these twin aspects as sui generis phenomena, which then cannot easily be ‘put back together’ to convey the synthetical nature of practice that Schatzki would advocate and that Kemmis et al. also hopes to achieve. Even for TPA, a recourse to a purported dialecticism, and the visual aids and meta rationality of infinity symbols or lemniscates to suture together a two-tier reality (see, for example, Kemmis & Edwards-Groves 2018, Kemmis 2022) cannot fully realise the understanding ‘that action determines the instances and structures that allegedly determine it’ (Sites, p. 233).

As does a medical autopsy, which takes a whole person and separates out the body parts onto a literal table to see ‘whodunnit’, the viewing platforms in Changing Practices, are likewise renting apart the self from the social to elaborate a methodological starting point for their scholarly interventions and conclusions – post facto. These ‘viewing platforms’ are described in Changing Practices as a ‘version’ of the visual/tabular depiction of TPA itself (see, Changing Practices, p. 27), which also suffers the same dyadic ontological rationality.

A stereoscopic vision?

Arguably, Kemmis et al. (2014c) are aware of danger of analytically renting apart individuals from their contexts. To shore up the now dissected corpus of the social, they use an infinity symbol over the table form that summarises TPA.² This attempted Frankenstinean semiotic re-animation, it is not explained in Changing Practices. However, the infinity symbol has been used previously by Kemmis (2012), to visualise a ‘stereoscopic view – the view from within and the view from without’ (p. 901), but also the relation between ‘“living well” (in the sense of living appropriately) and helping to create “a world worth living in”’ (p. 895), and also to indicate the use of TPA in tandem with other theoretical frames (e.g. Wilkinson, 2021). More recently, Kemmis (in Mahon et al., 2017) has suggested that ‘the symbol is intended to be read as a kind of flow’ (Kemmis in Mahon et al., 2017, p. 13, emphasis added) between individual (sayings, doings and relatings) and extra-individual (cultural-discursive, material-economic, and social-political arrangements) aspects of human sociality. Yet, in the same volume, Mahon et al. (2017) suggest that the infinity symbol indicates ‘true dialectical relationship, like the relationship of the chicken and the egg’ (p. 14), despite the conventional use of this metaphor as a philosophical device to illustrate the indeterminable nature of causality in a dependency between separate entities, not reciprocity nor synthesis.

There are myriad instances and variations in use of the infinity symbol in the TPA oeuvre (e.g. see Kemmis, 2022), which renders it as an ambivalent semiotic device. As Kemmis (2019) concedes, the lemniscate (or infinity symbol) is asked to ‘do some pretty heavy lifting when we ask it to represent these ideas’ (p. 28). However, these various usages invariably imply that the role of the infinity symbol to suture together ostensibly disparate entities or different levels of reality. For example, Wilkinson (2021) has recently argued that ‘a dual lens fosters a dialectical interplay between the differing goods of a practice that gives us a stereoscopic view’ (p. 58, emphasis added). Changing Practices also argues that a view of practice must be ‘stereoscopic’, i.e. comprising ‘both individual and systemic aspects’ (p. 165, emphasis added). A stereoscopic device, in reality, should result in the overlay of two images to produce a unitary but ‘stereo’ image.

The idea of a dialectical relation as a reciprocally mediating relationship between agents and practice architectures can be seen in the works predating Changing Practices, for example, in Kemmis and Grootenboer (2008), where the notion of practice architectures is first mentioned. In this earlier work, the authors use a simple double-sided arrow to contend that individuals shape societal conditions, and these conditions ‘dialectically and reciprocally shape individuals’ (p. 51), implying an interdependence between two or more entities. The notion of the dialectic appears equivalent to reciprocity, as it does in Changing Practices. In more recent book chapter, Kemmis (2022) revisits the notion of the dialectic to describe it as a ‘reciprocal mediation of the realms of the individual and the social’ (p. 18) and repeatedly uses the lemniscate to highlight this reciprocity. However, a dialectic should lead to a Hegellian-like synthesis (cf. Kojeve, 1969) turns the notion of reciprocal mediation of individual and extra-individual features into a conception that is more ontologically ‘flat’ but epistemologically ‘alive’ like the three-dimensional depth effect of stereoscopy. Yet, our more common encounters with the device, for example, the cinematic convention of depicting the view through binoculars is invariably rendered as a double circle, preserves duality, almost as if a resultant unitary synthetical rendering is beyond the intelligibility for the average audience. As such, one wonders if the infinity symbol or lemniscate is an unintelligible or ‘muddy’ resemiotising device.

Hopwood’s (2016), instead of a lemniscate, uses of the analogy of the Möbius strip, where a two-sided strip of tape is looped, twisted and then joined end to produce a one-sided shape. We think that the Möbius strip would be a better way to conceive of the social as a synthesis rather than a mutually informing dual-level reality. Hopwood also suggests that Schatzki’s approach is consistent with this metaphor of the Möbius strip, and himself contends that Changing Practices (and his own work in TPA) has had a tendency to erroneously see reality as micro actions contributing to macro phenomena and vice versa. Hopwood adopts a more filial sense of the Schatzkian flat ontology to argue for a more emergent sense of practice, where context and action are ‘intimately comingled’ (p. 133). However, the Möbius strip is a confounding non-orientable mathematical construct that we think is inaccessible as a basis for method or a ‘toolkit’. Even a theoretical recognition of ‘comingling’ does not belie our intuition that A TPA informed analysis could remain inescapably stereomorphic. i.e. intractably generating a two-tier virtuality of an otherwise ontologically flat actuality.

The cage of language or the invention of status?

As much as this language is seductive to our own sensibilities, we wonder if even Hopwood’s (2016) use of the Möbius strip might just be lulling us into the false sense that our cognition can escape the ‘cage’ of language (cf. Wittgenstein, 1965). However, a more pertinent question in relation to our focus in this article is how did TPA come to hold firm to Schatzki, while at the same time abandoning him? This philosophical yoking has led to a ‘muddying’ of the ontological foundations of TPA, as tensions are replaced by a misleading tidiness over time and shared use. As Schatzki (2016) laments, ‘scholars sometimes advocate an ontology but ignore it when investigating particular phenomena. Doing this amounts to not taking ontology seriously’ (p. 40). Yet, ontologies have consequences for theory.

The ways in which our studies are shaped by our ontological perspective, lead to a priori analytical frameworks and modes of their representation (e.g. text, tables and visual representations) that then mediates not only the questions asked but also influences the methods chosen to obtain and analyse our ‘data’. As Denzin and Lincoln (1998) have argued, ‘choices of conceptual framework, of research questions, of samples of the “case” definition itself, and of instrumentation all involve anticipatory data reduction’ (p. 184, emphasis added). Thus, visual devices like TOIs may aid analysis, but such ‘displays beget analyses, which then beget more powerful, suggestive displays’ (Denzin & Lincoln, p. 189). These visual elements of texts are typically purposed towards ‘[inter-modal or inter-semiotic] expansions of meaning’ (O’Halloran, 2011, p. 124). These ‘meta grammars’ can resemiotise (Iedema, 2003) or alter meanings as contentions and concepts move intratextually and become re-elaborated in multimodal representations and do various forms of ‘work’ including supporting the continuity of an enterprise despite its contradictions (Foucault, 1991).

For instance, while the conventional tabular approach to communicating visually is a seemingly benign feature, in this context of an academic text, it has a meta-function to persuade readers that the authors are rigorous in their methods by displaying an appropriate amount of ‘scientificity, or pretensions to it’ (Foucault, 1970, p. 184). These conventions are not benign intersemiotic devices but are examples of ‘political technologies’ that elaborate and exercise a ‘micro-physics of power’ (cf. Foucault, 1977b). In this sense, the use of tables is less an ‘invention’ than it is an integration of techniques and discourses that exploit the asymmetries between authors and audience, e.g. the pre-existing status differences between the academy and public, but also the differences in relative access to privileged forms of knowledge between those ‘in the know’ and newcomers who are being ostensibly ‘stirred in’. Such a strategy can leave little room for critique, codifying some aspects, prescribing others, and elaborating a domain of interest in which the discourse claims an authority (Foucault, 1991). Thus, while the Schatzkian ontological ‘foundation’ is emphasised throughout the TPA oeuvre, reputedly because of the key ideas derived from it, it also functions as a means of distinction.

We can see this discourse around status and distinctiveness in Changing Practices where the authors emphasise how ‘the theory that informs this book is different’ (p. 4, emphasis added), that it ‘differs from other [practice-orientated] writings’ and is ‘essential’ (p. 12, emphasis added), if one hopes for ‘history-making educational action’’ (p. 26, emphasis in original). The question of ‘what Is distinctive and significant about the theory of practice architectures?’ (Mahon et al., 2017, p. 15), is the type of question that academics normatively ask themselves as they attempt to establish their ‘territory’, each with ‘distinct language or at least a distinct dialect and a variety of ways of demonstrating its apartness from others’ (Becher & Trowler, 1989, p. 22). Our Foucauldian sensibilities leads us to argue that Schatzki’s (2002) ontology has remained valuable, despite the crucial differences, because it ‘locates the [TPA] in relation to other practice theories’ (Mahon et al., p. 17), which is a crowded field of all vying for their place in history.

Political theory and the politics of theory

The intra-textual moments in Changing Practices, the suturing together of arguments, the ‘borrowing’ of ideas, conjectures, conclusions but also the strategic occlusions – ‘in short, the said as much as the unsaid’ (Foucault, 1980, p. 194) – are not the consequence of disinterested choices. Affinities and/or Incommensurableness between key organising ideas requires judgements, inclusions and exclusions that are political as much as they are reasoned or pragmatic. For example, Changing Practices displays a strong commitment to its intended audience, in particular teachers as action-researchers because the transformation of education will ‘endure and be sustained only when people do it for themselves’ (Kemmis et al., 2014c, p. 8). Yet, it logically follows that building theoretical tools for practitioners cum action-researchers require translating the complexity of esoteric philosophy and the hyper-intellectualism of the academy into an accessible vocabulary and framing. As Kemmis et al. (2014c) argue, that the TPA framework ‘may seem a very abstract proposition, [but they] nevertheless believe that it is also eminently practical … connect[ing] readily with practice … ’ (p. 220).

We do not want to imply that the TPA is a ‘dumbing down’ of a more nuanced and complex practice-theoretic view. After all, as Wittgenstein (1965) reminds us, there are always linguistic impediments when we go about attempting to think beyond our current limits. However, what may have been intelligible and insightful in striving for translatability of ideas in TPA, for instance, the ‘sayings, doings and relatings’ mantra oft recited by TPA scholars, might have risked ‘making the social more real, more orderly, more predictable than it is’ (Ball, 2006, p. 4). As Foucault (1983) would offer, it is not that every choice is ‘bad’, but that ‘everything is dangerous’ (p. 231).

The making of useful theory

Authors invariably have political commitments and goals for their writing (Foucault, 1977b). In effect, texts are a ‘strategic proposition’ (Foucault, 1980, p. 144) bearing its authors’ politics and proclivities, and their ‘desire to persuade’ (Rooney, 1989, p. 1). This political orientation is prominent from the outset of Changing Practices and is well-rehearsed throughout the TPA oeuvre. As Kemmis at al. argue,

our intention in this book is not ‘just’ (one might say) theoretical in the sense that we aim to articulate a theoretical language that can be used to describe and interpret the world. Our aim is also practical. We have a message that we believe is crucial for the development of education in and for the twenty-first century. It is simple, but it has far-reaching consequences. (p. 6, emphasis in original)

Kemmis et al’.s words suggest that the theoretical treatment in Schatzki’s Sites is insufficient for their purposes. For instance, beyond foregrounding agency, Sites does not offer a general theory of change. As Schatzki (2002) argues,

my discussion likewise neither aims for systemicity nor seeks to fashion the forms and mechanisms it considers into a template that can be used either to order or to account for particular changes … The only general thesis … is that explaining change is a matter of charting the agency that constitutes and/or brings it about. I do not address the many epistemological and methodological issues that bear on the activities of uncovering, reporting, and charting agency. (p. 235)

The lack of focus on systematicity, the lack of a theory of change, and Schatzki’s setting aside of considerations about methodologies for excavating the social, does not make Sites a particularly ‘useful’ book, especially in relation to Kemmis et al’.s goals for practical intelligibility and impact. Timespace, in contrast, with its ‘hopes to develop concepts that are useful in empirical investigation’ (p. 7) is more amenable to the goal of making Kemmis et al’.s work ‘practical … [and] also critical … [for] practical people’ (p. 6–7, emphasis in original). However, considering how Changing Practices has ‘knowledge-constitutive’ (Habermas, 1978) interests in relation to the emancipation and transformation of education, it suggests that the theoretical choices made in Changing Practices is connected to the commitments, identity, and politics of its authors and advocates.

A transformational resource or a resource that cannot be transformed?

It has been long held since Changing Practices, that TPA is a ‘transformational resource’ (Mahon et al., 2017, p. 20, emphasis in original). Yet, Mahon et al’.s (2017) comments in their introductory chapter are also illustrative of the danger that this teleology can trump self-reflective criticality. In this edited volume, Mahon et al. draws on Winston Churchill’s appeal to the USA for support during the World War Two, where he said, ‘give us the tools and we’ll finish the job’ (p. 20). Such a framing directs the reader’s attention towards ‘use-value’, heroically so, aimed at fixing the ills of education. It stands to reason that, for advocates and adopters of TPA, it becomes a necessity to maintain rather than to question one’s ‘tools’ if one wants to get on with educational ‘history-making’. Namely, adopting a political theory requires one to ‘accept its positions, its ideology, and its motive for combat’ (Foucault, 1977b, p. 216). Yet, such a prominent discourse, through ‘repetition and sameness’ (Foucault, 1977b, p. 59), can delimit the ‘legitimate’ mode for engagement with the theory, precluding the chance that something might go otherwise.

We are not suggesting that educational research should be (or could be) value-free, but such ideological and political recruitment would impede (but not determine) the possibility of self-reflective challenge (cf. Foucault, 1979). Yet, we have a hunch that political interests have created an ‘epistemocentrism’ (Bourdieu & Wacquant, 1992, p. 254) that leaves TPA insufficient room to be reflexive of its social, political and moral commitments. We are also not contending that there is a total absence of retheorisation in relation to various aspects of the TPA. For example, works since Changing Practices exploring the socio-historical aspects of ‘studying education’ (Kemmis & Edwards-Groves, 2018), theorising emotions (Wilkinson, 2021) and rethinking the nature of learning (Kemmis, 2021) indicate that scholars working with TPA are continually striving to improve it. However, perhaps improvement is not what is needed, as none of these works have radically and ruthlessly attempted to destabilise theorisation in Changing Practices. Rather than remaining a ‘work in progress’ (Kemmis in an interview with Mahon, in Mahon et al., 2017, p. 228), the danger is that more or less iterative work will lead to the TPA framework becoming a settled ‘science’ (cf. Paulston, 1990) that can no longer see its incoherencies, becoming a ‘theoretical totalisation under the guise of “truth”’ (Foucault, 1977b, p. 217). This is a ‘truth’ we ourselves want to avert.

The alchemy of the author-function

It would be uncontroversial to suggest that Stephen Kemmis’ edifying contributions to TPA is without equal. The grasp of Kemmis over theoretical aspects in particular, how he features in edited volumes in expositions of the theory (e.g. Kemmis, 2019, Kemmis, 2022, Kemmis, 2023), is a prolific collaborator and supporter of a network of scholars using TPA (most recently in Reimer et al., 2023), as well as how theoretically oriented questions are often directed to Stephen Kemmis in various forums (e.g. interview in Mahon et al., 2017; NEARI in Ireland, 2022), evidences his sustained intellectual labour and influence. Such leadership has its merits, but such visibilities are double-edged. Through the passage of time, through repeated visibilities of authorship and citation, theory can grow as familiar as the author whose name becomes synonymous with it, demarcating both ownership and prohibition (Foucault, 1979).

Kemmis would likely be aghast at such a possibility in relation to his engagement with TPA, but the ‘author-function’ (Foucault, 1979, 1980) is not the vis a vis of the author alone. The author’s name does more than indicate the writer, it can have a ‘classificatory function [allowing] one to group together a certain number of texts, define them, differentiate them from and contrast them to others’ (p. 147). The author-function can also signify ‘a principle of a certain unity of writing … [serving to] neutralize the contradictions that may emerge in a series of texts’ (Foucault, 1979, p. 151) because the repeated presence of an author’s name in various texts, for example, denoting common usage of various ideas, implies affiliation and fidelity. The author-function can also be ‘transdiscursive’ in nature, that is, beyond a specific contribution to knowledge, a prolific writer like can establish ‘the possibilities and the rules for the formation of other texts’ (Foucault, 1979, p. 154). As Foucault argues ‘one can be the author of much more than a book – one can be the author of a theory, tradition, or discipline in which other books and authors will in turn find their place’ (p. 153, emphasis added). It is through these effects that texts can become authorised, sacralised, and indelibly yoked to an author’s name.

Collaborators and antecedent adopters may find that their linkages to the author-function can elevate their status, the visibility of their own work, but they also may discover that their ‘place’ is discursively marginalised. For instance, Kemmis’ collaborating authors in Changing Practices are hierarchised by ‘author order’ to denote the significance of their contribution. While this convention does not necessarily reveal the significance of each author’s contribution to a collaboration or account for synergetic effects, the elision of ‘corresponding’ authors’ names through in-text citation styles is commonplace. In-text can re-emphasise the significance of the first author in relation to their ‘et alia’ who become effaced. Thus, the author-function ‘is not defined by the spontaneous attribution of a discourse to its producer, but rather by a series of specific and complex operations’ (Foucault, 1979, p. 153).

The author-function can also establish limits, constrain divergence or direct attention through ‘modes of circulation, valorization, attribution, and appropriation’ (Foucault, 1979, p. 158). The author-function can demarcate the ‘rules of engagement’ or capture a discourse. As Foucault argues, the author-function ‘is a certain functional principle by which … one limits, excludes and chooses’ (p. 159). Even, intra-textual tensions, doubt and contradictory impulses of a sole author or collaborative authorship, can come under erasure because a text must ultimately be made docile by the time of its publication. This is a scholarly text’s ‘finished state. Homo academicus relishes the finished’ (Bourdieu & Wacquant, 1992, p. 219, emphasis in original). Consequently, the discursive effect of this erasure is that open systems of meaning and/or interpretation can become closed, and ‘theory building’ can give way to official theories (Bourdieu & Wacquant, 1992), even if unintended. In this way, the author-function extends its tyranny over the author themselves, even over the theory they have helped establish. As Kemmis recently confided in a webinar ‘if I was starting in 2008 again, now, I don’t think I’d call them practice architectures, but unfortunately after 15 years… . I can’t escape!’ (NEARI in Ireland, 2022). If Kemmis cannot ‘escape’ this demesne, maybe nobody can.

Conclusion

We are fully aware of the difficulty of attempting to re-invigorate theory, but also the risks involved in confronting the past. This paper’s paradoxical task, to pay homage to this past by destabilising it, runs the risk of offending those on whose shoulders we stand. However, our purpose here is not to criticise others who have historically contributed to the development of TPA, nor to simply attack the theory for its shortcomings. Instead, our hope is that we have destabilised it sufficiently to loosen its grip on our rationalities, so as to create a foyer to think otherwise; to indeed strengthen the theory by provoking scholarly critique about its merit and utility in contemporary research, and its transformative aspirations.

We began by exploring the encounter with Schatzki in Changing Practices, contending that choices made to depart from Schatzki’s synthetical ontology, in the efforts to produce and communicate a ‘practical’ set of theoretical tools and a focus on usage/utility, has ‘left behind’ theorisation, or worse, confused it. Our analysis suggests the choices made in relation to the theory were made with practitioners in mind and political goals in train, which suggests that the political urge to transform education and the need to provide an intelligible discourse for practitioners may be having deleterious effects on the development of TPA. However, there are social, political and cultural processes involved in the development of TPA, such as those implicated in the author-function, through which the potentials of hesitation and regret, disjunctions, and messy negotiations can all too easily be suppressed, sublimated or elided, which effectively erases the role these potentialities may have played in the development of TPA but also limits their potentially fruitful re-emergence.

We suggest that thinking anew with Schatzki’s (2002, 2016) ‘flat’ ontology may open avenues for scholars to reconsider the TPA tripartite framing and its post-facto analytics of suturing together of individual and extra-individual phenomena. If one is serious about keeping TPA ‘alive’ to the potential of transformation, both within the theory and beyond, then rather than an incremental approach, that is, one that is not beholden to more than it is informed by the past, a more ‘ruthless’ approach to theory building is needed. This will entail a resistance to the authority of the author-function and the weight of the ‘official’ canon of TPA, which will aid in loosening up the grip of the past over present potentiality for change.

We have admittedly only ‘scratched the surface’ of TPA in this paper. There are many threads to explore in depth and in detail, and we hope that our work might encourage others to engage in such an enterprise. This is perhaps ‘dangerous’ work, but to do otherwise, to fail to ruthlessly hold up our tools to scrutiny, may be equally dangerous for the future of the theory of practice architectures, and the potential it might otherwise achieve. In our final calculus, it would seem the first step to making TPA ‘dangerous again’ is to find the dangers within it.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Additional information

Notes on contributors

George Variyan

George Variyan is a lecturer in Master of Educational Leadership in the Faculty of Education at Monash University. His background includes teaching, learning and leading in schools in Australia and overseas. George’s engagement in research is based on a critically orientated sociology, which explores human agency in the relationship between education and society. Key interests include educational sociology, boys’ masculinities, social justice, climate activism and ethics.

Christine Edwards-Groves

Christine Edwards-Groves is Professor and ARC Fellow, Griffith University, Australia. She is recognised by international scholars as a leader in the field of literacy, middle leadership, educational action research, dialogic education and practice theory, evidenced by invitations to contribute to international research handbooks. She co-leads the Australasian Dialogic Education Network (ADEN) with colleagues from Auckland University. As a key researcher for the “Pedagogy Education and Praxis” International Research Network (PEP), she has contributed to international studies investigating how the education practices of student learning, teaching, professional learning and leading are ecologically connected in practices in schools. Her empirical and theoretical work in PEP investigates education practices from different intellectual and traditions (e.g., Scandinavia, Colombia, Netherlands, UK, Caribbean and Australia). Of note is a long-standing research collaboration with Professor Karin Rönnerman (Sweden) and Professor Peter Grootenboer (Griffith University) which has developed a strong empirical base for the theory of practice architectures and has led to the concepts of generative leadership and middle leading as a practice-changing practice.

Notes

1. Schatzki (2019) has continued this advocate that we ‘replace the micro-macro distinction of social analysis’ (p. 52) with a flat social ontology, which ‘denies the existence of fixed levels or hierarchies of phenomena in social life’ (p. 70).

2. Kemmis et al. (2014c) also use a lissajous pattern, with three linked loops, to indicate the relationship between the practices of leaders and teachers who ‘encounter one another amid [the third element of] practice architectures’ (p. 165).