People and practices in organizational learning

This article discusses the role of practices and people’s participation in practices in conceptual accounts of organizing, learning, and organizational learning. Specifically, the discussion takes its point of departure in Jean Lave and Etienne Wenger’s account of learning as legitimate peripheral participation in practices, and Theodore Schatzki’s practice theory account of organizing and organizations. Both accounts center on the role of practices as people come to know, and as changes occur in social activity and organizational settings. However, the two accounts are based on different ontologies. Borrowing the terminology of John Dewey and Arthur Bentley, Lave and Wenger instantiate a substantivist, and ultimately individualist, ontology, whereas Schatzki’s event ontology is relational. It is argued that both ontologies have merits of their own, but the article seeks to integrate the two approaches by utilizing Ole Dreier’s notion of the life trajectories of persons across social practices. In this perspective, organizational learning shows when people’s life trajectories are affected by the bundles of social practices they engage with, and when the bundles of social practices are transformed by the way people enact the


Introduction
In the 1990s, social theories of organizational learning were developed as a critical response to cognitivist conceptions that study organizational learning either as individuals' (cognitive) learning processes in organizational contexts (e.g. Argyris & Schön, 1978;March & Olson, 1987); or as a phenomenon in which organizations are conceived as entities that do the learning themselves (as wholes that learn analogically to individual learning) (e.g. Weick, 1991). In their seminal article, S.D. Noam Cook and Dvora Yanow (2011, p. 360) critique the perceived view of organizational learning: … we believe that organizational learning is not essentially a cognitive activity, because, at the very least, organizations lack the typical wherewithal for understanding cognition: They do not possess what people possess and use in knowing and learning-that is, actual bodies, perceptive organs, brains, and so forth.
In contrast to the individualist and cognitive model of organizational learning, Cook and Yanow envision organizational learning as a social and cultural phenomenon: … as a set of values, beliefs, and feelings, together with the artifacts and their expression and transmission (such as myths, symbols, metaphors, rituals), that are created, inherited, shared and transmitted within one group of people and that, in part, distinguishes that group from others (2011, p. 361).
Studying and conceptualizing organizational learning as participation in organizational practices has played a central role in social theories of learning over the last 25-30 years (Elkjaer, 2003;Brandi & Elkjaer, 2011;Gherardi, 2011). The introduction of the concept Communities of Practice (CoP) by Jean Lave and Etienne Wenger in the early 1990s has inspired many researchers to investigate organizational learning as a thoroughly social and situated activity (Lave & Wenger, 1991;Wenger, 1998;Brown & Duguid, 1996;Cook & Brown, 1999). More recently, theories of practices -or practice theory (PT) -have conceptualized organizations and organizing in terms of unfolding practices (Schatzki, 2002(Schatzki, & 2006Gherardi, 2006;Nicolini, 2012). Notably, the work of Theodore Schatzki has explicated organizations as bundles of social practices and material arrangements that are held together in specific ways as organizing happens.
This article sets out to discuss the role of practices and people's participation in practice(s) as conceptual accounts of organizing and learning. The article asks how organizational learning can be conceptualized through practice perspectives, in order to account for how people come to know and how organizational change comes about.
In what follows, the discussion takes its point of departure in Lave and Wenger's now classical account of learning as (legitimate peripheral) participation in CoP's, and Schatzki's PT account of organizing and organizations. Both of these accounts are 72 Forskning og Forandring centered on the role of practices as people come to know, and as changes occur in social activity. However, in their respective focus, the two accounts tend to construe coming to know in organizations in slightly different ways. I will lay out the central ideas in the two accounts and point to some of the commonalities and, more interestingly, investigate where they diverge in their mode of inquiry. Following John Dewey and Arthur Bentley's historical and conceptual analysis of modes of inquiry (1948/2008), I will discuss how the two approaches tend to understand their subject matter, and I will point to differences in focus and ambition among the two approaches that have implications for how organizational learning can be conceptualized and eventually studied.
These differences can easily be stipulated as absolute ontological discrepancies that set CoP and PT approaches apart. But, following Dewey's pragmatist philosophy, it can be argued that instead of construing (ontological) contrasts between the two, it is more productive to position the approaches as complementary: each theoretical approach helps address pertinent problems in relation to organizational learning that the other seems to neglect. Finally, Ole Dreier's attempt to transcend the narrow understanding of situatedness inherent in CoP and develop a theory of persons in structures of social practice is also considered as a potentially promising, but as yet underdeveloped, contribution to organizational learning research.

Organizational learning and communities of practice 1
Since Lave and Wenger introduced the concept of CoP it has indeed been used -and some would say misused -in a variety of ways serving different purposes. In reflecting on the notion of CoP almost 20 years after its introduction, Lave (2008, p. 283) mentions that the concept of CoP "… has taken on a life of its own, sometimes in felicitous and generative ways, but at other times in ways that give me a pause." Lave and Wenger intended CoP to be "… an informal label for a knot of ideas developed in the process [of analyzing situated learning]", but the label stuck and traveled into management -assisted by Wenger's subsequent writings (e.g. 1998;Wenger et al., 2002. See also an interview with Wenger-Trayner in Omidvar & Kislov, 2013). Today, CoP is an established concept that is taught in business schools and adopted in human resource management practices in companies as an organizational means to design knowledge sharing practices and stimulate organizational learning. However, Lave and Wenger originally introduced CoP in their efforts to theorize how learning could be understood in ways that did not presume the dichotomies of mind and body, rationality and emotions, etc. that are invested in the cognitive psychology account of learning. As indicated in the quote above, CoP was originally meant as an informal and epistemological label to understand how learning could be theorized in terms of participation in practice, and the notion of legitimate peripheral participation (Lave & Wenger 1991) was introduced as a productive theoretical perspective for conceptualizing learning and 1 This section draws on arguments and formulations presented in Buch (2015).
coming to know -and eventually processes of transformation in, among, and of practices (in organizations). Initially, therefore, CoP was intended to be an epistemological and analytical concept, but it has since 'traveled' and obtained still more ontological weight, both for some analysts in academia (including Wenger) and practitioners in business and industry. Many authors have expressed their worries in regard to this ontological drift of CoP and pointed to the inherent structural-functionalist heritage drawn upon when introducing 'community' into the equation (e.g. Nicolini, 2012).
As the CoP literature has focused primarily on understanding the mechanisms of learning as legitimate peripheral participation of individuals in groups where individuals are mutually engaged in a common enterprise and with a common history and repertoire, i.e. as a process where individuals belong in communities (Wenger, 2000, p. 203); the question of specifying the nature of the practices (of the communities), i.e. attending to the ontological status of practices, has remained undertheorized (cf. Gherardi, 2009). In this context, Brown and Duguid (2001, p. 203) observe that "[t]he appeal to community has tended to obscure the importance of practice …". In the CoP tradition, not much theoretical specification is offered in relation to practice(s), besides the fact that practice is social, situated, historical, material and cultural, and connotes the doings of the practitioners (Wenger, 1998, pp. 47-49 Stephen Kemmis et al. (2014, pp. 3-4)

Organizational learning and practice theory
Practice theory is not a unified theory, but should rather be seen as "… a broad family of theoretical approaches connected by a web of historical and conceptual similarities." (Nicolini, 2013, p. 1). 2 In this plethora of approaches, I draw on Theodore Schatzki's development of practice theory, as his account is widely referenced in the literature.
Schatzki's philosophical account of practice theory revolves around specifying a social ontology that highlights the nature and dynamics of human activity. Among practice theoreticians, Schatzki's contribution is, in my view, by far the most elaborate in terms of theorizing human activity and conceptualizing social practices. His exposition of practice theory is inspired by the early Martin Heidegger's event ontology (Schatzki, 2007) and the later Ludwig Wittgenstein's considerations on intelligibility (Schatzki, 1996), and sets out to understand human activity as unfolding in the form of indeterminate teleological actions in nexuses of practices and material arrangements (Schatzki, 2002(Schatzki, , 2010(Schatzki, , 2019). Schatzki does not develop a theory of (personal) learning (cf. Schatzki, 2017a, pp. 23-24), but his theoretical outlook makes it possible to theorize organizations and organizational change in ways that are relevant for understanding organizational learning (Schatzki, 2005(Schatzki, & 2006Loscher et al., 2019).
Schatzki conceives of organizations as meshes of social practices and material arrangements that are bundled in specific ways to continue to exist by the perpetuation and alteration of the involved practices and material arrangements (Schatzki, 2005, p. 476). In Schatzki's practice theory, the study of organizations is thus made possible as the study of social practices, the interplay among social practices, and the relation between social practices and material arrangements. Furthermore, in this perspective, the study of organizational learning will take the form of studying change Practices, in turn, are understood as organized human activity -as doings and sayings that are organized in specific ways. Practices are organized by the practical understandings of human agents, rules, teleoaffective structures and general understandings (Schatzki, 2002, pp. 77 ff.). Following Wittgenstein, Schatzki sees practical understandings as what makes the world intelligible -as sets of skills or capacities that underlie activity. Rules are "… explicit formulations, principles, precepts, and instructions that enjoin, direct, or remonstrate people to perform specific actions" (Schatzki, 2002, p. 77). Teleoaffective structures determine what makes sense to do; they are "… a range of normativized and hierarchically ordered ends, projects, and tasks, to varying degrees allied with normativized emotions and even moods." (Schatzki, 2002, p. 77). Finally, general understandings are senses of worth, value and nature that give an overall orientation and purpose to projects and tasks.
Organized as social practices, activities transpire amidst material arrangements.
In Schatzki's account, material arrangements are not part of social practices, but interact with practices in the sense that they can constitute, channel and prefigure social practices. Material arrangements comprise natural objects and artifacts, and they are essential to social practices in that they can afford and/or constrain the enactment of social practices. As an example, material arrangements, such as computers or case files, can make the continuation of certain organizational practices, such as e.g. accounting, easier or harder to enact. Taken together, the organization of social practices and material arrangements is configured in nexuses -and more extendedly -constellations of specific, but changeable, configurations. Doings, sayings, and material arrangements hang together in projects that in turn shape, and are shaped by, practice traditions.
In Schatzki's ontology, change and stasis are equally in need of explanation. Change is ubiquitous and fundamental in the plenum of practices and material arrangements, as events and processes unfold in social and material life: "… events and processes automatically introduce myriads of differences in the world." (Schatzki, 2019, p. x).
However, what amounts to 'change' -as opposed to mere and arbitrary differencesin this plethora of social and material transformations must be singled out by criteria of significance and relevance. As far as processes of organizational learning lead to transformations in and of organizations, they must thus be specified in relation to the significant differences that participants produce in relation to the organization as they learn.

Ontological issues
Even though the approaches of CoP and PT both focus on practice(s) when accounting for organizational learning and the study of organizational change, we have already Forskning og Forandring discerned some differences between the two accounts. Whereas CoP sees learning as integral to participating in (communities of) practices, Schatzki's version of PT does not comprise a theory of learning. There are good reasons why this is the case.
The theory of CoP's was developed as a theoretical account of (situated) learning -"… how learners inevitably participate in communities of practitioners and that the mastery of knowledge and skill requires newcomers to move toward full participation in the sociocultural practices of a community" (Lave & Wenger, 1991, p. 29). Wenger's subsequent development of the CoP theory (1998, pp. 143 ff.) in relation to issues of (personal) identity continued -and further stressed -the exploration of how individual persons (learners) come to know in (communities of) practice.
The focus of PT, in contrast, is not concerned with issues about how newcomers, or any other learners for that matter, come to know. Generally, it is about "… the constitution of social life: the nature of social existence, what it consists in, and the character of its transformation" (Schatzki 2002, xi) and, more specifically, it gives an account of organizational change and transformation (Schatzki, 2005(Schatzki, & 2006 Dewey and Bentley find one scheme that invokes inter-action as the central course of action: "where thing is balanced against thing in causal connection." (Dewey & Bentley, 1948 In opposition to inter-action, trans-action does not presuppose stable and unchanging substances in the explanatory scheme. Trans-action is a relational form of thinking about phenomena. Trans-actional approaches only accept descriptions and classifications tentatively and in a preliminary manner, "so that new descriptions of the aspects and phases of events, whether in widened or narrowed form, may freely be made at any and all stages of the inquiry" (Dewey & Bentley, 1948

Forskning og Forandring
For example, this recommendation would amount to not envisioning organizational actors as fixed entities with pre-existing interests, agendas, and objectives. Instead a trans-actional approach suggests conceiving actors as malleable and responsive practitioners that adapt to and impact their organizational environment in reciprocal processes (cf. Kriese, 1978).
Notably, though, Dewey and Bentley advise us to hesitate to elevate trans-actional modes of inquiry as superior to inter-actional modes: The issue is not baldly that of one or the other approach. It is not even an issue as to which shall be the basic underlying construction-since foundations in general in such questions are much less secure than the structures built upon them. It is, in view of the past dominance of the interactional procedure in most scientific enterprise, rather an issue of securing freedom for wider envisionment. (Dewey & Bentley, 1948).
Equipped with Dewey and Bentley's delineation of modes of inquiry, I will now turn to the approaches of CoP and PT to investigate their respective modes of inquiry.

Ontologies in communities of practice and practice theory
Unfortunately, the nature of the social in CoP is not explicitly discussed by Lave and Wenger. Nevertheless, it is possible to discern a social ontology in their account, since their investigations centre around how practitioners participate in, and form, CoP's where people learn.
For Lave and Wenger the unit of analysis is the 'newcomer' or 'apprentice' that comes to know (learn) in processes of legitimate peripheral participation in CoP's. As newcomers interact with old-timers, they acquire new skills, attitudes, and knowledge that eventually transform them as persons; through legitimate participation in the CoP, newcomers have the opportunity to become skillful, knowledgeable members of the community. Lave and Wenger (1991, p. 35) insist that "… learning is not merely situated in practice -as if it were some independently reifiable process that just happened to be located somewhere; learning is an integral part of a generative social practice in the lived-in world." Studying processes of learning becomes equivalent to studying how newcomers come to belong in CoP's: "The form that the legitimacy of participation takes is a defining characteristic of ways of belonging, and is therefore not only a crucial condition for learning, but a constitutive element of its content"

Anders Buch
The context of the learning process is the situatedness of the CoP. It is the interactions among the members of the community that transform individuals as they learn, and studying processes of learning is framed as the situated inter-actions of practitioners in CoPs. We see the mode of inquiry that Dewey and Bentley describe as inter-action as change and agency: learning is a phenomenon that is located in the "… Practice theory offers accounts of how the structure of the social world as practices delimits and defines the knowledges (and other items), the acquisition of which constitutes learning." (Schatzki, 2017a, p. 33). Practices are thus trans-actional in Dewey & Bentley's sense, in that they organize both knowing and the known, subject and object, in temporal-spatial processes of unfolding activities. PT is not a theory of learning and identity construction, but it is a theory that holds implications and resources for theoretical understandings of (personal and organizational) learning and identity.

Forskning og Forandring
It provides a vocabulary for understanding the social situatedness (i.e. the context) of learning paths in and among social practices and material arrangements. For PT, the context of certain forms of activities, e.g. learning activities, cannot be fathomed in local situations where interactional mechanisms are played out (in communities).
The context of these activities must rather be construed as sites, i.e. spaces or places where social events occur to become intelligible for the participants of these events as they happen. Sites are specific forms of contexts constituting and being constituted by their elements (practices and material arrangements). This means that the context for learning is not restricted to the (local) CoP, but must rather be understood as a particular way that the world is made intelligible through practices. Sites and practices are thus co-constitutive.
What is gathered from the above examination of CoP and PT is that the theories Are ontological differences problematic?
So far, the discussion has documented that the CoP and PT approaches differ in their modes of inquiry, and that they stipulate respectively a substantivist and a relational ontology. This indicates a fundamental discrepancy in their subject matter and 'first principles', i.e. their a priori assumptions about the constitution of the social, normally taken as a sign of theoretical incongruence, incompatibility, and contradiction. However, if we follow Dewey's lead, this conclusion is not warranted.
Dewey's pragmatism envisions 'the primacy of the practical' not primarily as a metaphysical principle that is stipulated to close an epistemological gap in human experience of the world, but rather as an anthropological and empirical claim about the problem-centredness of human endeavors (cf. Gimmler, 2018;Buch & Jensen, 2018). Dewey's instrumentalist approach urges us to see theories as tools that (social) scientists employ when they engage with the environment to make sense of what is happening in the world. Theories are thus more or less appropriate or useful according to the problems researchers wish to investigate and solve. Research does not need metaphysical ontological theory to get started (Dewey, 1938). In fact, Dewey argues that metaphysical commitments to specific ontologies might sometimes even hamper inquiry: When viewed from the standpoint of its position in the conduct of inquiry, the relativity theory rendered space and time themselves subjectmatters of inquiry instead of its fixed limits. In the Newtonian theory they had been treated as an Ultima Thule beyond which scientific inquiry could not possible go. These considerations may be used […] as an example of how submitting inquiry to ontological reference obstructs it. (original emphasis, Dewey & Bentley, 1948 Inquiry cannot, of course, escape ontological presumptions, in the sense that researchers always make judgements relative to a 'background' of ontological assumptions, but the scientific method, i.e. inquiry, must potentially submit all assumptions to critical scrutiny (although not simultaneously). For Dewey, ontological assumptions must be treated like any other assumptions, i.e. as representations of the world made by humans in specific contexts to serve specific purposes. Dewey's instrumentalism urges researchers to critically reflect on the usefulness of their theories and assumptions in relation to the problems under investigation, and revise them accordingly to fit the problems that they seek to solve: The different ontological assumptions of theories should thus not be seen as a mark of (absolute) incompatibility and conflict according to 'first principles', but rather as a sign of different priorities and focus according to 'last principles', i.e. according to the practical implications of the theories as tools for social research.
Complementary or integrative ontologies. Where to go with organizational learning?
The merits of CoP for theorizing organizational learning are already established in the literature (Lave & Wenger, 1991;Wenger, 1998;Brown & Duguid, 1996), and the resources of PT are also clearly pertinent for understanding processes of organizing and coming to know in organizations (Schatzki, 2005(Schatzki, & 2006Losche et al., 2019).
Although the two approaches rely on different ontological assumptions about the fabric of social reality, this divergence does not imply reciprocal exclusion.
Davide Nicolini (Nicolini, 2012, chapter 9;Nicolini & Monteiro, 2016) has convincingly argued in favor of an eclectic 'toolkit' approach that combines different sensitizing concepts in the study of (organizational) practice. The CoP approach, by zooming However, if we follow Dewey and Bentley's genealogical reasoning about modes of inquiry, we might eventually reach a more ambitious conclusion about the relationship between different practice theory approaches that deal with organizational learning.
Dewey and Bentley are clear that a trans-actional relational approach is not superior to inter-actional substantivist approaches. The important issue concerns "… securing freedom for wider envisionment" (Dewey & Bentley, 1948) that enables the continuous research process.
The Danish psychologist Ole Dreier has developed an account of learning that integrates insights from the literature of CoP and PT in giving an account of learning in structures of social practice (Dreier 2007(Dreier & 2008. Instead of situating CoP and PT as complementary theoretical accounts, Dreier synthesizes the theoretical perspectives by theorizing learning as personal (not individual) trajectories in social practice of: "Learning is seen as situated in social practice, and in the pursuit of most learning, persons are seen as moving around in social practice. They are involved in personal trajectories across various social contexts of practice." (Dreier 2008, p. 63).

Conclusion
In assessing the conceptual merits of CoP and PT for organizational learning, and the prospects of advancing the theoretical approaches to make them more relevant composed of sovereign individuals -aggregates of individuals -who learn to enter the interactional dances already available in organizations." (Kemmis et al., 2014, pp. 3-4), while still not decentring persons in the study of formations of social practices, as PT tends to do. Conceiving organizations as bundles of practices and material arrangements (Schatzki, 2005), while studying processes of learning as a social phenomenon that unfolds as people's life trajectories cross and intersect with social practices, enables us to delimit organizational learning as a social phenomenon that manifests itself when persons learn in bundles and constellations of social practices.
The phenomenon of organizational learning thus shows when people's life trajectories are affected by the bundles of social practices they engage with, and when the bundles of social practices are transformed by people's enactment of the practices -thus causing organizational change.
Organizational learning takes place when persons learn together in social practice and the organizational constellation of bundles of social practices are altered in significant ways by the persons who participate in the practices. This conception of organizational learning is transactional and processual in Dewey and Bentley's sense-it does not conceptualize persons (organisms) and organizations (environments of social practices) as distinct entities that inter-act. It construes the relationship between persons and organizations as co-constitutive in processes of trans-action. It acknowledges that people and practices are co-constitutive and mutually related in processes of transformation, and it avoids reducing organizational learning to a psychological phenomenon of cohorts of individuals that learn (as Argyris & Schön's (1978) and March & Olson's (1987) theorization of organizational learning tends to do). It also avoids conceiving organizational learning as organizational processes that are best understood by decentering people, and thus stipulating organizational learning as a phenomenon resulting from organizational entities that do the learning themselves (as Weick (1991) tends to do). Organizations and (bundles of) social practices should not be reified as entities that do the learning themselves. Dreier's acceptance of the ontologies of people (conceived as persons) and practices accentuates the transactional dance between persons and practices in organizational settings that constitutes processes of organizational learning.