From mediated actions to heterogeneous coalitions : Four generations of activity-theoretical studies of work and learning

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=hmca20 Mind, Culture, and Activity ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/hmca20 From mediated actions to heterogenous coalitions: four generations of activity-theoretical studies of work and learning Yrjö Engeström & Annalisa Sannino To cite this article: Yrjö Engeström & Annalisa Sannino (2020): From mediated actions to heterogenous coalitions: four generations of activity-theoretical studies of work and learning, Mind, Culture, and Activity, DOI: 10.1080/10749039.2020.1806328 To link to this article: https://doi.org/10.1080/10749039.2020.1806328 © 2020 The Author(s). Published with license by Taylor & Francis Group, LLC. Published online: 24 Aug 2020. Submit your article to this journal Article views: 213 View related articles View Crossmark data From mediated actions to heterogenous coalitions: four generations of activity-theoretical studies of work and learning Yrjö Engeström a and Annalisa Sannino b a University of Helsinki; b Tampere University ABSTRACT The article outlines four generations of theorizing and research on work and learning within the Finnish school of activity theory. The main focus of the paper is on the evolution of the unit of analysis through the four generations, from mediated action to a collective activity system, to multipleinterconnected activity systems, and most recently to heterogenous work coalitions aimed at resolving wicked societal problems. We examine how learning and agency are conceptualized in each generation, and what kinds of interventions are conducted in the projects presented as examples of the four generations. We conclude that the four generations of activitytheoretical studies of work and learning continue to co-exist. To meet the challenges of work in today’s world, activity theory must seriously engage in the development of its fourth generation, building on ideas and instruments developed by the preceding generations. Introduction Work activities were not a central focus in the empirical studies of the Russian founders of culturalhistorical activity theory. However, as activity theory has been adopted in western capitalist countries, studies of work have emerged as one of its main areas of application. In this article, we will examine the evolution and current challenges of these studies. Activity theory is pursued in multiple variations around the world. Our own work has evolved within what is sometimes called the Finnish school of activity theory. Initiated in the early 1980s by an informal group, the Finnish school has developed since 1994 in the Center for Research on Activity, Development and Learning CRADLE at University of Helsinki, and since 2018 also in the RESET research group at Tampere University. This approach is known for its modelling of activity systems as prime units of analysis, for its emphasis on the object-oriented and contradiction-driven character of activity, for the theory of expansive learning, as well as for the more recent methodology of formative interventions and studies of transformative agency by double stimulation. Much of the empirical work in CRADLE has focussed on work activities under the rubric of developmental work research (DWR; see Engeström et al., 2005). More recent topics include transformations in communities and social movements, and most recently in RESET the formation of heterogenous work coalitions tackling critical societal problems and forging practical alternatives to capitalism (Sannino, 2020, in press). In what follows we will trace the formation of central ideas and findings generated in the Finnish school. Obviously, other approaches within activity theory may find our ideas and arguments limited in various ways. However, rather than trying to cover all the variations of activity-theoretical research, we find it more useful to focus on our own lineage and theoretical perspective, as a contribution to further dialogue and collaborative discourse with other strands of work in activity theory and related approaches. CONTACT Yrjö Engeström [email protected] CRADLE, University of Helsinki, Sipoo 04130, Finland MIND, CULTURE, AND ACTIVITY https://doi.org/10.1080/10749039.2020.1806328 © 2020 The Author(s). Published with license by Taylor & Francis Group, LLC. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http:// creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way. In the next section, we will present the notion of four generations of activity theory and the lenses we will use to discuss examples of research representing the four generations. In the subsequent sections, we will examine the four generations one at a time, focusing on one major research project as an example for each generation. We will conclude with comments on methodological challenges for activity-theoretical studies of work. Four generations of cultural-historical activity theory The development of activity theory may be understood as a succession of four generations of theorizing and research. Each of the generations developed its own prime unit of analysis. The first generation was embodied in L. S. Vygotsky’s work. Even though Vygotsky occasionally wrote about “systems of activity” (Vygotsky, 1997b, pp. 20–21), he did not propose or conceptualize activity as a basic unit of analysis. We agree with Zinchenko’s (1985) conclusion that for Vygotsky the prime unit of analysis was that of culturally mediated action. In research on work, Scribner’s (1997) studies are powerful representatives of this generation. It was Leont’ev (1978, 1981) who worked out the second-generation unit of analysis, namely the concept of activity. Activity is a relatively durable system in which the division of labour separates different goal-oriented actions and combines them to serve a collective object. Object is what the activity is oriented toward. As the true motive of the collective activity, the object gives activity its identity and direction. The object is durable and constantly under construction; it generates a perspective for possible actions within the activity. As such, the object is not reducible to conscious goals; those are connected to discrete and relatively short-lived actions. The object of an activity is typically difficult to define for the participants. Medical practitioners may agree that the object of their work activity is health and illness, but if asked to specify the object further, each practitioner tends to give a somewhat different characterization, depending on the personal history of the individual and his or her position in the division of labour within the activity system. An activity system is more than a mechanical sum of its components. An activity weaves together its own dynamic context, “in activity theory [. . .] contexts are activity systems. The subsystem associated with the subject-mediator-object relationship exists as such only in relationship to the other elements of the system. This is a thoroughly relational view of context” (Cole, 1996, p. 141). The studies reported in the collection Putting Activity Theory to Work (Engeström et al., 2005) largely represent the second generation. Typically, the focus is on the analysis of a single, relatively well-bounded activity system and its transformations. As activity systems are increasingly interconnected and interdependent, many recent studies take as their unit of analysis a constellation of two or more activity systems that have a partially shared object. Such interconnected activity systems may form a producer–client relationship, a partnership, a network, an alliance, or some other pattern of multi-activity collaboration. The formation of minimally two activity systems connected by a partially shared object may be regarded as the prime unit of analysis for third-generation activity theory. Increasingly complex runaway objects with broad societal ramifications, such as climate change or pandemics, connect large numbers of activity systems across national borders (Engeström, 2009). Such objects tend to transcend the boundaries between the history of a specific activity, the history of a singular society, and the history of humankind. The emerging fourth generation of activity theory zooms on heterogenous work coalitions aimed at resolving critical societal problems, or runaway objects, and creating sustainable alternatives to capitalism (Sannino, 2017, 2020; Sannino & Engeström, 2018). The four generations share certain foundational ideas. All of them see that work needs to be analyzed as object-oriented practice, mediated by instruments, and changing through its inherent contradictions. Work is to be understood in its constant development and transformations, making learning a central aspect of work. Transformative agency and willful action are of crucial importance 2Y. ENGESTRÖM AND A. SANNINO in performing and shaping work. The four generations agree that formative interventions aimed at finding new possibilities and potentials are the most appropriate way to study work. In the following sections, we examine the four generations of activity-theoretical studies of work. Our main focus is on the evolution of the unit of analysis through the four generations. We will also pay attention to how learning and agency are conceptualized in each generation, and what kinds of interventions are conducted in the projects presented as examples of the four generations. The first generation: discovering historicity in work actions An early study of the Finnish school of activity theory had focussed on the work and thinking of janitorial cleaners employed by a large commercial cleaning company (Engeström & Engeström, 1984, 1986). Their task was to clean offices and other facilities, following carefully measured time schedules and work instructions. The researchers videotaped a number of cleaners working on standard office rooms. Some of their ways of working struck the researchers as peculiar. For example, when the practitioners were asked to vacuum the floor of an office room, many of them first dragged the vacuum cleaner to the rear end of the room, then started vacuuming from there, moving clumsily backwards toward the door. When they were asked to mop the floor, many of them would move the mop on the floor as they were if washing the floor rather than moving loose dirt toward a pile with the help of the mop. When these actions were shown to the practitioners on the video, some of them were embarrassed and laughed at their own performance. When asked why they were conducting the actions in these ways, several practitioners referred to the ways they had learnt at home. I always start from the farthest corner. At home, too. The same with mopping. So that I don’t leave my own footprints there. Well, in vacuum cleaning you actually don’t leave footprints, but I am used to that. At home I have a floor where even a single drop of water shows. I guess I’ve learned from that. [Subject 11] (Engeström & Engeström, 1986, p. 12) The researchers interpreted these findings as evidence of layers of history actively influencing the present-day actions of the practitioners. In the European scale, the strengthening of cleaning was obviously connected with the rise of the bourgeoisie and the protestant ethics, placing high premium on discipline, purity and order. In Finland, the first decades of [the 20th] century were a period of intensive propaganda for modern home-making and regular weekly cleaning, against the spread of diseases (especially tuberculosis) and parasites. This resulted in the breakthrough of the model of home cleaning which has persisted till today. The central idea is to clean all furniture and surfaces, especially the floors, once a week (on Saturday, nowadays on Friday) very thoroughly, with plenty of water when possible. Now the objective basis for this model has vanished as general conditions of hygiene and medical care as well as the quality of housing have improved. Thus, we are left with the persistent form of the model of home cleaning, turned more or less into a ritual. The original struggle against diseases and parasites has been replaced with the objective of absolute visible cleanness and orderliness – an elusive objective that can be reached only momentarily. [. . .] The rules of home cleaning are those of regularity and status quo. Cleaning takes place regularly and is composed of regular components; its unspoken norm is to restore the order and appearance which have been disturbed during the week after the previous cleaning. All this has predominantly tacit, implicit character.” (Engeström & Engeström, 1986, p. 6) In commercial cleaning, a radically different model has emerged. Wage labourers with minimal specific training perform this work. Effective floor cleaning machinery and powerful chemicals are used, and standardized written work instructions have replaced the tacit traditions of home cleaning. The object of work is now a carefully measured and normed “cleaning area”, and the outcome is an agreed-upon “appropriate level of cleanness”. The worker is a member of a centralized cleaning service. The rules are those of evening-shift wage labour. This mass-production model of cleaning has led to impressive increases in efficiency, but also to increasingly aggravated contradictions of its own. MIND, CULTURE, AND ACTIVITY 3 The historical model of craft-like home cleaning influenced, sometimes dominated, the everyday actions of the cleaners working in an environment of mass production. This created a tension which the research team at the time called “the cleaner’s bad conscience”. Workers repeatedly expressed their dissatisfaction with the kind of cleanness they were able to achieve in their work. They saw their work as degenerated home cleaning. This led them to try more than was prescribed in the work instructions: more repetitions, more efforts at reaching visible ’absolute cleanness’ in the image of weekly home cleaning. This often meant excessive stress and time pressure at work, connected to mounting health problems. So should cleaners be taught to reject home cleaning and adapt to the norms of mass production? That would lead to submission to Taylorization and increasing pressure on the pace of work, consequences of capitalist modernization in its patronizing and oppressive sense. The researchers found it more meaningful to invite the cleaners to identify emerging new possibilities to move beyond mass production. The example of the cleaners contains a general lesson. History is always present in human activity. Layers of historically earlier forms of the activity can be both constraints and resources. They persist in routine actions, in ways of thinking, in material artefacts and in rules. If one tries to understand actions without historicity, consequential phenomena such as the cleaners’ bad conscience are easily dismissed as arbitrary irrational features, even pathologies, of certain individuals or classes of people, to be eliminated or, at best, ignored. The unit of analysis in the cleaning study was mediated work action. The structure of such an action was represented with the help of Figure 1 (Engeström & Engeström, 1984, p. 17). Focus on work actions nourished detailed studies of actual work performances and their cognitive requirements. Variation and anomalies found in work actions opened up an avenue for a historical understanding of work. Focus on action highlights the possibility of subjectification of the worker. The worker is seen as a subject who can design his or her own ways of performing the required actions. In the cleaning study, the model of mediated action was primarily an instrument for the researchers’ analyses. To our surprise, we found occasions in which cleaners themselves started on their own initiative to use the model to discuss their work. An obvious limitation of action as a unit of analysis is that it does not explicitly address the social relations and organizational embeddedness of work actions. In other words, it runs the risk of locating explanations for disturbances and problems as well as innovations and emancipation in work exclusively in the individual worker. In the cleaning study, learning was analyzed with the help of Norman’s (1982) framework of three types of learning: accretion, structuring, and tuning. This framework works well when one examines the internalization of well-defined skills, such as the standard actions of cleaning work. The framework may not be very useful if one analyzes workers’ efforts to transform their work and to generate qualitatively new actions (Engeström & Sannino, 2012). Figure 1. Mediated work action as a first-generation unit of analysis in the cleaning study (Engeström & Engeström, 1984, p. 17). 4Y. ENGESTRÖM AND A. SANNINO In the cleaning study, workers’ cognitive orientations to work were seen as the key to the agency. It was assumed that if workers understood the historically evolving nature of their work actions, including the contradictions behind troublesome experiences such as the cleaner’s bad conscience, they would be able and willing to move forward, beyond the restrictive orientations of both home cleaning and rationalized cleaning. The intervention conducted in the project consisted of a three-day training course aimed at teaching the cleaners to analyze and design their own cleaning actions. The new orientation was built on an expanded notion of the object, based on conscious integration of three requirements: healthiness or hygiene, functionality and economy, and visible cleanness (Engeström & Engeström, 1986). Three months after the course, critical work actions of the cleaners who participated in the course were significantly more in line with the new orientation than those of a comparison group (Engeström, 1995, pp. 167–168). The second generation: trancing contradictions and expansive possibilities in systems of work activity In the early 1990s, a large project on the work of municipal courts and their judges were carried out by researchers of the Finnish school of activity theory (Engeström, 1995, 1996; Engeström et al., 1991, 1992; Haavisto, 2002). This project is an example of the second generation of activity theory at work. The project was aimed at helping two pilot courts to implement locally the general guidelines of a nationwide court reform that would be legally enforced in all courts a few years later. Implementation was understood as a process of experimentation, innovation and learning (Pressman & Wildavsky, 1984). The researchers followed, recorded and analyzed trials before the reform and subsequently also after the reform. In particular, the researchers tried to identify and cultivate creative innovations devised and used by judges in anticipation of the full-scale reform. The court proceedings in Finland were traditionally very formal and non-interactive, based on lengthy written briefs read aloud in front of the judge. At the same time, the judges traditionally left it to the attorneys to decide what issues to cover and how long to continue with the exchange of briefs. In other words, the old proceedings were both formal and unrestricted. This often meant multiple hearings in a single court case, extended over a period of several months. The new legislation aimed at proceedings based on informal oral discussion actively controlled and guided by the presiding judge. This should lead to a compact process in which all points of view come to light in one and the same hearing. This transformation was intertwined with a transition from the traditional notion of justice as material truth to the idea of negotiated justice and pragmatic compromise. A crucial aspect of second-generation studies is an analysis of the historically formed contradictions in the work activity. This requires that the unit of analysis is expanded to encompass the entire collective activity system. The generic structure of an activity system was modelled by Engeström (1987, p. 78, 2015, p. 61). In second-generation studies, the model of the activity system is commonly used interactively between the researcher-interventionists and the practitioners to analyze the past, present and future of the activity. In the model, the subject refers to the individual or subgroup whose position and point of view are chosen as the perspective of the analysis. Object refers to the raw material or problem space at which the activity is directed. The object is turned into outcomes with the help of instruments, that is, tools and signs. The community comprises the individuals and subgroups who share the same general object. Division of labour refers to horizontal division of tasks and vertical division of power and status. Rules refer to the explicit and implicit regulations, norms, conventions and standards that constrain actions within the activity system. The object of the activity is an invitation to interpretation, personal sense making and societal transformation. One needs to distinguish between the generalized object of the historically evolving activity system and the specific object as it appears to a particular subject, at a given MIND, CULTURE, AND ACTIVITY 5 moment, in a given action. The generalized object is connected to societal meaning, the specific object is connected to personal sense (Leont’ev, 1978). The analysis of the activity system of one of the courts, Vantaa District Court, led to the representation in Figure 2. Systemic contradictions in the activity are depicted with the help of lightning-shaped double-headed arrows. There was a polarization of court cases: both the number of simple routine cases and the number of complicated cases were increasing and overshadowing the “ordinary cases”. The growing number of cases connected to business life meant more demanding and complicated issues. The judges also told in their interviews of pressures to make fast decisions and to produce more grounded justifications for their judgments. Despite this transformation in the object of the court, the rules guiding the proceedings remained the same. The contradiction between the rules and the object was expressed in prolonged court processes, where numerous postponements compelled the judge to return to a case several times. This fragmentation diminished the judges’ possibilities to master the proceedings, and compromised the quality of decision-making. This mechanism produced excessive workload and stress experienced by the judges, sometimes leading to serious health issues and even decisions to leave the court for less burdensome forms of legal work. The rules were also internally contradictory: alongside with the vestiges of the old established practice, there were pressures to take into account the forthcoming norms of the new legislation, especially the active role to be taken by the judge in the courtroom. Another basic contradiction in the court activity appeared between the changing object of the work and the rigidly segmented division of labour. The judges were located in their departments, the office personnel in their offices, and the court trainees, who still provided an important labour resource, were scattered around different departments and offices. The potential for collaboration among different personnel groups was not efficiently exploited. The long-established historical tradition of the judge working alone was evident: the judges in Vantaa District Court appeared to be individual judges working alone on cases allocated to them. In the courts project, the concept and model of expansive learning (Engeström, 1987; Engeström & Sannino, 2010) was systematically used. Learning was now understood as a collective process of creating and acquiring something that is not yet there. An expansive learning process proceeds from Figure 2. The activity system of Vantaa District Court and its contradictions in 1990 as a second-generation unit of analysis in the court study (Engeström et al., 1992, p. 154). 6Y. ENGESTRÖM AND A. SANNINO questioning the existing practice to analyzing it, then to modelling, examining and implementing a new solution (Figure 3). To transcend or resolve the contradictions depicted in Figure 2, Haavisto (2002) traced the possibilities of expansive learning in trials conducted in the Vantaa court. She identified incremental innovations, or “spearheads of transformation,” generated by a number of judges (Figure 4). These included trials in which the clients (not just attorneys) began to take active initiatives, and proceedings in which the judges intervened actively to promote settlements between the parties. In these spearheads, new ways of talking emerged, such as instructional talk of the judge directed at lay clients taking initiatives in the hearing. At the same time, new tensions also emerged with the spearheads, such as a tension between increased client initiatives made possible by the informality and the increased emphasis on active control and leadership on the part of the judge (arrow A in Figure 4). Arrow B in Figure 4 represents a tension between the intrinsically formal character of the talk of legal professionals and the informal character of the talk of lay clients. Arrow C represents a tension between controlled/restricted and uncontrolled/unrestricted discourse in the hearings, inherent in the alternative logics that were used to legitimate the new proceedings: “the first one assumes that the truth will emerge in free and unlimited communication, the second that it will evolve in communication that is regulated and rationalized by procedures” (Haavisto, 2002, p. 295). Whereas the intervention in the cleaning study took the shape of a short and rather straightforward training course, in the courts project the intervention was multi-faceted and lasted 3 years. The researchers conducted a series of workshops in each court, inviting the practitioners to analyze and redesign their work activity. For example, in the Vantaa court, five multi-professional planning groups were formed, with the task of designing different aspects of a new overall working model for the court. The first one of the planning groups was responsible for designing a new organization for the court. The group met 18 times, including one meeting organized for the entire personnel of the court to discuss the group’s provisional proposals. Like many of the interventions in second-generation studies, this process contained various complementary learning events and sequences. Some of them were formal training sessions organized by the researchers, others were more informal design Figure 3. The cycle of expansive learning (Engeström & Sannino, 2010, p. 8). MIND, CULTURE, AND ACTIVITY 7 workshops led by judges who took the initiative to promote peer learning (Engeström, 1995; Haavisto, 2002). In the second generation of activity theory, the agency was mainly an implicit notion, understood as expansive movement from isolated individual subjects engaged in their specific work tasks toward collaborative and collective subjects engaged in transforming the entire work activity. This movement is driven by recurring disturbances and troubles generated by historically accumulated inner contradictions in the activity system. Power becomes something that can be generated from below, by grasping the contradictions and by re-forging the activity to transcend the contradictions. Individuals gain agency and power by joining their efforts and constructing what Leont’ev characterized as “motive-goals”, referring to a merger of conscious goals of individual actions and the motive of the entire collective activity: “(. . .) a different fate is created when the principal motive-goal is elevated to a truly human level and does not weaken man but merges his life with the life of people, with their good” (Leont’ev, 1978, p. 134). This implies re-orchestration of social relations at work, creation of generative microcosms of collaboration and design of alternative futures. The third generation: negotiated knotworking among multiple activity systems Discontinuity and fragmentation of care is a critical issue in hospital work, especially in care for patients with multiple chronic illnesses. In the late 1990s, researchers of the Finnish school of activity theory were invited to conduct a project hosted by the Children’s Hospital in Helsinki, aimed at improving the coordination and continuity of care for children with chronic illnesses. The problem was particularly acute among children with multiple or unclear diagnoses. Such children often drifted between caregiver organizations without anyone having an overview and overall responsibility of the child’s care trajectory. This put a heavy burden on the families (Engeström, 2001). The Children’s Hospital study was a formative intervention project which used a version of the Change Laboratory method (Engeström et al., 1996), called Boundary Crossing Laboratory. 1 Figure 4. Expansive spearheads in the zone of proximal development of court work (Haavisto, 2002, p. 298). 8Y. ENGESTRÖM AND A. SANNINO transformative agency develops to the point that the staff member discovers new capabilities he had ignored before and qualities in the resident that he had equally ignored (Sannino, in press). Nobody told the staff member to have a cup of coffee with the resident. This is a volitional action taken by the staff member in response to the conflict of motives he or she is experiencing. Essentially the transformative agency process is a deliberately initiated conditioning process by an individual or a collective who resolves to learn a new way of working. This is crucially a self-designed and selfenforced conditioning (rather than a classic behavioristic one) which builds up into novel modes of action and expansive learning. On the basis of 25 years of activity-theoretical research using Change Laboratory formative interventions (Sannino & Engeström, 2017; Sannino et al., 2016b) in which double stimulation has been systematically implemented, we have strong empirical evidence to claim that the stepwise expansive learning cycle and the transformative agency process mutually induce one another. As we embark on fourth-generation intervention studies, this connection gains new significance. The intertwined learning cycles are longer and more complex, typically riddled with obstacles and setbacks. In these conditions, the process of transformative agency by double stimulation is of critical importance for sustained expansive learning to occur and needs to be mobilized throughout the intersecting cycles of expansive learning (Sannino, 2020, in press). As already indicated above, the first Change Laboratory intervention of the homelessness project was conducted with the entire staff of a housing unit that embarked on an effort to reorganize itself in 2018. This intervention is analyzed in more detail by Sannino (2020, in press). The unit offers supported housing for young residents, each having his or her own apartment and rental contract. It has nearly 100 apartments for a clientele that is considered the hardest to find housing for, due to criminal records, debts, substance abuse and mental health problems. Staff is present 24/7. The unit is located in a well-off, politically conservative neighbourhood and has received negative press frequently ever since its opening. In public debate, the residents are often considered as “addicts” who are deemed “everyone’s least favorite neighbors”. In the housing unit, the staff characterized their zone of proximal development in terms of moving from “being a controller to being a coach and a fellow traveler”. Such a transition was a demanding effort as the tensions inherent in the traditional model of guarding and controlling were identified and worked out by the staff in the Change Laboratory sessions. The participants analyzed the change process they were carrying out in the housing unit, using the model of the cycle of expansive learning as their instrument for analysis. They have modelled their new way of working and defined the direction of its development. To continue the development, the participants identified 15 spearhead projects whose implementation is followed and supported. The residents largely welcomed the change as a perspective toward regaining their self-esteem and self-determination. The results of the transformation process in the unit have been rather remarkable and it has become a model for many other units in the nation aiming at similar transformations. Two related Change Laboratories followed in 2019, one with city level actors in the city of Tampere and the other one with state-level actors. Each Change Laboratory consisted of a series of six sessions in which the research team and the participants jointly analyzed factors that hamper successful outcomes with homelessness and co-designed novel solutions. The videotaped sessions are transcribed for analyses using discourse analysis methods devised for conversations taking place in longitudinal Change Laboratory interventions (Engeström et al., 2013; Engeström & Sannino, 2011; Sannino, 2008; Sannino et al., 2016b). With regard to the role of the participants as initiators and shapers of the new model of activity, the three CLs discussed here were similar to the CLs of earlier generations of CHAT. In the first CL conducted in the housing unit – the one already analyzed in some detail – the analysis of the current contradictions, the zone of proximal development, the new model of the activity, and the spearheads for next steps of development were all co-constructed and decided upon by the participants, with the interventionist-researchers in the role of provoking, supporting and MIND, CULTURE, AND ACTIVITY 15 recording their efforts (Sannino, 2020, in press). There were several episodes in which the participants rejected or substantively altered suggestions coming from the researcherinterventionists. The three Change Laboratories were intertwined in various ways. A practitioner from the local housing unit presented experiences of the first CL and answered questions at one of the city level CL sessions. One practitioner from the housing unit CL and one from the city level CL participated in the national level CL. Video clips from the housing unit CL and city level CLs were used as stimuli for joint analyses and design efforts in the national level CL’s intervention sessions. The multiple ways in which the three interventions and learning cycles interacted require careful analysis which is currently underway. The interventions at the ground level of the housing unit and at the intermediate level of the city served as impulses for modelling solutions for the challenges to be met at the national level. This crosslevel and cross-sector interaction and coalition building represents a new challenge and opportunity for expansive learning at work (Sannino, 2020). Also, these intertwined Change Laboratories initiate further learning cycles aimed at establishing novel heterogeneous coalitions toward the next steps of the strategy. These dynamics demand careful studies and new intermediate conceptual tools to be developed within activity theory. Expansive learning in this context consists of joint reconceptualization efforts by a flexibly developing coalition of diverse actors pursuing the eradication of homelessness and realization of the functional independent living of each former homeless person in a home of one’s own. This is a societally contradictory and tension-filled object, due to the financial costs and required extensive transformation of mindsets, practices and material arrangements. The different actors are driven by their different partial objects; the challenge is to enable them to debate, negotiate and co-design the emerging overall object so as to reach sufficient coherence, stability and motivational force to generate determined and resilient collective activity over a lengthy period of time. Strong agentive actions of commitment and material implementation are needed to pursue such an object (Sannino, 2008, 2010). The concept of transformative agency by double stimulation (TADS) (Sannino, 2015a, 2020, in press) aptly captures this type of willful pursuit of enacting utopias for the common good. Fourth-generation studies involve a radical expansion of social relations. Dialogue, trust and collaboration need to be built among large numbers of diverse activity systems and their actors, crossing professional, functional and administrative boundaries both horizontally and vertically. We are only beginning to develop conceptual tools and methods to analyze and foster these processes of generating dynamic social cohesion around a shared object among heterogeneous activities. The notion of fourth-generation activity theory presented here consists of a research-based inquiry conducted in close collaboration with key operative stakeholders working for and with the homeless from the ground level of frontline work to the city level and the state level. The role played by the research team represents ways in which theoretical tools can contribute to bringing together multiple parties and serve as mediators for collaborative analyses to meet critical societal transformations. The approach may, in the long run, contribute to the establishment of globally interlinked experimental hubs across universities and their partners, focused on pressing challenges of our time. Conclusion Our account of the evolution of activity-theoretical research through the four generations may be summarized with the help of Table 1. The four generations of activity-theoretical studies of work continue to co-exist. To meet the challenges of work in today’s world, activity theory must seriously engage in the development of its fourth generation. But a fourth-generation study builds on ideas and instruments developed by the preceding generations. The intertwined learning cycles depicted in Figure 6 are intimately connected to the interacting activity systems depicted in Figure 5, and the analyst needs to move between the two. 16 Y. ENGESTRÖM AND A. SANNINO Detailed studies of critical work actions pioneered by the first generation do not lose their relevance – but they need to be done in a broader perspective of the activity systems and societal challenges in which the actions are embedded. The triangular models of the units of analysis of the first three generations have proven to be effective instruments for participatory collective analysis and design. Can the proposed fourthgeneration unit of analysis (Figure 6) serve practitioners, communities and researchers in the same way? We believe it can. The first detailed analyses of the interplay of coalescing expansive cycles are in progress, and many new methodological problems need to be solved. We already have evidence that the closely related model of transformative agency by double stimulation (Figure 7) is serving well in interventions and analyses (Hopwood & Gottschalk, 2017; Jalasi, 2020; Sannino, in press). From the very beginning, activity theory has prioritized “transforming experiments” (Bronfenbrenner, 1977) or “genetic-modeling experiments” (Zuckerman, 2012) as methods of research that go beyond the given and generates new, emancipatory forms of activity. We call these methods formative interventions. They are the core of the emerging methodology of fourthgeneration activity-theoretical research (Engeström et al., 2014). Here methodology is understood as the bridge between theory and data. In other words, the methodology is more than a collection of specific methods or techniques. It puts forward and implements a theory-driven set of principles, or “an argumentative grammar” (Kelly, 2004), upon which the choice of specific methods is based, starting from data collection and reaching all the way to conceptual interpretation of the findings. The argumentative grammar of activity-theoretical methodology is condensed in the principles of double stimulation (Sannino, 2015a, 2015b) and ascending from the abstract to concrete (Engeström et al., 2012). Table 1. Overview of the four generations. First generation Second generation Third generation Fourth generation Object and problem Challenge in individual learning or development Collective developmental contradictions demanding on espansive solution Developmental contradictions within and between interconnected activity systems A critical societal challenge or crisis demanding a multilevel and cross-sectoral solution Unit of analysis Mediated action Collective activity system Minimally two interacting activity systems with a partially shared object Coalescing cycles of expansive learning in a heterogenous coalition of activities facing a critical societal challenge Concept of learning Internalization of given skills and knowledge Expansive learning cycle generating what is not yet there Expansive learning cycle involving boundary crossing and horizontal sideways learning Horizontal and vertical interplay between multiple coalescing cycles of expansive learning Concept of agency Agency as grasping the historically evolving nature and emancipatory possibilities of one’s actions Agency as expansive movement from individual subjects and their tasks toward collective subjects transforming their activityn Agency as recognition and negotiation of differences and complementarities Transformative agency by double stimulation Typical intervention Training aimed at emancipatory understanding and mastery of one’s actions Longitudinal process of collective analysis and redesign of the activity → emergence of the Change Laboratory method Change Laboratory and Boundary Crossing Laboratory Multiple interconnected Change Laboratories, from local to municipal, regional, national and international levels – with longitudinal follow-up MIND, CULTURE, AND ACTIVITY 17 Aimed at expansive learning, Change Laboratory interventions put into use the principles of double stimulation and of ascending from the abstract to the concrete, typically with a carefully planned sequence of tasks and learning actions. This does not mean that the interventionist’s plan is smoothly implemented. To the contrary, participants in formative interventions commonly take over the process at some point and generate deviations from the interventionists’ intentions. These deviations reveal gaps between the interventionists’ object and the participants’ objects – gaps that need to be negotiated. The deviations, gaps and negotiations are important instances of emerging transformative agency among the participants (Engeström & Sannino, 2012). There is a global push toward formative interventions in multi-activity constellations and coalitions which may include local communities, social movements, educational institutions, private companies, public service agencies, non-governmental associations, and policy-making administrative and political bodies (Gray & Purdy, 2018; Wiewiora et al., 2018). Not accidentally, collective initiatives for such bold formative interventions are increasingly coming from the global south (Lotz-Sisitka et al., 2017; Vilela, 2019). The need for these interventions typically stems for contradictions connected to the quest for social and economic equity and ecological sustainability. Such fourth-generation Change Laboratories face methodological and practical challenges that are different from those of earlier CLs and require concerted research efforts in their own right. The founders of activity theory, from Vygotsky to Leont’ev, Il’enkov and Davydov, faced revolutionary challenges and also oppressive counter-forces (Sannino, 2011). Today the revolutionary challenge for activity theory is to develop and put to use its conceptual foundations and methodological solutions in the service of formative interventions in multi-activity constellations aimed at creating and enacting alternatives to capitalism. This is also a way to deepen our understanding of human potentials at work in the cultural-historical conditions of the 21 st century. Note 1. We first used the notion of boundary crossing in work on expert cognition (Engeström et al., 1995). Subsequently, Kerosuo (2003) developed the concept in studies of health care. By the notion of boundary, we refer to both formal and practically enacted definitions and limits of membership, competence and authority of organizations, functions or professions. Boundaries typically become visible and articulated when actors try to access something on the other side of the boundary and encounter obstacles or constraints in this quest. Disclosure statement No potential conflict of interest was reported by the authors. References Bratus, B. S., & Lishin, O. V. (1983). Laws of the development of activity and problems in the psychological and pedagogical shaping of the personality. Russian Education & Society, 21(3), 38–50. Bronfenbrenner, U. (1977). Toward an experimental ecology of human development. American Psychologist, 32(7), 513–532. https://doi.org/10.1037/0003-066X.32.7.513 Cole, M. (1996). Cultural psychology: A once and future discipline. Cambridge University Press. Engeström, Y. (1987). Learning by expanding: An activity-theoretical approach to developmental research. OrientaKonsultit. Engeström, Y. (1995). Kehittävä työntutkimus: Perusteita, tuloksia ja haasteita [Developmental work research: Foundations, results, and challenges]. Valtionhallinnon kehittämiskeskus. Engeström, Y. (1996). The tensions of judging: Handling cases of driving under the influence of alcohol in Finland and California. In Y. Engeström & D. Middleton (Eds.), Cognition and communication at work (pp. 199-232). Cambridge University Press. Engeström, Y., Engeström, R., & Vähäaho, T. (1999). When the center does not hold: The importance of knotworking. In S. Chaiklin, M. Hedegaard, & U. J. Jensen (Eds.), Activity theory and social practice: Cultural-historical approaches (pp. 345–374). Aarhus University Press. 18 Y. ENGESTRÖM AND A. SANNINO Engeström, Y. (2001). Expansive learning at work: Toward an activity theoretical reconceptualization. Journal of Education and Work, 14(1), 133–156. https://doi.org/10.1080/13639080020028747 Engeström, Y. (2008). From teams to knots: Activity-theoretical studies of collaboration and learning at work. Cambridge University Press. Engeström, Y. (2009). The future of activity theory: A rough draft. In A. Sannino, H. Daniels, & K. D. Gutiérrez (Eds.), Learning and expanding with activity theory (pp. 303–328). Cambridge University Press. Engeström, Y. (2015). Learning by expanding: An activity-theoretical approach to developmental research (2nd ed.). Cambridge University Press. Engeström, Y. (2018). Expertise in transition: Expansive learning in medical work. Cambridge University Press. Engeström, Y., Brown, K., Christopher, C., & Gregory, J. (1991). Coordination, cooperation and communication in courts: Expansive transitions in legal work. The Quarterly Newsletter of the Laboratory of Comparative Human Cognition, 13, 88–97. Engeström, Y., & Engeström, R. (1984). Siivoustyön hallinta ja työntekijöiden laadullinen koulutustarve [The mastery of janitorial cleaning work and the workers’ qualitative need for training]. ServiSystems. Engeström, Y., & Engeström, R. (1986). Developmental work research: The approach and the application of cleaning work. Nordisk Pedagogik, 6, 2–15. Engeström, Y., Engeström, R., & Kärkkäinen, M. (1995). Polycontextuality and boundary crossing in expert cognition: Learning and problem solving in complex work activities. Learning and Instruction, 5(4), 319–336. https://doi.org/10. 1016/0959-4752(95)00021-6 Engeström, Y., Haavisto, V., & Pihlaja, J. (1992). Alioikeudet uuden työtavan kynnyksellä: Kehittävän työntutkimuksen sovellus tuomioistuinlaitoksessa [Municipal courts facing a new way of working: An application of developmental work research in the court system]. Yliopistopaino. Engeström, Y., Lompscher, J., & Rückriem, G. (Eds.). (2005). Putting activity theory to work: Contributions from developmental work research. Lehmann’s Media. Engeström, Y., Nummijoki, J., & Sannino, A. (2012). Embodied germ cell at work: Building an expansive concept of physical mobility in home care. Mind, Culture, and Activity, 19(3), 287–309. https://doi.org/10.1080/10749039.2012. 688177 Engeström, Y., Rantavuori, J., & Kerosuo, H. (2013). Expansive learning in a library: Actions, cycles and deviations from instructional intentions. Vocations and Learning, 6(1), 81–106. https://doi.org/10.1007/s12186-012-9089-6 Engeström, Y., & Sannino, A. (2010). Studies of expansive learning: Foundations, findings and future challenges. Educational Research Review, 5(1), 1–24. https://doi.org/10.1016/j.edurev.2009.12.002 Engeström, Y., & Sannino, A. (2011). Discursive manifestations of contradictions in organizational change efforts: A methodological framework. Journal of Organizational Change Management, 24(3), 368–387. https://doi.org/10. 1108/09534811111132758 Engeström, Y., & Sannino, A. (2012). Whatever happened to process theories of learning? Learning, Culture and Social Interaction, 1(1), 45–56. https://doi.org/10.1016/j.lcsi.2012.03.002 Engeström, Y., Sannino, A., & Virkkunen, J. (2014). On the methodological demands of formative interventions. Mind, Culture, and Activity, 21(2), 118–128. https://doi.org/10.1080/10749039.2014.891868 Engeström, Y., Virkkunen, J., Helle, M., Pihlaja, J., & Poikela, R. (1996). Change laboratory as a tool for transforming work. Lifelong Learning in Europe, 1(2), 10–17. Gray, B., & Purdy, J. (2018). Collaborating for our future: Multistakeholder partnerships for solving complex problems. Oxford University Press. Haavisto, V. (2002). Court work in transition: An activity-theoretical study of changing work practices in a Finnish district court. University of Helsinki, Department of Education. Hopwood, N., & Gottschalk, B. (2017). Double stimulation “in the wild”: Services for families with children at-risk. Learning, Culture and Social Interaction, 13, 23–37. https://doi.org/10.1016/j.lcsi.2017.01.003 Jalasi, E. M. (2020). An integrated analytical framework for analysing expansive learning in improved cook stove practice. Learning, Culture and Social Interaction, 26, 100414. https://doi.org/10.1016/j.lcsi.2020.100414 Kelly, A. E. (2004). Design research in education: Yes, but is it methodological? The Journal of the Learning Sciences, 13 (1), 115–128. https://doi.org/10.1207/s15327809jls1301_6 Kerosuo, H. (2003). Boundaries in health care discussions: An activity theoretical approach to the analysis of boundaries. In N. Paulsen & T. Hernes (Eds.), Managing boundaries in organizations: Multiple perspectives (pp. 169–187). Palgrave. Leont’ev, A. N. (1978). Activity, consciousness, and personality. Prentice-Hall. Leont’ev, A. N. (1981). Problems of the development of the mind. Progress. Lotz-Sisitka, H., Mukute, M., Chikunda, C., Baloi, A., & Pesanayi, T. (2017). Transgressing the norm: Transformative agency in community-based learning for sustainability in southern African contexts. International Review of Education, 63(6), 897–914. https://doi.org/10.1007/s11159-017-9689-3 McDermott, R. P. (1980). Profile: Ray L. Birdwhistell. The Kinesis Report, 2(3), 14–16. Norman, D. A. (1982). Learning and memory. W. H. Freeman. Pleace, N., Knutagård, M., Granfelt, R., & Culhane, D. (2016). The strategic response to homelessness in Finland: Exploring innovation and coordination within a national plan to reduce and prevent homelessness. In N. Nichols & MIND, CULTURE, AND ACTIVITY 19 C. Doberstein (Eds.), Exploring effective systems responses to homelessness (pp. 426–442). The Homelessness Hub Press. Pressman, J. L., & Wildavsky, A. (1984). Implementation: How great expectations in Washington are dashed in Oakland; Or, why it’s amazing that federal programs work at all, this being a saga of the Economic Development Administration as told by two sympathetic observers who seek to build morals on a foundation. University of California Press. Programme of Prime Minister Sanna Marin’s Government. (2019, December 10). Inclusive and competent Finland: A socially, economically and ecologically sustainable society. https://julkaisut.valtioneuvosto.fi/bitstream/handle/ 10024/161935/VN_2019_33.pdf?sequence=1&isAllowed=y) Sannino, A. (2008). From talk to action: Experiencing interlocution in developmental interventions. Mind, Culture, and Activity, 15(3), 234–257. https://doi.org/10.1080/10749030802186769 Sannino, A., Daniels, H., & Gutierrez, K. (2009). Activity theory between historical engagement and future-making practice. In A. Sannino, H. Daniels, & K. Gutierrez (Eds.), Learning and expanding with activity theory (pp. 1–15). Cambridge University Press. Sannino, A. (Ed.). (2015a). The emergence of transformative agency and double stimulation: Activity-based studies in the Vygotskian tradition. Learning, Culture and Social Interaction, (4), 1–3. http://dx.doi.org/10.1016/j. lcsi.2014.07.001 Sannino, A. (2015b). The principle of double stimulation: A path to volitional action. Learning, Culture, and Social Interaction, 6, 1–15. https://doi.org/10.1016/j.lcsi.2015.01.001 Sannino, A. (2016). Double stimulation in the waiting experiment with collectives. Integrative Psychological and Behavioral Science, 50(1), 142–173. https://doi.org/10.1007/s12124-015-9324-4 Sannino, A. (2017). Researching work and learning for enacted utopias: The struggle to overcome homelessness. Keynote address at the 10th International Conference on Researching Work and Learning, South Africa: Rhodes University. Sannino, A. (2020). Enacting the utopia of eradicating homelessness: Toward a new generation of activity-theoretical studies of learning. Studies in Continuing Education, 42(2), 163–179. https://doi.org/10.1080/0158037X.2020.1725459 Sannino, A. (in press). Transformative agency as warping: How collectives accomplish change amidst uncertainty. Pedagogy, Culture & Society. doi: 10.1080/14681366.2020.1805493 Sannino, A., & Engeström, Y. (2017). Co-generation of societally impactful knowledge in change laboratories. Management Learning, 48(1), 80–96. https://doi.org/10.1177/1350507616671285 Sannino, A., & Engeström, Y. (2018). Valuable innovations out of nonsense? Expansive organizational learning and transformative agency in the Mann Gulch disaster and in the Finnish homelessness strategy. Teoria E Prática Em Administracão, 8(2), 60–79. https://doi.org/10.21714/2238-104X2018v8i2S-40728 Sannino, A., Engeström, Y., & Lemos, M. (2016b). Formative interventions for expansive learning and transformative agency. Journal of the Learning Sciences, 25(4), 599–633. https://doi.org/10.1080/10508406.2016.1204547 Sannino, A., & Laitinen, A. (2015). Double stimulation in the waiting experiment: Testing a Vygotskian model of the emergence of volitional action. Learning, Culture, and Social Interaction, 4, 4–18. https://doi.org/10.1016/j.lcsi.2014. 07.002 Sannino, A. (2010). Teachers’ talk of experiencing: Conflict, resistance and agency. Teaching and Teacher Education, 26, 838–844. Sannino, A. (2011). Activity theory as an activist and interventionist theory. Theory & Psychology, 21(5), 571–597. https://doi.org/10.1177/0959354311417485 Scribner, S. (1997). Mind and social practice: Selected writings of Sylvia Scribner. Cambridge University Press. Spinuzzi, C. (2019). “Trying to predict the future”: Third-generation activity theory’s codesign orientation. Mind, Culture, and Activity, 27(1), 4-18. https://doi.org/10.1080/10749039.2019.1660790 Vilela, R. A. (Eds.). (2019). Collaborative development for the prevention of occupational accidents and diseases: Change laboratory in workers’ health. Springer. Vygotsky, L. S. (1997a). Self-control. In R. W. Rieber (Ed.), The collected works of L. S. Vygotsky: The history of the development of higher mental functions (Vol. 4, pp. 207–219). Plenum. Vygotsky, L. S. (1997b). The problem of the development of higher mental functions. In R. W. Rieber (Ed.), The collected works of L. S. Vygotsky: The history of the development of higher mental functions (Vol. 4, pp. 1–26). Plenum. Wiewiora, A., Smidt, M., & Chang, A. (2018). The ‘how’ of multilevel learning dynamics. European Management Review, 16(1), 93-115. https://doi.org/10.1111/emre.12179 Wright, E. O. (2013). Transforming capitalism through real utopias. Irish Journal of Sociology, 21(2), 6–40. https://doi. org/10.7227/IJS.21.2.2 Zinchenko, V. P. (1985). Vygotsky’s ideas about units for the analysis of mind. In J. V. Wertsch (Ed.), Culture, communication and cognition: Vygotskian perspectives (pp. 94–118). Cambridge University Press. Zuckerman, G. A. (2012). Developmental education: A genetic modeling experiment. Journal of Russian and East European Psychology, 49(6), 45–63. https://doi.org/10.2753/RPO1061-0405490603 20 Y. ENGESTRÖM AND A. SANNINO