Introduction

The focus of this study is on the construction of shared meaning and practices as part of an emerging community of game makers. As such, the research questions address cultural factors that act as barriers to participation in digital making, and therefore require a method of research that allows the detailed analysis of complex, emergent learning environments. To achieve this, I adopt cultural historical activity theory (CHAT) as a guiding framework, operationalised through a formative intervention approach and incorporating adaptations from design-based research (DBR). This chapter explores these theoretical frameworks and key concepts relevant to the development of the thesis and the underlying ethos of this study. It begins by introducing the foundational concepts of AT and then explores third-generation activity theory (3GAT), developed by Engeström and the Helsinki school. In order to enhance the 3GAT approach, several key concepts from the work of the Laboratory of Comparative Human Cognition (LCHC) in San Diego are discussed. While AT forms the basis of this study, elements from other theoretical approaches within the learning sciences, particularly DBR, are incorporated to address the utility and practical application of the research [@hoadley_creating_2002-1; @barnett_ecosystem_2019]. This section justifies the synthesis of these theoretical choices. To address the study’s focus on designing an intervention aimed at fostering participant agency within the game-making process, the following section of the chapter explores key conceptions of agency within CHAT, specifically in the form of activist and transformative approaches. To demonstrate how CHAT is applied in research, the chapter also examines several forms of formative interventions, providing brief descriptions of how these concepts are operationalised in applied models.

The foundational concepts of activity theory

Activity Theory (AT) emerged as part of a movement known as the social turn in psychology [@sugiman_social_2008], marking a shift from a purely psychological approach towards a more social understanding of learning and human behaviour. While this shift has influenced various disciplines, the focus here is on its impact on education and learning processes. The rationale behind moving away from the assumption that human mental activity should be studied in isolated, context-free settings [@stetsenko_activity_2005] concerns not only the validity of results [@brown_design_1992], but also increases the potential relevance of the research for both designers and participants [@barnett_ecosystem_2019]. This perspective transcends traditional transmission models of learning and embraces the value of group-based, participatory processes.

Before outlining the core concepts driving Activity Theory (AT), a brief historical overview helps to contextualise them. For Marx [-@marx1975marx], human endeavour and change are based on activity, with the subject of that activity being the individual or group driving it [ @blunden_interdisciplinary_2009]. Vygotsky [-cole_mind_1978] built upon Marxist ideas, emphasising the importance of human activity as a socially mediated process, and applied these core concepts to the broader understanding of learning [@blunden_interdisciplinary_2010]. While Vygotsky’s student Leontiev developed Vygotsky’s work into a prototypical activity theory (AT) [-@leontev_problem_1974], AT stagnated in the repressive context of the USSR [@blunden_interdisciplinary_2010] and saw limited uptake in Western academia until Cole translated Vygotsky’s work [@cole_mind_1978]. Cole’s efforts brought these foundational ideas into the realm of education and psychology during the broader social turn in the discipline.

The work of Brown [-@brown_design_1992], a prominent psychology researcher who advocated the importance of context within experimental design, helped legitimise this focus and facilitated its broad adoption across diverse fields. While the interdisciplinary nature of the social turn encouraged pluralism and knowledge-sharing across fields as diverse as education, anthropology, and computer science, the field of learning sciences has been criticised for lacking conceptual coherence in understanding learning and human activity on a deeper level [@hoadley_learning_2011]. The growing work in these varied fields, sometimes grouped under the term learning sciences, reflects a move away from randomised control trials and their attempt to isolate context, shifting from a view of learning as an internal psychological process to a more situated perspective drawing on cultural psychology [@hoadley_learning_2011]. Within cultural psychology, scholars such as Wertsch [-@wertsch_vygotsky_1985] and Cole [-@cole_cultural_1996] explored the critical role of cultural mediation in human development, particularly in relation to learning communities and contexts. Nardi and Kaptelinin [-@kaptelinin_acting_2009] outline AT as a series of foundational principles that underpin broader theories, specifically: object-orientedness, tool mediation, the psychological processes of internalisation and externalisation, hierarchical structure of activity, and the ongoing development of activity [@wertsch_concept_1981]. These principles are now briefly introduced.

A useful starting point to explore key concepts in AT is Vygotsky’s formulation of the relationship between subject, object, and tools. Activity Theory is object-oriented, meaning that human behaviour and learning are driven by the goals or objectives of the individual or collective subject [@engestrom_learning_1987-1]. Kaptelinin and colleagues [-@kaptelinin_activity_1995] note that while the materialist roots of AT, stemming from Marxist theory, have led some interpretations to focus on the object as the physical entity being worked on and transformed by activity, conceptual objects are also included in Leontiev’s [-@leontev_problem_1974] conception of Activity Theory . Thus, the term object is best understood within the context of a broader motivation or objective. To represent the subject-object relationship in a way that addresses Vygotsky’s resolution of the subject-object, mind-body, and Cartesian dualism problems [@still1991mutual], this relationship was developed by Leontiev into a triadic model that includes tools which mediate the activity [@engestrom_learning_1987-1]. See Figure 3.1 below.

{width=60%}

Fig. 3.1 Essential components of activity outlined by Vygotsky

Turning to mediation via tools, Vygotsky’s [-@cole_mind_1978] concept of tools is broad , encompassing language, physical tools, and representational concepts. Mediation, the process of using a tool to work on an object(ive), is particularly relevant to this study, as the learning environment involves diverse and dynamic forms of tool use. Physical, digital, and conceptual tools evolve as products of the experiences of others in society, containing evidence of cultural adaptation and evolution. This study positions pedagogies as forms of mediational strategy that can be employed by both facilitators and learners [@donato_sociocultural_1994; @gutierrez_re-mediating_2009].

For Vygotsky, mediation via conceptual tools involves the interconnected processes of internalisation and externalisation of activity [@kaptelinin_activity_1995-1]. Learners both absorb cultural knowledge through participation in shared practices and contribute back to collective understandings through their own actions. This reciprocal process is well described in learning contexts by Rogoff [-@rogoff_observing_1995] via a personal process she calls participatory appropriation. This framing broadens the perspective beyond traditional transmission-based learning models and supports a more situated understanding of participation. Rogoff [-@rogoff_observing_1995, p.154] further identifies this appropriation as consisting of three key planes, of personal, interpersonal, and cultural processes, which are revisited in later chapters as part of the study’s analysis of learning in community settings.

To describe activity systems with greater granularity, Leontiev [-@leontiev_activity_2009] employed the concepts of actions and operations within activities. Together these elements constitute a vertical hierarchy of activity structure. At the highest level, activity is driven by an overall motive, while actions represent smaller sub-goals necessary to achieve this broader objective (see Figure 3.2 below).

{width=80%}

Figure 3.2 - Hierarchical structure of activity [@leontiev_activity_2009].

As we move down this hierarchy, we encounter a series of unconscious operations that are performed fluently during the execution of actions. The concept of fluency is key to this study, making it essential to locate and describe the chains of operations and actions involved in the activity. Kuutti [-@kuutti_activity_1995] explains that operations begin as conscious actions involving orientation and execution, but with practice, the orientation phase fades and the action collapses into a fluid, automatic operation. This understanding of internalisation connects closely with how fluency is developed in digital learning environments.

Local activity systems can be viewed as nested within broader activity systems. For instance, the development and teaching of a series of lessons may occur within the larger activity of a school or university [@lewin_developing_2018-1; @barab_using_2002]. Therefore, the process of selecting which activity system to analyse in depth involves identifying an appropriate unit of analysis [@blunden_germ_2014]. In Barab et al.’s [-@barab_using_2002] research on changes to a university course, the use of varied scopes and different framings of activity are used, at times focusing in on the specifics of evolving tool use or participant interaction as an object of activity, to build up a detailed examination of the tensions and processes within a complex educational environment. Using a smaller activity system than what is typically represented in Engeström’s [-@engestrom_learning_1987-1] approach is justified not only for enabling more detailed analysis but also for increasing the potential replicability and practical utility of findings for other researchers and practitioners [@barab_using_2002]. Similarly, in this thesis, while game making is studied as a community project within a larger system, interpersonal activity is also examined through an activity system of a smaller scope. This aspect is explored in more detail in Chapter 4.

In terms of development, activity theory perceives activity as constantly evolving within its context. The theory is rooted in Marxist dialectical materialism, which posits that activity is a dynamic system best understood through its transformations [@ilyenkov1960dialectics]. A significant driver of transformation in activity is the emergence and resolution of contradictions and conflicts that arise from the accumulation of tensions within the activity itself [@ilyenkov1960dialectics; @blunden_activity_2023]. Engeström [@engestrom_learning_1987-1] contributed the concept of expansive learning to activity theory, which traces the changes and growth of the object being worked on. This transformation often occurs as tensions extend beyond a single activity system to involve multiple systems, acknowledging the influence of actors moving between them [@engestrom_expansive_2001]. The next section introduces third-generation activity theory (3GAT), which extends the framework to include expansive learning and the dynamics of interacting activity systems.

CHAT, 3GAT and expansive learning

As activity theory (AT) has been adopted and developed beyond its Soviet origins, researchers have introduced new concepts within varied schools of practice. The cultural-historical approach and cultural-historical activity theory (CHAT) refer to the broader application of these principles, which may also integrate related concepts from sociocultural frameworks. The term third-generation activity theory (3GAT) has emerged from the work of Engeström and the Helsinki school’s interpretation of CHAT. This section introduces key concepts related to third-generation activity theory (3GAT) and its application through formative interventions, specifically: a community layer of activity, contradictions and tensions between system elements, double stimulation, and rising to the concrete.

Third-generation activity theory (3GAT) represents an important development of conceptual tools for applying activity theory to authentic learning settings. Engeström’s concept of the activity system model extends Leontiev’s work through a triangular representation that highlights key cultural aspects of the activity in question. Engeström [-@engestrom_learning_1987-1] added a new tier to the model to visually represent Leontiev’s focus on cultural factors. Specifically, these cultural factors include the wider community, which encompasses those involved in the activity beyond the subjects themselves, the division of labour, which may occur through either the distribution of tasks among community members or a more vertical power structure, and the emergence of rules or norms that guide evolving community behaviours [@sannino_cultural-historical_2018].

{width=60%}

Figure 3.3 - Engeström’s representation of community concepts an activity system

While these visual representations are novel, the main advances of third generation activity theory (3GAT) lie in its expanded focus on the role of interacting activity systems, the movement of subjects between systems, and the potential for collaborative work on a shared object. The recognition of broader activity systems and the transfer of motivations and practices of actors into new systems encapsulates Engeström’s perspective on activity theory [@spinuzzi_trying_2020-1]. This shared, or expanded, object being worked on by more than one activity system of expansive learning forms an important base for formative in [@sannino_formative_2016] [@engestrom_methodological_2014]. To clarify this relationship, an example in context is beneficial. In research on a joint activity involving schools, Cakir and colleagues [-@cakir_contradictions_2022] depict the joint activity as an objective shared between the school, family, and researchers.

{width=90%}

Figure 3.4 - An object shared between different activity systems in context from [@cakir_contradictions_2022]

The conception of joint activity emphasises the importance of the roles of researchers and other stakeholders within these interventionist and participatory processes. Analysing activity through its transformation, particularly in relation to the contradictions and tensions among system elements, draws on Marxist conceptions of dialectics and forms the foundation for the concepts of expansive learning. In 3GAT, this view of transformation through collaborative work between activity systems on a new, expanded object is termed expansive learning [@engestrom_design_2011].

Building on the foundational concepts of dialectical materialism, 3GAT allows for a granular exploration of tensions and contradictions as key sources of change in developing activity systems [@engestrom_expansive_2001]. It achieves this through careful description and analysis of system elements, their interactions, and the contradictions that emerge. Some tensions arising between participant motives and different parts of an activity system may be resolved quickly, while others persist and develop into impasses. These unresolved contradictions can serve as motivational forces that prompt agentic responses. A later section explores Sannino’s perspective on such conflicts as a source of transformative agency. Following Kuutti [-@kuutti_activity_1995] and Sannino [-@sannino_cultural-historical_2018], this study adopts the term contradictions to encompass concepts of tension, barriers and misalignment. Following Engeström [-@engestrom_discursive_2011, p. 340], these are treated as “manifestations of contradictions” 1. Additionally, Kamanga [-@kamanga_contradictions_2021] draws attention to the value of identifying system congruencies or strengths within activity systems.

Double stimulation refers to a process involving the use of tools to transcend contradictions that emerge in joint activity. Engeström’s [-@valsiner_cultural-historical_2007, p.492] early work on formative interventions frames a primary stimulus within an intervention as “the problem itself”, while a potential series of secondary stimuli are introduced by researchers as part of the ongoing intervention. Engeström and Sannino [-@engestrom_methodological_2014] explain that the process of double stimulation is active and goes beyond simple mediation via existing tools. The stimuli, whether presented by researchers as suggested tools or discovered in the environment by participants, are actively imbued with meaning and transformed by the participants through their use. Engeström [-@engestrom_methodological_2014, p.121] notes that it is likely that such stimuli are “ejected or reshaped by the participants, and the participants typically pick up or invent devices of their own”.

To track these transformations, researchers often identify a germ cell or unit of analysis. These concepts can be traced through the works of Vygotsky and Leontiev back to Hegelian and Marxist ideas of holistic understandings of phenomena [@blunden_activity_2023]. Forms of activity are viewed as concrete instances and iterations of underlying germ cell concepts, which exist in an abstract form. For example, Marx’s analysis identifies complex and diverse patterns of economic activity within the capitalist system as stemming from an underlying, prototypical concept of commodity exchange in pre-industrial economies. The works of Blunden [-@blunden_interdisciplinary_2010; -@blunden_germ_2014; -@blunden_activity_2023] and Davydov [-@davydov_problems_2008] provide comprehensive justification for the utility of identifying a germ cell or unit of analysis in relation to collaborative and learning projects. Blunden [-@blunden_unit_2020] explains that the germ cell and the unit of analysis are closely related concepts: the former foregrounds the developmental aspect, while the latter serves an analytical function. In this thesis, both are relevant. My role as researcher and practitioner requires attention to how design elements develop and how they are examined in analysis. One practical challenge in applying this concept lies in the stability and duration required to observe germ cell development. These developmental aspects are explored in the chapters that follow.

Formative Interventions as a CHAT methodology

The preceding section has outlined key theoretical tools within third-generation activity theory. These concepts are often operationalised through what Engeström and colleagues term formative interventions, a research methodology that blends interventionist aims with systemic analysis. Formative intervention plays a central role in applying CHAT to real-world contexts, offering a structured means of provoking and analysing transformation within activity systems. Before turning to more contextually adaptive approaches from the LCHC tradition, it is helpful to briefly clarify how formative intervention is understood in different schools of CHAT. Cole and Engeström [-@valsiner_cultural-historical_2007] use the term formative intervention to describe their interventionist research based on the principles of CHAT and overcome practical challenges in aligning with perceived rigour of social science, given the variation in environmental factors in the research process.

Engeström and Sannino [-@sannino_formative_2016; -@engestrom_discursive_2011] in the Helsinki school highlight the importance of double stimulation and identifying a germ cell that can be applied more widely through a process of rising to the concrete. Examples of specific instances of formative interventions from the Helsinki school are examined in the final section of the chapter. Their approach tends to emphasise structured, multi-phase interventions such as the Change Laboratory.

By way of contrast, Cole [-@valsiner_cultural-historical_2007] bases his initial interpretation of formative interventions on the concept of mutual appropriation. Rather than introducing a predefined conceptual tool, LCHC interventions evolve in response to the needs and actions of participants and community partners. The following section locates this concept within a wider ecology of formative interventions and design experiments emerging from the Laboratory of Comparative Human Cognition (LCHC) in San Diego.

CHAT concepts emerging from the Laboratory of Comparative Human Cognition (LCHC)

This section outlines key CHAT concepts related to the development and analysis of culture through joint activity [@valsiner_cultural-historical_2007] that have emerged from the work of the Laboratory of Comparative Human Cognition (LCHC) in San Diego; in particular, the research on the after-school intervention Fifth Dimension (5thD) [@cole_mind_1997]. LCHC’s approach to formative interventions is iterative, responsive and flexible. It shares characteristics and some common lineage with design-based research, and has been influenced by culturally rich, descriptive approaches such as communities of learners [@rogoff_developing_1994] and communities of practice [@lave_situated_1991]. This section tackles key concepts relevant to this study, namely: mutual appropriation, third spaces, idiocultures, funds of knowledge, and design-based approaches.

Mutual appropriation

Cole [-@valsiner_cultural-historical_2007; -@cole_fifth_2006] highlights the need to align research interventions with the ongoing aims of educational or community partners to maximise their sustainability beyond the research period. The concept of mutual appropriation has been utilised by Cole to explain and explore the development of the practicalities and structure of the research programme. The setting of 5D interventions predominantly consisted of after-school clubs run by community partners, guided by university researchers and supported by undergraduate student volunteers. The main activities were playful tasks aimed at providing fun, along with additional objectives of building literacy. Cole [-@cole_fifth_2006] noted that for an activity to be sustainable, the goals of researchers, student volunteers, funders, and local partners needed, as much as possible, to align . To achieve this, Cole and colleagues describe a mutual process of adaptation in joint activities between institutions, which not only reforms activities to make them suitable for ongoing partnership but also leaves a historical trace of evolving activities that serve as a valuable basis for cultural analysis [@downing-wilson_design_2011].

Third spaces and funds of knowledge

The concept of third space is helpful in the context of this study, in particular as a space between home life and formal education, rooted in sociocultural understandings of learning. Cole’s approach to formative interventions involved creating educational programmes founded on motivations of sustainability, mutuality, and a utopian ideal of creating a new activity system. This new system would thus be insulated from some of the more debilitating aspects of mainstream schooling and broader societal forces [@blunden_formative_2023]. Drawing on Homi Bhabha’s [-@bhabha_location_2012] postcolonial concept of the third space as a site of hybridity and negotiation between cultural meanings, Gutiérrez [-@gutierrez_rethinking_1999] recontextualised the term within education to theorise learning environments that bridge students’ everyday knowledge with academic practices. This process of blending the knowledge and discourses of the first space (home life) with the expectations and norms of the second space (school or institution) into a new third space has been explored in various studies [@moje_working_2004; @gutierrez_developing_2008; @bryant_academic_2009]. The work of Moje [-@moje_working_2004] in particular helps to demonstrate how the concept of third space can be operationalised within educational settings. She examined how home-based discourses might be productively integrated with school-based literacies, while also highlighting the challenges and tensions involved in constructing truly hybrid learning environments.

Funds of knowledge (FoK), a term popularised by Luis Moll [-@moll_funds_1992], one of the research partners in the 5th Dimension (5D) project, positions home-based discourses and practices as valuable resources that educators can draw upon to support learning in school and other educational settings. This approach aims to foster more equitable and inclusive learning experiences, particularly for students whose home cultures are significantly different from the dominant school discourse. FoK refers to the knowledge and skills acquired through participation in everyday activities outside formal schooling, which teachers can use to enhance classroom learning and support learner motivation. Moll and colleagues [-@moll_funds_1992] found that the cultural practices, skills, and traditions of Latino families were largely absent from mainstream school curricula, contributing to deficit views of these communities.

The concept emerged from research conducted in Latino communities in the United States, where Moll [-@moll_funds_1992] defined FoK as the knowledge household members acquire through lived experience and social practice. In the original study, researchers conducted home visits to inform the design of after-school programmes. However, limitations of this approach included its adult-focused design and the fact that only a subset of families were visited [@rajala_utopian_2023]. Rajala and colleagues [-@rajala_utopian_2023] critique the application of FoK in third space pedagogies, arguing that superficial uses risk reinforcing cultural stereotypes and reproducing deficit narratives. They also highlight practical barriers, such as the time-intensive nature of gathering meaningful data on students’ home cultures. In response, they advocate for the concept of funds of identity, which supports the co-construction of identity artefacts with participants as a means of representing and engaging with home cultures in more dynamic and participant-led ways [@subero_mediation_2018].

Repertoires of practice within three planes of activity

This section explores how concepts such as linguistic and cultural-historical repertoires and idiocultures have been used to understand and support culturally responsive learning environments. Rogoff and Gutiérrez’s [-@gutierrez_cultural_2003] initial motivation in their exploration of repertoires was to challenge reductive and overgeneralised notions of learning styles, particularly when applied to non-dominant communities. The researchers highlight the value of identifying potential funds of knowledge, framed as “linguistic and cultural-historical repertoires”, that learners bring to educational settings [@gutierrez_cultural_2003, p.22]. Following Fine [-@fine_boys_1987], Cole [-@cole_fifth_2006, p.32] employs the term idioculture to describe the emergence of a novel, localised micro-culture within the 5th Dimension (5thD) programme. In line with Downing-Wilson [-@downing_wilson_sociogenesis_2011], these idiocultures do not emerge in isolation; rather, they reflect the intersecting traits of diverse cultural systems represented by programme stakeholders [@cole_idiocultural_2017]. Cole [-@cole_idiocultural_2017] emphasises the utility of idiocultures both as a means of examining cultural variability across sites and as a practical basis for adapting programme activities to participants’ lived experiences .

The work of Rogoff and Gutiérrez [-@rogoff_cultural_2003] on linguistic and cultural-historical repertoires also seeks to prevent misrepresentation of cultural practices as innate traits, a pattern they identify as especially damaging when linked to racial assumptions. They stress that researchers and educators should avoid designing interventions based on presumed characteristics of learners. Instead, they call for contextualised inquiry into existing repertoires of practice, which can then be extended through collaborative activity. This position builds on Rogoff’s [-@rogoff_childrens_1993] foundational framework for analysing participation across three interconnected planes: personal, social, and community.

A central aim of analysing behaviour on the social plane is Rogoff’s effort to move beyond the binary between teacher-led and learner-led approaches, as also noted by Mascolo [-@mascolo_beyond_2009]. In addition, identity is highlighted as a key factor, but one that requires complex, multi-layered inquiry grounded in situated historicity rather than reliance on cultural stereotypes. This is important for this thesis, as it allows for an approach that transcends binary distinctions between instruction and discovery, supporting a more nuanced view of participation. This theme is explored in more depth in Chapter 6, particularly through the varied use of game design patterns.

Design-based research (DBR) in relation to formative intervention

This section introduces the theoretical foundations of design-based research (DBR), exploring its characteristics, critiques, and points of overlap with formative interventions grounded in cultural-historical activity theory (CHAT). Brown’s [-@brown_design_1992] provocation on the importance of context anticipated the growth in conceptual diversity and practical applications that characterise what is now referred to as design-based research (DBR). DBR has diverse interpretations, which stem from a set of generally agreed principles [@penuel_emerging_2014; @mckenney_educational_2021]. Easterday and colleagues [-@easterday_design-based_2014] characterise the core elements of DBR as: research as a form of intervention, iteration, involvement of participants in the evolution of designs, and flexibility in research outcomes based on how events unfold. Cobb and colleagues [-@cobb_design_2003, p.13] summarise the characteristics of design-based approaches as “extended (iterative), interventionist (innovative and design-based), and theory-oriented enterprises whose ‘theories’ do real work in practical educational contexts”.

DBR is common in the domain of learning sciences, particularly in advancing practice and theory related to innovative learning tasks [@penuel_emerging_2014]. For more detailed definitions of DBR and its place within a socio-economic funding structure, see Sandoval [-@sandoval_conjecture_2014] and Hay [-@hay_constructivism_2001, p. 3]. Barab and Squire [-@barab_design-based_2004] describe the inherent messiness of design-based research, which poses a challenge for researchers attempting to present results in a coherent and usable form. There is an ongoing tension between staying closely aligned with the specific, rich context of the research, and stepping back to generalise findings without becoming lost in abstraction. This balance is crucial for ensuring that research remains both contextually grounded and broadly applicable. Hoadley and Campos [-@hoadley_design-based_2022] further explore DBR outcomes beyond design refinement. These include the development of design processes, domain theories, design principles or patterns, and ontological innovations. They also identify new hypotheses and transformative learning experienced by researchers themselves. These varied outcomes highlight the broad applicability of DBR and the importance of ensuring that design processes and theoretical insights are communicated clearly to enhance the utility of research findings.

A key aspect of DBR is its emphasis on the rapid iteration of resources based on continuous feedback rooted in practice. In an educational context, this process helps produce pedagogies and resources that respond to the needs of learners [@cobb_design_2003]. This iterative responsiveness is also visible in how participants navigate the design environment itself. A further lens through which to view design-based research is the concept of affordances, commonly used in HCI and educational technology design to describe the perceived possibilities for action provided by a tool or environment [@kaptelinin_affordances_2012]. While affordances have often been analysed at the level of individual tool-user interaction, recent scholarship has expanded this to consider collective and sociocultural dimensions of affordance. In both HCI and DBR, there is a shared interest in how participants respond to contradictions in activity by appropriating or modifying tools to suit emergent needs. As Karanasios and Allen [-@karanasios_moving_2021] suggest, users often engage with digital tools in creative ways to overcome obstacles or redesign workflows, an idea echoed in DBR’s emphasis on iterative refinement and participant agency. In this thesis, learners’ responses to design tensions, such as how they adapted or misused templates, reshaped their digital tools, or requested changes to the learning environment, can be understood through this sociocultural framing of affordance. Such moments serve as evidence of agency, aligning with the CHAT principle that contradictions can act as drivers of transformation when supported by responsive design. These overlapping features and tensions provide a useful basis for comparing DBR and formative intervention more directly.

Tensions and alignments between DBR and formative intervention

The concerns of the previous sections indicate shared issues between schools of CHAT and DBR researchers. In an editorial of a dedicated edition of Mind, Culture, and Activity, Cole et al. [-@cole_designing_2014] note existing interactions and introduce the proposal of Penuel [-@penuel_emerging_2014] to link FI and DBR explicitly by proposing and defining the term “formative intervention research” as a useful way to encapsulate the alignment of FI and DBR via their concern for historically emerging tensions and the importance of expanding the agency of the participants in interventions.

Within the same issue, Engeström and colleagues [-@engestrom_methodological_2014] resist this proposition, insisting that formative interventions should contain a clear articulation of double stimulation and the process of rising to the concrete within the framework of Engeström and Sannino’s [-@sannino_formative_2016; -@engestrom_methodological_2014; -@engestrom_discursive_2011] understanding of these concepts. Engeström’s article [-@engestrom_methodological_2014] further develops his previous critique of DBR [-@engestrom_learning_2009]: that it lacks rigorous examination of the dynamic and historical nature of context within CHAT conceptions of activity, and that it tends to be overly focused on a closed, linear approach to refining the learning design, which prioritises researchers’ goals and limits the input and agency of participants.

Other critiques focus on institutional constraints and research practices. O’Neill [-@oneill_understanding_2016] emphasises that DBR researchers often fail to invest sufficient time in familiarising themselves with the existing learning practices of the contexts they study—an oversight that runs counter to the more contextually focused approaches encouraged by CHAT. He also offers factors which help explain perceived deficits in the latter’s analysis of historical context. These include the timescales and institutional priorities that shape DBR in funded research environments. An additional factor at play is the context in which interventions are often applied. The formative interventions of the Helsinki school are frequently undertaken in stable, institutional work environments, particularly in European contexts, where stakeholder roles and systemic structures are more predictable. In contrast, learning environments associated with DBR, especially in less formal or rapidly evolving contexts, do not share the same stability [@spinuzzi_trying_2020-1]. Given the reliance of formative interventions of the Helsinki group on this relative stability, this difference may limit the utility of such approaches in more fluid settings.

In response to these tensions, several authors call for rethinking assumptions. Bakker [-@bakker_design_2018] challenges Engeström’s characterisation of DBR as having closed goals, particularly critiquing Engeström’s portrayal of the work of Collins and colleagues [-@collins_design_2004] as a misrepresentation. Bakker [-@bakker_design_2018, p.17] argues that these researchers did “not want to fix variables or educational goals upfront”. He echoes the call of Cole, Penuel, and O’Neill [-@cole_cultural-historical_2018] for more dialogue and greater recognition of the shared traditions between CHAT and design research.In addition, Penuel [-@penuel_emerging_2014] emphasises the eclectic methodological approach in design-based research as a necessary means of responding flexibly to context. Bakker also views this flexibility as a strength of DBR, resisting calls for greater formalisation of the design process from other researchers [@ormel_researchpractice_2012; @mckenney_systematic_2013].

More recent work signals increasing convergence. Recent work by Engeström and colleagues [-@cakir_contradictions_2022] suggests a shift toward a more collaborative stance on DBR, citing a new generation of DBR which involves multiple stakeholders as co-designers in dynamic educational contexts. In particular, they propose that formative interventions can enrich DBR by offering a robust framework incorporating the concepts explored in previous section [@cakir_contradictions_2022]. This shift highlights a growing consensus on the potential benefits of incorporating cultural-historical concepts into DBR approaches.

This study draws on overlapping principles from design-based research and formative intervention, combining their strengths to support collaborative learning in a dynamic, informal setting. While the discussion so far has explored how these traditions align and diverge, the next section turns to concepts of agency that underpin the study’s analytical focus. These ideas offer a bridge between theory and practice, helping to explain how learners navigated the intervention and how the research design evolved in response. The chapter then closes by examining how selected approaches have operationalised these concepts, informing the flexible methodology developed in Chapter 4.

Concepts of agency within CHAT

To help later exploration of RQ3, which interogates how participants develop agency in an evolving community of game makers, it is important to clarify the concept of agency, particularly in the context of learning environments. Agency and autonomy are often either vaguely defined or have contested, interconnected definitions in studies [@huang_autonomy_2013]. The question of who or what has agency is a philosophical one with practical applications in designing learning settings [@matusov_mapping_2016]. For this thesis, in line with a cultural-historical approach, agency is viewed as a cultural phenomenon experienced by individuals in relation to their peers and their learning environment [@hopwood_forward_2022]. In this view, agency is not a fixed trait but something that emerges dynamically within activity systems. To situate this view within the varied and evolving concepts of agency in educational settings, I draw on notions of instrumental, transformational and relational agency within the framework of activity theory. These forms of agency offer a broad perspective on how individuals navigate and reshape their learning environments. While this section outlines the theoretical framing of agency, a more detailed exploration of how agency developed in practice is provided in Chapter 7.

Instrumental agency refers to the capacity to achieve goals through purposeful action, often framed as the ability to get things done [@edwards_steps_2005]. This concept presents agency as a set of learnable or improvable capacities. Matusov and colleagues [-@matusov_mapping_2016] highlight how educational environments can take a technological or instrumentalist turn, where educators focus on increasing learner capacity to complete specific tasks. Many intervention-based studies, particularly in educational technology, seek to develop this form of agency by introducing new tools and tracking learners’ progress in using them. These studies often aim to identify barriers to task completion and refine design iterations accordingly.

Within a CHAT perspective, instrumental agency emerges through successful tool mediation. As learners gain fluency with tools, actions that once required effort may become automatised, aligning with Leontiev’s [-@leontev_activity_1978] account of the transition from conscious actions to unconscious operations. This shift reflects the stabilisation of practice through repeated mediation. Understanding instrumental agency through an activity-theoretical lens, therefore, requires attention to how tools shape action over time and how mediated activity becomes internalised as operational routine.

The Helsinki school’s focus on transformative agency highlights the importance of intervention and transformation as desirable aspects of research [@engestrom2006development; @haapasaari_emergence_2016; @sannino_formative_2016]. In recent years, this has been framed through the concept of transformative agency through double stimulation (TADS). Similarly to TAS, the objective of TADS is to increase the agency of participants in research interventions. In addition, TADS researchers aim to develop a broader understanding of agency from a CHAT perspective. As outlined above, double stimulation describes a secondary stimulus introduced beyond the first stimulus, understood as the motivation of the activity. In the Change Lab (CL) interventions, for example, the active use of a secondary stimulus can both analyse and facilitate agency by provoking reflection [@sannino_principle_2015].

Studies of transformative agency, for example within CL, often involve a significant longitudinal aspect. However, the TADS framework has been applied in research over much shorter time frames and in more naturalistic settings such as hospital practice [@hopwood_forward_2022; @hopwood_volitional_2022], higher education [@grant_double_2022], and digital education [@aagaard_teacher_2022]. The concept of volitional action by subjects as a foundation for transformation is used in these shorter term studies to explore the processes of decision forming and implementation [@virkkunen_dilemmas_2006; @sannino_double_2015]. Here, the secondary stimuli are used by participants as a spur and then as leverage in volitional acts. To clarify how this process works, Sannino [-@sannino_double_2015] describes how participants use secondary stimuli to deliberately shift their course of action when they encounter contradictions or difficulties. This highlights the active effort of participants to overcome tensions and dilemmas in their activities through intentional action using conceptual or material tools introduced during the intervention. While not all attempts, such as directly asking a peer or teacher for help, will succeed, some will take hold and lead to transformation for the student. This may also trigger a broader shift in the educational environment itself. This idea is particularly relevant to the setting of this study, where the learning environment is co-created by both participants and researchers.

Stetsenko’s [-@stetsenko_critical_2020] transformative activist stance (TAS) presents a politicised view of agency that challenges both individualist and overly structural conceptions. She critiques views of agency that reduce it to either personal traits or render individuals as powerless within wider cultural systems. Instead, TAS positions agency as a dynamic interplay between social tensions and human volition, grounded in an urgent call to transform unjust conditions. From this perspective, Stetsenko [-@stetsenko_hope_2020] advocates a deliberately partisan approach to research, opposing the detachment and neutrality often associated with postmodern or post-humanist positions. This framing is relevant to the ethical commitments of this study, particularly in its alignment with activist research that aims to reshape the learning conditions of participants, not simply observe them.

Relational agency has multiple interpretations. Stetsenko [-@stetsenko2020radical] critiques ecological and sociomaterial theories of agency, including those of Latour [-@latour_reassembling_2005], Barad [-@barad_meeting_2007], and Giddens [-@giddens_constitution_1984]. While these approaches usefully explore distributed activity across human and material systems, she argues that they often neglect power, political struggle, and the transformative potential of human action. In contrast, Edwards and Mackenzie’s [-@edwards_steps_2005] conception of relational agency builds on activity theory and retains a focus on the historical and political dimensions of practice. Relational agency extends the idea of instrumental agency by recognising the social capacity to interpret and respond to others’ actions, particularly through collaboration. It involves individuals working with others to expand a shared object of activity, where goals may shift and evolve. In educational contexts, this form of agency is often constrained by institutional routines, but remains a valuable lens for analysing participation in community-based learning environments.

For Edwards [-@edwards_building_2011, p.34], relational agency involves developing a shared understanding of an activity’s object and adjusting one’s actions in response to others’ evolving interpretations of that object. Feedback is often a key part of this process. As explored in Chapter 2, Gutiérrez and colleagues’ work on 5thD interventions [-@digiacomo_relational_2016-1; -@gutierrez_developing_2008; -@digiacomo_seven_2017] offers a compelling example of relational agency in practice, particularly in after-school learning contexts concerned with identity and equity. Their analysis highlights how relational and material forms of feedback contributed to what they term relational expertise, a shared capacity to engage purposefully in joint activity. In the context of this intervention, the researchers argue that “relational and material feedback elements of the activity helped to create an environment with relational agency” [@digiacomo_relational_2016-1, p.148]. This example illustrates how specific educational design features, such as structured opportunities for reflection and feedback, can support forms of agency that are both social and developmental.

Framing agency in this way positions it as a socially situated and evolving capacity, shaped by participation in collective activity and the tensions that emerge within it. It highlights the importance of learners not only acting within a designed environment, but also adapting and reshaping it through joint activity. While this chapter establishes the conceptual groundwork for understanding agency in the study, Chapter 7 returns to this theme in greater depth, examining how different expressions of agency emerged and developed across the intervention.

Variations and synthesis in interventionist CHAT and DBR research

Despite shared interests in transformation and participation, activity theory and design-based research have produced divergent methodological models. These approaches may be driven by different underlying principles or practical considerations. For example, in this study, rather than analysing larger workplace systems over extended periods, the focus is on the analysis of systemic tensions and resulting transformations within the context of shorter interventions, as seen in diverse research [@barab_developing_2002; @barab_using_2002; @lewin_developing_2018]. This section outlines key approaches relevant to this study, including the Change Laboratory, Learning Labs, design narratives, and social design-based experiments (SDBEs), each of which offers a distinct take on interventionist research. The aim is not to provide an exhaustive review, but to highlight methodological features that inform the design of this study. Elements from these traditions are integrated into the approach developed in the following chapter.

The Change Laboratory model

The Change Laboratory (CL) is the key implementation of expansive learning in workplace environments proposed by the Helsinki school [@engestrom_putting_2007]. CL interventions typically involve a series of workshops with key stakeholders, facilitated by researchers. The structural model of third-generation activity theory (3GAT) is used to identify and reflect on contradictions in current practices, with the goal of collectively envisioning and enacting change. This model is well suited to relatively stable work environments where roles, tools, and community structures are clearly defined.

In the CL model, the mechanism of double stimulation plays a central role. The first stimulus is often empirical data, such as video or observational artefacts, which help participants surface contradictions. The second stimulus takes the form of conceptual tools or mediational artefacts introduced to support reflection and reorganisation of the activity. Double binds are particularly important in this process. They mark moments where the constraints of the existing system become intolerable, creating a space for participants to take up new tools or concepts and reframe the situation. Sannino [-@sannino_principle_2015] highlights how these auxiliary stimuli can support volitional engagement, enabling the emergence of transformative agency.

Learning Labs

A related approach is the Learning Lab, which adapts similar workshop structures to school and educational contexts [@cakir_contradictions_2022; @bal2011culturally]. While it shares core features with the CL model, the Learning Lab is often used to study how educational innovations emerge and unfold, rather than to directly produce them. As such, it tends to operate at a more analytical level, focusing on how learning environments evolve through participation and reflection. This meta-level positioning means that Learning Lab studies are particularly attentive to how tensions and shifts play out, often through linguistic analysis of interactions between diverse stakeholders in tailored reflective workshops.

One key contribution of this tradition is its detailed treatment of tension and contradiction as productive elements of change. In particular, Engeström and Sannino [-@engestrom_discursive_2011] offer a typology of four forms of discursive contradiction observed in participant dialogue: dilemma, conflict, critical conflict, and double bind. These categories describe different ways that participants express and work through tensions in activity settings. For example, dilemmas involve expressions of uncertainty or mixed evaluations; conflicts feature direct disagreement; critical conflicts often include emotional or moral struggle; and double binds occur when all available options seem inadequate, prompting the need for a new direction.

This typology aligns closely with the earlier discussion of double stimulation. While the model was introduced through CLs, the same explanatory framework can be applied to Learning Labs. The presence of discursive tensions, particularly double binds, helps explain how expansive learning unfolds in real time, especially when participants are given tools to articulate, confront, and work through contradictions. This framing is foundational to the present study’s approach to both design and analysis, and shapes how agency, innovation, and tool use are interpreted in the chapters that follow.

Design narratives

Design narratives emerged in the learning sciences as a response to the limitations of communicating design-based research (DBR) within traditional journal formats, where space and context are often constrained [@hoadley_creating_2002]. They offer a way to bridge the tension between richly situated accounts and the need for generalisable insights. Hoadley [-@hoadley_creating_2002-1] introduced design narratives as a means of providing detailed, reflective accounts of the learning design process. These narratives document the problems encountered, the iterative changes made, and the reasoning behind key decisions [@hoadley_creating_2002-1; @brase_knowledge_2024; @bell_theoretical_2004]. Bakker [-@bakker_design_2018] similarly notes the challenge of representing the full complexity of design work within conventional publishing formats. In this study, a design narrative is particularly valuable in conveying the intricate details of the learning design. This thesis format allows for a fuller exploration of these aspects, notably in Chapter 5, where I use extended description to examine how tensions emerged and were responded to during the activity. I argue that design narratives can help further align DBR with CHAT-informed interventions. By incorporating systemic concepts from activity theory, they offer a more holistic account of both the practical and analytical development of an intervention.

A key influence on this approach is the Fifth Dimension project, described in Cole’s work [-@cole_fifth_2006]. The 5thD programme exemplifies a CHAT-informed design narrative situated in informal, community-based settings [@nicolopoulou_design_2009; @downing-wilson_design_2011]. It shows how both educational and research designs were shaped by feedback from participants, student volunteers, and educators. Design adaptations emerged in response to concrete tensions and local cultural practices, rather than being fixed in advance. As discussed in Chapter 2, 5D also offers a model of thick description and negotiated meaning across institutional and everyday contexts. Researchers did not impose a static design but remained responsive to the lived activity of the site. This underscores the idea that design narratives are not only retrospective records but also cultural tools that shape and are shaped by the learning activity itself.

Designing for equity and transformation: from DBIR to SDBEs

Efforts to scale context-sensitive educational innovation have led to methodological developments that extend core principles of design-based research (DBR). One such development is design-based implementation research (DBIR), which emerged from Penuel and colleagues’ work [-@penuel_design-based_2021] and seeks to address the challenge of producing change that is both sustainable and equitable across institutional systems. DBIR retains DBR’s emphasis on iteration and collaboration but explicitly prioritises long-term implementation, practitioner partnership, and system-level coordination. It offers a pragmatic orientation to research, foregrounding organisational learning and equity from the outset, especially in contexts like inquiry-based learning where resources may be constrained.

While DBIR represents a more formalised response to scaling equity-oriented design, related approaches have taken shape in less institutionalised contexts. For example, the methodological and political aims of DBIR share a strong resemblance with social design-based experiments (SDBEs), a related approach developed by Gutiérrez and colleagues [-@gutierrez_social_2016-1]. SDBEs build on traditions of formative intervention and DBR but reorient the work around questions of identity, power, and social justice. While DBIR often operates in more formal institutional settings, SDBEs are typically situated in informal or hybrid learning environments where learners’ repertoires and cultural histories are central to the design process.

This study draws on these overlapping traditions, but given its focus on a fluid, co-designed learning space involving multiple stakeholder voices, the approach aligns most closely with the logic of SDBEs. In particular, SDBEs support the creation of third spaces where community and institutional knowledge intersect, and where feedback functions not just as an instructional device but as a material and relational element in cultivating relational expertise and relational agency [@digiacomo_relational_2016-1]. These concerns echo the aims of DBIR but are more explicitly addressed in the analytical and practical toolkit of SDBEs, making them especially relevant to the intervention context explored in this thesis.

Conclusion: from theory to intervention

This chapter has outlined the theoretical foundation for the study, combining concepts from design-based research (DBR) and cultural-historical activity theory (CHAT) to support an interventionist methodology. I have traced how their overlapping concerns with participation, iteration, and transformation provide a productive framework for the study’s design. At the same time, attention has been paid to their distinct theoretical commitments. DBR foregrounds flexible design processes in learning environments, while CHAT centres the systemic analysis of contradictions and the development of agency.

A central focus has been the treatment of agency, particularly through concepts of relational and transformative agency. These offer tools for analysing how participants engage with and reshape the learning environment. The role of contradictions, double stimulation, and expansive learning has been emphasised as mechanisms through which new forms of activity and identity emerge. The chapter also examined specific interventionist models, including the Change Laboratory, Learning Lab, and social design-based experiments (SDBEs), to identify methodological approaches relevant to the present study.

The thesis draws on elements from across these approaches, particularly the emphasis on iterative design, analysis of contradictions, and joint activity. The intervention explored in this study did not adopt a single model wholesale, but instead developed a hybrid approach tailored to the demands of a fluid, informal setting. In doing so, it reflects the tradition of CHAT-informed research that treats interventions not only as means of change but also as analytical lenses through which learning and development can be understood.

The next chapter outlines the methodological design of the study, including how the theoretical concepts introduced here were operationalised through collaborative practices, data collection, and analysis.

  1. For a comprehensive analysis of systemic contradictions, see the work of Engeström and Cakir [@cakir_contradictions_2022; @engestrom_discursive_2011]. ↩