Growth mindset as an emergent outcome: autonomy and strategy use in second language learning

1 Introduction

Growth mindset interventions have attracted sustained interest in language education, yet their effects have proven disappointingly modest and inconsistent. Meta-analytic reviews consistently find small effect sizes and high variability across contexts (Macnamara and Burgoyne, 2023; Sisk et al., 2018), and Yeager and Dweck (2020) themselves acknowledge that mindset beliefs produce strongest effects only when learners operate in environments where effort is genuinely consequential. This pattern suggests a fundamental theoretical problem: if growth-oriented beliefs are treated as stable internal predictors that independently drive motivation and achievement, the repeated failure of belief-targeted interventions to produce durable effects is difficult to explain. The present article proposes an explanation, and with it a reconceptualization of the construct itself.

The central argument is that growth mindset is better understood as an emergent outcome of accumulated learning experience than as a primary causal predictor of achievement. Learners do not develop adaptive beliefs through instructional messaging about the malleability of intelligence; they develop them through environments in which deliberate, strategic action is visibly and repeatedly connected to improvement. This reconceptualization is not merely terminological. It has direct consequences for how autonomy, strategies, and mindset relate to one another theoretically, and for what kinds of pedagogical designs are likely to produce durable belief change. Specifically, this article advances a claim that further develops an underexplored proposition in the SLA literature: that language learning strategies function not only as behavioral tools for improving performance, but as belief-formation events through which growth-oriented interpretations of ability are constructed and internalized. This represents the central theoretical contribution advanced in the present framework.

This reorientation connects to broader developments in second language acquisition (SLA) research, which has increasingly moved toward perspectives emphasizing learner agency, self-regulation, and adaptive participation (Dörnyei and Ryan, 2015; Mercer and Dörnyei, 2020; Ushioda, 2020). Paradigmatically, the Douglas Fir Group (2016) proposed a transdisciplinary framework for SLA that foregrounds the dynamic, layered interplay of individual, social, and macro-contextual processes in language learning development, and several subsequent accounts have explored intersections between autonomy and motivational constructs within specific educational contexts (Edsall, 2023). The present article extends this line of inquiry by identifying the precise mechanism through which autonomy-supportive environments translate into belief change. Within this shift, learner autonomy has emerged as a foundational condition enabling productive engagement with challenge (Benson, 2011; Lamb, 2017), and self-regulated learning research has demonstrated that beliefs, strategies, and contextual affordances operate recursively rather than independently (Pintrich, 2000; Zimmerman, 2002). What the literature has not yet provided is a coherent account of the mechanism through which autonomy-supportive environments translate into belief change. The present framework proposes that language learning strategies may represent a central mediating mechanism through which autonomy-supportive environments become translated into growth-oriented beliefs (Burns and Richards, 2018; Dörnyei, 2005; Ushioda, 2011).

Learner autonomy has long been recognized as a foundational construct in language education, traditionally defined as learners’ capacity to take responsibility for their own learning (Holec, 1981). Subsequent scholarship has expanded this view by conceptualizing autonomy as multidimensional, socially mediated, and context-sensitive, involving behavioral, cognitive, and relational dimensions (Benson, 2011; Lamb, 2017). Research consistently links autonomy with greater persistence, strategic engagement, self-regulated learning, and achievement (Little, 1991; Teng and Zhang, 2020). Increasingly, autonomy is understood not simply as an individual trait, but as something that instructional environments either enable or constrain.

In parallel, growth mindset has become highly visible in educational psychology and, more recently, in SLA research. Growth mindset refers to the belief that abilities can be developed through effort, effective strategies, and support (Dweck, 2006). Such beliefs have been associated with resilience, persistence, and adaptive responses to challenge (Lou and Noels, 2024; Yeager and Dweck, 2020). However, recent meta-analytic and critical work has raised important questions regarding the size, consistency, and transferability of mindset effects across contexts (Macnamara and Burgoyne, 2023; Sisk et al., 2018). These findings suggest that mindset may be less a stand-alone causal variable than a construct whose influence depends on broader learning processes.

Despite their conceptual proximity, learner autonomy and growth mindset have rarely been examined within a unified explanatory framework. Autonomy research has primarily focused on agency, self-management, and pedagogical conditions, whereas mindset research has centered on beliefs about ability and motivation. As a result, insufficient attention has been paid to how learners’ experiences of autonomy may shape belief formation, or how beliefs may emerge through sustained participation in autonomy-supportive learning environments.

This separation reflects a broader limitation in educational research: the tendency to treat psychological constructs as isolated predictors rather than as interdependent components of developmental learning processes. Research on self-regulated learning demonstrates that beliefs, strategies, agency, and contextual affordances operate recursively rather than independently, with each shaping the conditions under which the others function (Pintrich, 2000; Zimmerman, 2002). Within this account, language learning strategies may function as a critical mediating mechanism, translating pedagogical opportunities into behavioral engagement and adaptive interpretations of learning experience.

Against this background, the present article proposes an integrated conceptual framework linking learner autonomy, language learning strategies, and growth mindset. Specifically, it argues that growth-oriented beliefs are unlikely to arise directly from instructional messaging or isolated mindset interventions. Rather, they emerge gradually when learners repeatedly experience that effort, strategic action, and self-directed engagement can lead to improvement.

Accordingly, this article advances three interrelated claims. First, growth mindset should be reconceptualized as an emergent and context-sensitive outcome rather than a primary causal driver of achievement. Second, learner autonomy should be understood not merely as an endpoint of instruction, but as a generative condition that enables productive engagement with learning processes. Third, language learning strategies serve as mediating mechanisms through which autonomy-supportive environments become experientially meaningful to learners.

By synthesizing insights from SLA, self-regulated learning, and motivational psychology, this article contributes to ongoing calls for more process-oriented and theoretically integrated models of language learning. In doing so, it offers a framework that repositions the relationship between pedagogy, learner behavior, and belief systems within a dynamic and recursive account of development.

Methodological note. This article is a theoretically grounded conceptual synthesis rather than a report of original empirical research. It draws selectively on published peer-reviewed scholarships in SLA, self-regulated learning, and motivational psychology. Sources were identified through searches of Scopus, Web of Science, and Google Scholar using the terms learner autonomy, growth mindset, language learning strategies, self-regulated learning, and second language acquisition, and were selected based on relevance to the central argument, methodological rigor, and recency. Preference was given to meta-analyses, longitudinal studies, and theoretically influential works. This article does not claim to report a systematic review and makes no attempt to exhaustively catalogue the available literature; its purpose is to identify convergent patterns that support the proposed framework and to derive pedagogically grounded implications.

2 Literature review

2.1 Learner autonomy as a situated capacity

Learner autonomy remains one of the most influential constructs in language education. Initially conceptualized as learners’ ability to take charge of their own learning (Holec, 1981), subsequent scholarship has significantly broadened this definition. Contemporary perspectives view autonomy not as simple independence, but as a situated capacity shaped by cognitive resources, social relations, and pedagogical opportunities (Benson, 2011; Lamb, 2017). Importantly, Little et al. (2017) demonstrated through longitudinal classroom research that autonomy develops through sustained democratic dialogue and shared decision-making between teachers and learners, a finding that underscores the inherently relational and pedagogically contingent nature of the construct. More recently, Reinders and Benson (2017) argued that autonomy research must attend more carefully to the social and contextual conditions under which self-directed learning becomes possible, extending the construct well beyond Holec’s (1981) original focus on adult learners in self-access settings (Benson, 2013; Benson and Voller, 1997).

Empirical research consistently associates autonomy with positive learning outcomes, including greater persistence, engagement, strategic behavior, and academic achievement. Particularly relevant is its close relationship with self-regulated learning processes such as goal setting, monitoring, and reflective adjustment (Teng and Zhang, 2020). These findings suggest that autonomy is best understood as an enacted capacity rather than a stable learner trait: something that learners do, rather than something they possess.

At the same time, autonomy is unevenly distributed across contexts. Institutional constraints, exam-oriented curricula, and teacher-centered norms may restrict opportunities for learner agency (Littlewood, 1999). This indicates that autonomy is not simply possessed by learners but produced through interaction between learners and learning environments. Critically for the present argument, when environments restrict agency, learners may also be deprived of the experiential conditions through which growth-oriented beliefs would ordinarily develop.

2.2 Language learning strategies and self-regulated engagement

Language learning strategies have long been linked to successful language development (O'Malley and Chamot, 1990; Oxford, 1990). More recent research, however, has moved beyond cataloguing strategies toward understanding how strategic behavior supports self-regulated learning (Gao, 2007; Oxford, 1990).

Studies show that explicit strategy instruction can strengthen learners’ ability to plan, monitor, evaluate, and adapt their learning processes, leading to improved performance across skills domains (Teng, 2022; Teng and Zhang, 2020). Strategy use has also been associated with increased self-efficacy, sustained engagement, and metacognitive awareness (Flavell, 1979; Oxford, 2017; Wenden, 1998).

Importantly, strategies may serve functions beyond performance enhancement. When learners experience that deliberate strategy use improves outcomes, they may reinterpret success as controllable rather than fixed. In this sense, strategy use may influence not only behavior, but also learners’ beliefs about learning and ability. It is this dual function, namely behavioral regulation and belief formation, that positions strategies as theoretically central to the present framework.

2.3 Growth mindset in SLA: promise and limits

Growth mindset has attracted increasing attention in SLA because language learning often involves uncertainty, repeated errors, and delayed progress. Learners who hold growth-oriented beliefs may be more willing to persist, experiment, and tolerate difficulty (Cheong et al., 2024; Fan et al. 2024; Lam and Zhou, 2022; Lou and Noels, 2024; Lou et al., 2022; Mercer and Ryan, 2010; Sadoughi and Hejazi, 2024).

Nevertheless, the broader educational literature has become more cautious regarding mindset effects. Meta-analyses indicate that the relationship between growth mindset and achievement is generally modest, while intervention outcomes vary substantially across settings (Macnamara and Burgoyne, 2023; Sisk et al., 2018). Importantly, Yeager and Dweck (2020) themselves have acknowledged that mindset effects are strongest when learners also encounter challenges within contexts that make effort consequential, a qualification that aligns closely with the argument advanced here (Bai and Wang, 2023; Jiang et al., 2024; Rong et al., 2024).

A growing implication of this work is that mindset may be contextually produced and behaviorally reinforced rather than simply internally held. Learners’ beliefs about ability are likely shaped by repeated experiences of failure, progress, feedback, and perceived control. Crucially, Zarrinabadi et al. (2021) demonstrated directly that autonomy support predicts the development of growth-oriented language mindsets, providing an empirical basis for the theoretical pathway proposed in the present article (Wang et al., 2021).

Recent evidence further suggests that growth mindset develops within broader self-regulatory and interpersonal systems rather than functioning as an isolated learner characteristic. For example, Dong (2024) found that learners’ meta-lay theories regarding their teachers’ beliefs predicted growth mindset, which in turn mediated the use of online self-regulated learning strategies. These findings reinforce the view that growth-oriented beliefs are shaped by learners’ interpretations of their learning environments and social relationships. However, whereas Dong (2024) conceptualizes growth mindset primarily as a mediator between contextual perceptions and self-regulation, the present framework conceptualizes growth mindset as an emergent outcome of sustained experiences of learner autonomy, strategic engagement, and perceived progress across educational contexts.

2.4 The integrative gap

Although learner autonomy, strategic learning, and growth mindset are frequently discussed in adjacent literatures, they are rarely theorized together. This constitutes a notable omission because each construct addresses a different dimension of the same developmental process: autonomy concerns learner agency and the conditions that enable it; strategies concern the regulation of action through which agency becomes effective; and mindset concerns the interpretive beliefs through which learners make sense of their own progress and capacity.

Research on self-regulated learning suggests that these dimensions are mutually influential rather than independent (Pintrich, 2000; Zimmerman, 2002). Learners who experience meaningful control over learning may engage more strategically; strategic success may strengthen perceived competence; and repeated experiences of improvement may foster growth-oriented beliefs. What literature has not yet provided is a coherent account of how these processes connect over time and through what mechanisms autonomy-supportive environments translate into belief change. Addressing this gap is the central purpose of the framework proposed in the following section (Zimmerman and Schunk, 2011).

3 Proposed framework and theoretical contributions

The central argument of this article is that the relationship between learner autonomy and growth mindset has been conceptually underdeveloped because the two constructs have typically been studied in isolation. While autonomy has been framed as a matter of learner control and self-direction, mindset has largely been treated as a belief-based predictor of motivation and achievement. This separation obscures the developmental processes through which beliefs about ability may emerge from learners’ experiences of agency, regulation, and progress.

To address this gap, the present article proposes a process-oriented framework in which language learning strategies function as the key mediating mechanism linking autonomy-supportive environments to the gradual formation of growth-oriented beliefs. Rather than viewing mindset as an antecedent variable that independently causes success, the model positions mindset as an emergent outcome of repeated engagement in contexts where learners experience that effortful and strategic action can produce improvement. The framework is summarized in Figure 1 and proceeds through six interconnected components.

The framework is also compatible with Complex Dynamic Systems Theory perspectives in SLA, which emphasize emergence, non-linearity, reciprocity, and continual adaptation across multiple interacting components of development (de Bot et al., 2007; Larsen-Freeman and Cameron, 2008). From this perspective, growth-oriented beliefs are not viewed as fixed psychological traits but as evolving products of ongoing interactions among learners, strategies, experiences, and contextual affordances (Xu and Feng, 2024).

This sequence should not be interpreted as linear or deterministic. Instead, each component interacts recursively with the others over time. Learners’ beliefs influence future engagement, engagement shapes subsequent experiences, and those experiences in turn reinforce or weaken beliefs. This recursive, bidirectional account is consistent with self-regulated learning research, which treats beliefs, strategies, and contextual conditions as mutually constitutive rather than sequentially ordered (Pintrich, 2000; Zimmerman, 2002). It is also consonant with the transdisciplinary framework proposed by the Douglas Fir Group (2016), which characterizes SLA as a dynamic system in which individual, social, and institutional dimensions continuously interact and co-shape developmental trajectories over time (Lantolf and Thorne, 2006).

3.1 Repositioning growth mindset as an emergent outcome

A first contribution of the framework is its reconceptualization of growth mindset. Much existing research implicitly treats mindset as a relatively stable internal belief that precedes engagement and predicts performance. However, recent evidence suggests that mindset effects are modest and strongly mediated by contextual and behavioral variables (Macnamara and Burgoyne, 2023; Sisk et al., 2018).

The present framework responds to this limitation by repositioning mindset as developmentally constructed rather than psychologically prior. Learners are unlikely to adopt growth-oriented beliefs simply because they are told to do so. More plausibly, such beliefs emerge when learners repeatedly encounter environments in which effort, strategic adjustment, and persistence are visibly connected to progress. This is consistent with Dweck’s (2017) own more recent theorizing, which acknowledges that mindset beliefs are dynamically responsive to contextual experience rather than fixed properties of the individual. In SLA specifically, the framework builds on the empirical observation by Zarrinabadi et al. (2021) that perceptions of autonomy support directly predict language mindset development, a finding that provides preliminary empirical grounding for the pathway depicted in Figure 1.

3.2 Extending learner autonomy as a generative condition

A second contribution lies in extending conventional understandings of learner autonomy. Rather than conceptualizing autonomy solely as the learner's capacity to manage learning independently, the present model treats autonomy as a relational and contextually enabled condition, produced through interaction between learners and pedagogical environments (Benson, 2011; Reeve and Cheon, 2021).

Autonomy-supportive classrooms provide meaningful choice, rationale, reflective dialogue, and opportunities for decision-making. These experiences matter not only because they increase engagement, but because they help learners perceive a contingency between action and outcome, namely that what they do makes a difference to what they learn. In doing so, they create the experiential basis from which adaptive beliefs may develop. This shifts autonomy from being an instructional endpoint to becoming a generative condition within learning systems.

3.3 Reframing strategies as developmental mechanisms

The primary theoretical contribution of the present framework concerns the role of language learning strategies, and it is here that the model departs most sharply from prior accounts. Strategy research has historically pursued one of two questions: which strategies are associated with better learning outcomes, and how can strategy use be trained (O'Malley and Chamot, 1990; Oxford, 1990; Teng and Zhang, 2020). Both questions treat strategies as instrumental, as means toward the end of improved performance. The present framework proposes a fundamentally different account: that strategies are not only behavioral tools but belief-formation events, and that this dual function is precisely what makes them theoretically indispensable to understanding how growth-oriented beliefs develop in second language learners.

The behavioral function of strategies is well established. Planning, monitoring, revising, and problem-solving during tasks improve performance outcomes across language skill domains (Teng, 2022; Teng and Zhang, 2020), and explicit strategy instruction strengthens metacognitive awareness and self-regulated engagement (Flavell, 1979; Oxford, 2017; Wenden, 1998). What the existing literature has not adequately theorized, however, is what happens psychologically when a strategy works: when a learner deliberately selects an approach, applies it, and observes an improvement in their output. The present framework argues that this moment is not merely a performance event but an interpretive one: the learner acquires experiential evidence that their own deliberate action was consequential. This evidence, accumulated across repeated strategic episodes, can be understood as a mechanism that constitutes the raw material from which growth-oriented beliefs are constructed.

This account draws theoretical support from two adjacent literatures that strategy research has rarely engaged with together. First, Bandura’s (2001) social cognitive theory identifies enactive mastery experiences (direct experiences of successfully executing a behavior) as the most powerful source of self-efficacy beliefs. Strategy use that produces visible improvement is precisely such an experience: the learner does not merely receive feedback that they have improved; they experience themselves as the agent of that improvement. Second, Schunk and Greene (2018) demonstrate that self-regulatory success experiences strengthen learners’ beliefs about the controllability of their own learning processes — a finding that maps directly onto the effort–outcome attribution that growth mindset theory treats as foundational (Dweck, 2006; Yeager and Dweck, 2020). Taken together, these bodies of work suggest that the psychological mechanism through which strategies produce belief change is attribution: repeated experiences of strategic success shift learners’ causal attributions from fixed ability toward effort and method, the very shift that constitutes growth-oriented belief formation.

Critically, this account also explains why stand-alone mindset interventions, which communicate the malleability of ability through explicit messaging, so frequently fail to produce durable belief change (Macnamara and Burgoyne, 2023; Sisk et al., 2018). Telling learners that ability is malleable is not equivalent to providing them with the experiential evidence that makes malleability personally credible. The present framework argues that strategies are the mechanism through which such evidence is generated: they convert the abstract proposition that effort leads to improvement into a concrete, personally enacted experience. In this sense, strategies do not merely support growth mindset; they are the primary experiential basis through which growth mindset becomes possible.

Two boundary conditions on this argument should be stated explicitly. First, the belief-formation function of strategy use depends on progress being visible and attributable to the learner's own action. Where feedback is absent, delayed, or externally attributed, as in highly teacher-directed or exam-driven environments, strategic engagement may improve performance without generating the interpretive evidence that belief formation requires. The strategy works, but the learner does not experience themselves as the cause of it working. Second, the relationship between strategy use and belief formation is not automatic but requires metacognitive scaffolding: learners need to be explicitly supported in connecting their strategic choices to their observed outcomes. Oxford’s (2017) reflective strategy instruction and Teng’s (2022) metacognitive writing instruction both provide pedagogical models for this scaffolding, though neither frames it explicitly in terms of belief formation. The present framework proposes that this reframing is theoretically necessary, and that making the attribution process visible to learners is the critical pedagogical mechanism through which strategy instruction becomes a belief-building practice.

3.3.1 Self-Efficacy

Importantly, the present framework distinguishes growth mindset from self-efficacy, although the two constructs are closely related. Self-efficacy concerns learners’ beliefs about their capability to perform specific tasks successfully (Bandura, 2001), whereas growth mindset concerns broader beliefs about the malleability of ability itself (Dweck, 2006). Strategic success experiences may initially strengthen self-efficacy by increasing learners’ confidence in their capacity to perform particular language tasks. Over time, however, repeated experiences in which effortful and strategic action consistently lead to improvement may contribute to a more generalized belief that language ability can be developed. In this sense, self-efficacy and growth mindset should be viewed as complementary rather than interchangeable constructs, with self-efficacy potentially functioning as one pathway through which strategic experiences contribute to broader beliefs about learning and ability (Bandura, 1997).

3.4 A recursive model of learning development

Finally, the framework advances a recursive model of language learning development. Once learners begin to interpret progress as controllable, they may engage more willingly with challenge, persist longer, and experiment more strategically. These behaviors then generate further opportunities for success, reinforcing adaptive beliefs and completing a positive developmental cycle.

Conversely, environments lacking autonomy support or strategic scaffolding may produce cycles of disengagement in which learners interpret difficulty as evidence of fixed limitation. The recursive logic of the model therefore applies bidirectionally: just as positive cycles may reinforce growth-oriented beliefs, negative cycles may deepen fixed-ability interpretations, a dynamic with important implications for pedagogical design that are elaborated in Section 5.

4 Discussion

The framework proposed in this article has implications that extend beyond the three constructs it directly addresses. By reconceptualizing the relationship among autonomy, strategy use, and mindset, the model contributes to broader theoretical debates in SLA research and raises specific empirical and pedagogical questions that the field has not yet adequately addressed.

4.1 Theoretical implications

The most significant theoretical implication of the framework concerns the status of the growth mindset construct within language learning research. The tendency to treat mindset as a relatively stable belief that predicts motivation, persistence, and achievement has generated a substantial empirical literature, but one whose effect sizes have proven disappointingly modest and whose intervention outcomes have been inconsistent across contexts (Macnamara and Burgoyne, 2023; Sisk et al., 2018). The framework proposed here offers an explanation for this pattern: if mindset is partly an outcome of accumulated learning experience rather than solely a stable individual difference, then interventions that target beliefs directly, without modifying the learning environments that produce them, are unlikely to achieve durable effects.

This interpretation does not abandon mindset theory. It engages with it more precisely by locating belief formation within a broader developmental ecology. Dweck’s (2017) own recent theorizing supports this move: she acknowledges that mindset beliefs interact dynamically with contextual experience and that the conditions under which effort is perceived as consequential matter as much as the beliefs themselves. The present framework operationalizes this insight for language learning contexts by identifying learner autonomy and strategic engagement as the mechanisms through which such conditions are created and sustained.

A second theoretical implication concerns the construct of learner autonomy. Traditional accounts, particularly those associated with Holec (1981) and the early European tradition, emphasized autonomy as an individual capacity, treating it as something learners either possess or lack. The present model, consistent with more recent sociocultural and self-determination perspectives (Noels et al., 2019; Reeve and Cheon, 2021), treats autonomy as relationally produced. Agency is not merely exercised; it is enabled or constrained by pedagogical structures. This reframing is supported by empirical work showing that autonomy develops most robustly in classrooms characterized by ongoing negotiation and reflective practice (Little et al., 2017), and by theoretical arguments that self-directed learning capacities are always shaped by the social and institutional conditions in which learners are embedded (Reinders and Benson, 2017). This reframing carries methodological consequences: studies examining the effects of autonomy on learning outcomes need to measure not only learner dispositions but also the specific environmental affordances that make autonomous action possible or impossible (Ryan and Deci, 2020).

A third implication concerns the role of language learning strategies within motivational and belief-oriented accounts of SLA. Strategy research has historically proceeded along a separate track from motivational psychology. The present framework proposes that these tracks intersect precisely at the point where behavioral engagement produces experiential evidence for or against growth-oriented beliefs. Strategies are therefore theoretically interesting not only as performance-enhancing behaviors but as belief-formation events, a framing that may open productive new directions for both strategy research and motivational psychology in SLA.

4.2 Research implications

The framework raises several specific directions for future empirical research. First and most directly, future studies should examine whether strategy use mediates the relationship between autonomy-supportive instruction and the development of growth-oriented language mindsets. Zarrinabadi et al. (2021) established that autonomy support predicts mindset, but the intervening mechanism remains underspecified. Multi-variable designs employing mediation analysis or structural equation modelling could test whether strategic engagement and progress visibility account for this relationship. Comparable research designs have been applied to university transition contexts, where shifts in learners’ social identities and perceived agency have been shown to influence academic engagement and persistence (Edsall, 2024); future research might profitably examine analogous dynamics within language learning settings. Self-access learning environments, in which learners make autonomous choices about resources and activities outside formal instruction, offer particularly promising sites for this work, since they make learner agency and strategic decision-making unusually visible (Gardner and Miller, 1999).

Second, longitudinal research designs are essential. The recursive dynamics proposed in Figure 1 cannot be captured in cross-sectional data, which offers only static snapshots of what the model treats as developmental processes unfolding over time. Tracking learners across semesters or academic programs, and measuring beliefs, strategic behavior, and perceived progress at multiple time points, would provide substantially stronger evidence regarding the proposed recursive cycle.

Third, greater methodological diversity is warranted. Quantitative studies could model indirect pathways between autonomy support, strategic engagement, and mindset development. Qualitative and mixed-methods approaches could examine how learners narrate their experiences of control, progress, and difficulty over time, providing the kind of process-level evidence that quantitative designs alone cannot supply.

Finally, and importantly, future research should investigate negative recursive cycles as explicitly as positive ones. Learners who repeatedly encounter difficulty without strategic support, or who operate in environments that attribute outcomes to fixed ability rather than to effort and method, may strengthen fixed beliefs and disengage from strategic action. Understanding how such cycles form, persist, and can be interrupted is as theoretically important as understanding their positive counterparts.

4.3 Contextual considerations

The relationships proposed in this article are unlikely to operate uniformly across all instructional settings. Institutional constraints, assessment cultures, class size, teacher beliefs, and sociocultural norms may each shape the opportunities learners have for autonomous action and strategic engagement. In highly exam-driven systems, for example, pedagogical choices may be constrained in ways that limit learner agency and experimentation, reducing the frequency with which learners experience the effort–progress connection that the framework identifies as foundational to belief change. Research on learner identity transitions in higher education further suggests that contextual shifts, such as moving from secondary to university study, can substantially alter the conditions under which autonomous engagement is possible, with implications for how beliefs about ability are formed and revised across different institutional phases (Edsall, 2024).

Similarly, in strongly teacher-centered contexts, the meaning of learner autonomy may be understood and enacted quite differently than in explicitly learner-centered environments. Strategies that are effective in one context may be less available or less valued in another. This suggests that growth-oriented belief formation should be understood as context-sensitive rather than universal, a point that has practical implications for any attempt to generalise findings from one educational setting to another.

4.4 Boundary conditions of the framework

A theoretical framework that does not specify when it does not apply is unfalsifiable, and unfalsifiable frameworks have limited scientific value. The present section therefore identifies four conditions under which the positive recursive cycle proposed in Figure 1 is predicted to stall or break down. These boundary conditions are not incidental qualifications; they are derivable predictions of the framework itself, and as such they constitute an empirical agenda for future research.

The first boundary condition concerns the visibility of progress. The framework predicts that strategic engagement leads to growth-oriented belief formation only when improvement is perceptible to the learner. Language learning is a domain in which progress is notoriously slow and difficult to detect, particularly in early and intermediate stages. Where learners engage strategically but cannot observe any traceable improvement, because feedback is absent, because gains are too gradual to be perceived, or because assessment focuses on absolute performance rather than development, the experiential evidence that belief formation requires will not accumulate. The positive cycle therefore depends critically on pedagogical designs that make progress visible, as elaborated in Section 5. Where such designs are absent, even learners in autonomy-supportive environments may not develop growth-oriented beliefs.

The second boundary condition concerns attributional scaffolding. As argued in Section 3.3, the belief-formation function of strategy use depends not only on producing improvement but on learners attributing that improvement to their own deliberate action. In the absence of explicit reflective scaffolding (opportunities to connect strategic choices to observed outcomes) learners may attribute improvement to teacher quality, task difficulty, luck, or natural aptitude rather than to effort and method. Where such misattribution is chronic, strategic success will not generate growth-oriented beliefs even when improvement is objectively present. This prediction connects the framework to attribution theory (Weiner, 1985) and suggests that metacognitive reflection is not merely a pedagogical nicety but a structural requirement of the belief-formation pathway.

The third boundary condition concerns challenge calibration. The framework predicts that positive recursive cycles require challenge to be within a range that is effortful but attainable, analogous to Vygotsky (1978) zone of proximal development. Tasks that are too easy produce improvement without effort, which provides no evidence for the effort–outcome connection. Tasks that are too difficult produce effort without improvement, which may actively strengthen fixed-ability interpretations by confirming that trying harder does not help. The growth-oriented belief formation pathway therefore operates only within a calibrated difficulty range, and instructional designs that do not maintain this calibration will predictably disrupt the positive cycle.

The fourth boundary condition concerns institutional and cultural context. The framework is grounded in a conception of learner agency that assumes meaningful choice, reflective dialogue, and consequential action are available to learners within their instructional environments. In strongly teacher-directed, examination-oriented, or shame-avoidant cultural contexts, these conditions may be structurally unavailable regardless of individual teacher intentions. Where institutional cultures consistently attribute outcomes to fixed ability, through streaming, competitive grading, or public error correction, the interpretive conditions required for belief formation are undermined at the systemic level. The framework therefore predicts weaker effects in such contexts, and this prediction should be empirically tested across diverse educational systems. The geographic concentration of existing autonomy and mindset research in specific cultural contexts (Lou and Noels, 2024; Zarrinabadi et al., 2021) means this boundary condition remains largely untested.

4.5 Limitations

Several limitations of the present article should be acknowledged. First, and most fundamentally, the framework is conceptual rather than empirical. The proposed relationships among autonomy, strategy use, and mindset development remain theoretically inferred and require direct empirical testing across diverse learner populations and instructional contexts before their scope and boundary conditions can be established.

Second, the framework synthesizes literature from multiple traditions (SLA, self-regulated learning, and motivational psychology) that employ different definitions, measures, and theoretical assumptions. Some degree of conceptual friction is unavoidable in such integration, and the proposed model may exceed current empirical consensus at certain points.

Third, although the model emphasizes recursive and bidirectional dynamics, it remains heuristic rather than formally operationalized. The framework identifies the components of the proposed developmental process and the directionality of relationships among them, but does not specify the precise quantitative parameters or temporal dynamics that a fully formalized model would require. These limitations reflect the article's primary purpose: to provide a theoretically coherent and empirically motivated platform for future inquiry, rather than to offer a complete and fully tested account.

5 Pedagogical implications

The framework proposed in Section 3 generates three specific pedagogical implications. These are not general good-teaching advice; each derives from a specific theoretical claim within the framework and would not follow from prior instrumental accounts of strategy use alone.

5.1 Move beyond stand-alone mindset interventions

Many educational initiatives attempt to improve learner outcomes by delivering direct mindset messages, for example telling learners that intelligence is malleable or that effort produces growth. While such messages may have motivational value in the short term, the present framework suggests their effects are likely to remain limited unless supported by learners’ actual classroom experiences. This is consistent with evidence that even well-designed mindset interventions produce durable effects only when learners are also positioned within environments where effort and improvement are genuinely connected (Yeager and Dweck, 2020), and that digitally delivered mindset programmes, while scalable, show modest benefits that vary substantially by prior achievement and contextual support (Paunesku et al., 2015). Teachers may therefore be better served by embedding growth-oriented principles within everyday pedagogical practice rather than treating mindset as a discrete intervention target. Learners are more likely to internalize adaptive beliefs when they repeatedly experience that revision, persistence, and strategic adjustment lead to observable improvement.

5.2 Teach strategies as belief-formation practices

The framework's central claim, that strategies are belief-formation events as well as performance tools, has a direct instructional consequence: strategy instruction must be designed to make the connection between strategic action and observed improvement explicitly visible to learners. It is not sufficient to teach strategies procedurally. Learners must be supported in attributing their improvement to their own deliberate choices, because it is this attribution, rather than the improvement itself, that generates growth-oriented beliefs. Instruction should therefore include structured reflection on the relationship between specific strategic decisions and their outcomes. Classroom modelling might include:

planning before undertaking complex tasks

monitoring comprehension or production during tasks

revising and adjusting after receiving feedback

selecting and switching strategies when difficulties arise

reflecting explicitly on what produced improvement

When learners develop metacognitive understanding of the connection between strategy use and outcomes, they are better positioned to interpret improvement as the product of deliberate action, which is precisely the interpretation that growth-oriented beliefs require.

5.3 Make progress visible and attributable

The boundary conditions identified in Section 4.4 make clear that progress visibility is not a pedagogical nicety but a structural requirement of the belief-formation pathway. When improvement is imperceptible, strategic engagement cannot generate the experiential evidence that growth-oriented beliefs require, regardless of how autonomy-supportive the environment is. Teachers must therefore design learning environments in which progress is concrete, traceable, and regularly foregrounded, and in which learners are explicitly supported in connecting that progress to their own strategic action rather than to external factors. This design imperative is supported by research on effective feedback, which demonstrates that feedback oriented toward learning processes and development over time is substantially more beneficial than feedback focused on performance levels alone (Hattie and Timperley, 2007), and by work on sustainable assessment practices that argues for building learners’ capacity to monitor and judge their own work across time (Boud and Falchikov, 2006). Useful practices include:

portfolio assessment in which earlier and later work are preserved and compared

draft-to-final comparisons that make revision visible

repeated performance tasks using the same or parallel prompts

self-rating checklists completed at multiple points across a course

feedback focused explicitly on development over time rather than on fixed performance levels

When learners can see and articulate evidence of their own growth, beliefs about improvement become experientially grounded rather than abstract, which according to the framework is a necessary condition for their durability.

6 Conclusion

This article has proposed an integrated and process-oriented framework for understanding the relationship among learner autonomy, language learning strategies, and growth mindset in second language learning. The framework makes a specific theoretical argument: growth-oriented beliefs are more accurately understood as emergent outcomes of repeated learning experience than as primary predictors of achievement. Learners do not develop adaptive beliefs through messages about the malleability of intelligence alone; they develop them through environments in which effort, strategic engagement, and self-directed action are visibly and repeatedly connected to progress.

Within this model, learner autonomy functions not as an endpoint of instruction but as a generative condition: the pedagogical context that makes consequential action possible. Language learning strategies serve as the mediating mechanism through which pedagogical opportunity is transformed into psychologically meaningful experience. Together, these components create recursive cycles in which belief and behavior continuously co-develop, each shaping the conditions for the other.

The framework contributes to current debates in three specific ways. First, it offers an explanation for the modest and inconsistent effects of stand-alone mindset interventions: such interventions target beliefs without modifying the learning environments through which beliefs are formed. Second, it extends learner autonomy research by connecting autonomy to belief formation, a connection that has received preliminary empirical support (Zarrinabadi et al., 2021) but has not yet been theorized in process terms. Third, it repositions language learning strategies as belief-formation events as well as performance-enhancement tools, opening a new dimension for strategy research that connects it to motivational and psychological accounts of SLA.

Several limitations qualify these contributions. The framework is conceptual rather than empirical, and the proposed relationships require direct testing across diverse contexts before their scope and boundary conditions can be established. The recursive dynamics depicted in Figure 1 demand longitudinal research designs that have not yet been applied to this specific constellation of constructs in SLA. And the framework's integration of multiple theoretical traditions inevitably involves some simplification of the complex theoretical debates within each.

Future research should empirically test the proposed pathways, particularly the mediating role of strategic engagement between autonomy support and mindset development, using longitudinal, mixed-methods, and mediation-based designs. Equal attention should be given to negative recursive cycles, in which restricted agency and absent strategic scaffolding may reinforce fixed-ability beliefs and further disengage learners from productive effort.

Ultimately, this article argues for a reconceptualization of language learning development: not as the product of isolated psychological variables acting independently on achievement, but as a dynamic system in which pedagogy, behavior, and belief continuously co-develop over time. This view is consistent with the transdisciplinary understanding of SLA advanced by the Douglas Fir Group (2016), which emphasizes that language development is always simultaneously individual, social, and contextual, and cannot be adequately understood by privileging any single dimension in isolation. The framework therefore does not deny the importance of mindset. Rather, it repositions mindset within a broader developmental system in which beliefs emerge through sustained interactions among agency, strategic engagement, feedback, and perceived progress. Growth mindset, in this account, is not something learners bring to the classroom; it is something that well-designed classrooms help them build.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI was used to edit the original draft for an academic tone of language.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Mikhail Gradovski, University of Stavanger, Norway

Reviewed by: Dominic Edsall, Kyoto Prefectural University of Medicine, Japan

Baljinder Singh Maghar Singh, SEGi University, Malaysia