Abstract
Albert Bandura's theory of self-efficacy is one of the most cited constructs in the history of psychology. The phenomenon it describes is real: explicit declarations of future orientation predict subsequent performance differences across a wide range of domains. What this paper argues is that the ontological story Bandura built around this finding is wrong. Self-efficacy is not an internal cognitive property of isolated individuals. It is a coordinative technology, an interrecursive commitment structure that reorganizes the recursive ecology within which action subsequently occurs. Living Value Theory offers a reinterpretation that preserves what is genuinely valuable in the self-efficacy tradition while dissolving its characteristic confusions about location, causation, and mechanism.
I. A Half-Century of Overconfidence
Albert Bandura's theory of self-efficacy is one of the most cited and applied constructs in the history of psychology. Since its formal introduction in 1977, the literature has grown to tens of thousands of studies, spanning domains from academic achievement to clinical intervention, athletic performance to organisational behaviour. Bandura presented self-efficacy as a general psychological mechanism. His research and that of the tradition he created has been taken to demonstrate that what people believe about their own capacities is a significant determinant of what they subsequently achieve.
The phenomenon is real. The practical applications have helped people. The correlations are genuine. What this paper argues is something more precise than a general critique: that the ontological story Bandura built around his finding is wrong, and that this wrongness has had systematic consequences for what the tradition has been able to see and what it has been unable to see.
Because self-efficacy has been mistaken as an individual cognitive property, interventions have been designed to modify what goes on inside isolated individuals rather than to restructure the mediational configurations and recursive ecologies within which those individuals act. This is not merely a theoretical error. It is a practical one, with consequences for the design of educational programmes, clinical interventions, and organisational change efforts.
Living Value Theory offers a reinterpretation that preserves what is genuinely valuable in the self-efficacy tradition while dissolving its characteristic confusions about location, causation, and mechanism.
II. What Bandura Thought He Found
Bandura's theory of self-efficacy emerged from his broader social cognitive theory, which was itself a response to the limitations of behaviourism. Behaviourism had located the determinants of action in the external environment, specifically in reinforcement histories. Bandura argued, correctly, that this was insufficient: what people believe about themselves and their capacities matters for what they subsequently do, and these beliefs cannot be reduced to reinforcement histories.
Several features of Bandura's account are worth noting precisely, because they reveal exactly where the ontological error enters. First, Bandura located self-efficacy inside the individual, as a cognitive property, specifically as a belief about one's capacity to perform specific behaviours. Second, he treated this property as causally efficacious: the belief is what produces the performance difference. Third, he treated it as measurable through verbal self-report: asking people to rate their confidence that they can perform specified tasks gives access to the relevant internal state. Fourth, he treated it as domain-specific but generalisable: people have different self-efficacy beliefs for different tasks, but the mechanism is the same across domains.
The measurement apparatus that self-efficacy research relies on encodes all these assumptions. Participants are asked to rate their confidence that they can perform specified tasks. The ratings are treated as measurements of an internal cognitive state that exists independently of the reading.
None of these assumptions is self-evident, and several are directly challenged by the LVT account. But before developing that account, it is worth acknowledging what Bandura saw clearly and what remains genuinely valuable in the tradition he created. He was right that something happens in the relationship between explicit self-statement and subsequent performance. He was right that this something is real, robust, and practically significant. He was right that it is not simply reducible to past reinforcement history, and that something about the organism's relationship to its own future capacities matters for what it subsequently achieves. The phenomenon is there. The ontological story is wrong.
III. The Distributed Character of Capacity
The first and most fundamental error in Bandura's framework is the location of self-efficacy inside the individual. Consider what it actually means to say that someone can climb a mountain, conduct an orchestra, raise a child, perform surgery, or write a novel. In each case, the capacity described is not located inside any individual. It is distributed across the five mediations through which coordination takes place.
Embodiment matters. The climber's capacity depends on her fitness, acclimatisation, injury history, and the physical state of her body on the day of the attempt. None of this is inside her psychology. The surgeon's capacity depends on the fine motor coordination she has developed over years of practice, a coordination that lives in her hands and nervous system rather than in her beliefs about herself. The orchestral conductor's capacity is partly in her hearing, her proprioception, her capacity to be affected by sound in ways that allow real-time coordination with dozens of musicians simultaneously. These are embodied capacities, not cognitive properties.
Being-with matters. Almost every capacity that self-efficacy research concerns is exercised in relation to others. The climber has companions whose competence, attention, and reliability are part of the capacity. The surgeon has a theatre team. The conductor has her orchestra. The novelist has editors, readers, and the entire literary tradition within which her work is located. The researcher in Bandura's experiments who rates her confidence that she can perform some task is already embedded in a social context that shapes what confidence means and how its expression functions. Capacity, in every case, is co-produced through ongoing relations rather than residing in an isolated individual.
Dwelling matters. The conditions of light, temperature, altitude, season, and time of day within which a task is performed are part of the capacity, not mere external constraints on an underlying psychological property. The climber's actual capacity varies with altitude and weather. The surgeon's varies with fatigue and the time of day relative to her circadian rhythms. The novelist's varies with the quality of the space and time available for writing. None of this is incidental to the capacity; it is constitutive of it.
Multimateriality matters. The climber's capacity is partly in her equipment: rope, harness, boots, protection. The surgeon's is partly in her instruments. The conductor's is in the acoustic design of the concert hall. The novelist's is in her access to the library, the computer, the uninterrupted hours. Material affordances and constraints are not separable from capacity; they are part of what capacity is. When Bandura asks whether someone can do something, the question assumes that the doing is a property of the person rather than an emergent feature of the entire mediational configuration.
Multisymbolization matters. What it means to have written a novel, conducted a successful surgery, or climbed a mountain is partly constituted by the symbolic frameworks within which these activities are recognised, evaluated, and understood. The capacity to do these things includes the capacity to make sense of them within the symbolic systems that give them their significance. A surgeon operating in an unfamiliar institutional and symbolic context is less capable than the same surgeon in her home environment, not because her internal beliefs have changed but because the symbolic infrastructure that makes her competence operative has changed.
The self-efficacy questionnaire collapses all of this into a single number: how confident are you that you can do this? This is not simply a pragmatic simplification. It is an ontological claim: that the relevant variable is inside the person and can be indexed by their verbal report. LVT argues that this claim is wrong, and that its wrongness has systematic consequences for what the self-efficacy tradition has been able to see and what it has been unable to see.
IV. The Recursivity Problem
The second error in Bandura's framework concerns recursivity. Bandura treats self-efficacy as a property that causes performance. The person has a belief; the belief causes them to persist; persistence causes better performance. This is a linear causal story. LVT argues that the relationship between self-statement and performance is not linear but recursive, and that this recursivity is precisely what makes the phenomenon interesting.
Consider what actually happens when a person makes a confident self-statement about a future performance. The statement does not merely express a pre-existing internal state. It enters the social world as a fact. Others hear it or know of it. The person themselves has now said something that can be confirmed or disconfirmed by subsequent events. The statement reorganizes the field of expectations, obligations, and possibilities within which the subsequent performance will occur. This is not a causal story about an internal property producing an external outcome. It is a story about how a social act restructures a recursive ecology.
The recursivity operates at multiple levels simultaneously. At the level of the individual, the declaration creates a loop between the present self and the future self: the future self will inhabit a world in which this declaration was made, and this fact will be part of what the future self has to reckon with. At the level of the social field, the declaration creates loops between the individual and others: their expectations, their responses, their willingness to support or challenge or hold accountable. At the level of the symbolic field, the declaration creates loops between the individual and the cultural meanings that give the declaration its weight: the social significance of having said something, the implications for identity and standing of having committed to something publicly.
Bandura's framework cannot see any of this because it is committed to a linear, individualist ontology. The self-efficacy belief is inside the person; it causes behaviour; the behaviour produces outcomes. The recursivity, the social restructuring, the temporal reorganization of the field of possibilities: all of this is invisible to a framework that begins by locating the relevant variable inside the individual.
LVT distinguishes three fundamental recursivity domains: non-recursive, self-recursive, and interrecursive. This distinction is crucial for understanding what self-efficacy research has actually been measuring.
Non-recursive processes are those in which the elements of the process do not respond to expectations about them. Gravity does not adjust its behaviour based on what physicists expect it to do. Chemical reactions do not proceed differently because chemists believe they will. In these domains, the relationship between belief and outcome is purely epistemic: accurate beliefs help you navigate the process, but the process itself is indifferent to your beliefs about it.
Self-recursive processes are those in which the organism's relationship to its own states and processes is part of what determines those states and processes. Pain is partly constituted by the organism's relationship to nociceptive signals. Mood is partly constituted by the organism's relationship to its own affective states. In these domains, the relationship between belief and outcome is genuinely causal, but the causation runs through the organism's own self-regulatory processes rather than through the social field.
Interrecursive processes are those in which multiple parties recursively affect each other's expectations, orientations, and behaviours. Most of the domains in which self-efficacy research has been conducted are interrecursive: academic performance, athletic competition, clinical outcomes, organisational behaviour. In these domains, what one party expects and declares affects what others expect and do, which affects what the first party subsequently does, which affects what others subsequently expect, and so on.
The crucial observation is that self-efficacy's effects are overwhelmingly concentrated in interrecursive domains. The mechanism is not the internal belief but the social restructuring that the declaration of belief performs. Bandura's generalisation across these domains conflates processes that operate quite differently. The result is a framework that works reasonably well in interrecursive domains, where the declaration does real coordinative work, but systematically overpredicts in non-recursive domains, where the declaration can only affect the individual's own self-regulatory processes and cannot restructure the social field.
V. The Domain Problem
The domain problem is related to the recursivity problem but distinct from it. Bandura claimed that self-efficacy is a general psychological mechanism that operates across all domains of human performance. LVT argues that this generalisation is unwarranted, and that the evidence for it is systematically misleading.
Consider the difference between climbing a mountain and conducting an orchestra. Both are domains in which self-efficacy research has been conducted, and in both, some correlation between self-efficacy ratings and performance has been found. But the mechanisms are quite different. The climber's performance depends heavily on physical factors, equipment, and conditions that are largely independent of the social field. The conductor's performance depends heavily on the recursive coordination of dozens of musicians whose responses to her are partly constituted by their expectations of her, which are partly constituted by her expressed confidence, which is partly constituted by their responses to her. The self-efficacy questionnaire captures something real in both cases, but what it captures is different in each case, and the difference matters.
In the climbing case, the self-efficacy rating is partly capturing something about the climber's actual physical preparation and readiness, which is genuinely predictive of performance. It is also partly performing a pre-performance declaration that reorganizes the climber's own self-regulatory processes. But the social restructuring effect is limited, because the mountain does not respond to the climber's confidence.
In the conducting case, the self-efficacy rating is doing something quite different. It is partly a declaration of orientation that enters the social field of the orchestra and restructures their expectations and responses. The musicians who believe their conductor is confident perform differently than musicians who believe their conductor is uncertain, not because of any internal property of the conductor but because of the interrecursive dynamics of the ensemble. The self-efficacy questionnaire, in this case, is not measuring an internal property at all. It is participating in the recursive ecology that it purports to measure.
The evidence for the generality of self-efficacy is therefore misleading. The correlations are real, but they are produced by different mechanisms in different domains. In non-recursive domains, the correlation is modest and is explained by the relationship between self-assessment and actual preparation. In interrecursive domains, the correlation is stronger and is explained by the social restructuring that the declaration performs. Bandura's framework, committed to a single internal mechanism, cannot see this difference and therefore cannot explain why the correlations vary in the ways they do.
VI. The Measurement Performs What It Measures
There is a further problem with the self-efficacy paradigm that is perhaps the most theoretically significant of all: the measurement instrument does not measure a pre-existing state. It constitutes a new state.
Consider what actually happens when a research participant fills out a self-efficacy questionnaire. They are asked, typically in a formal institutional context, to rate their confidence that they can perform specified tasks. The act of filling out this questionnaire is not a neutral measurement of an ongoing cognitive state. It is a formal declaration, made in a socially significant context, of the participant's orientation toward future tasks. The participant is not reporting a belief; they are performing a commitment.
This is not a measurement of a pre-existing cognitive property. It is a miniature sankalpa: a formal declaration of future orientation that enters the social world and reorganizes the recursive ecology within which subsequent performance will occur. The questionnaire does not measure self-efficacy; it produces it, in the sense of producing the social fact of a declared commitment that will subsequently affect performance through the interrecursive mechanisms described above.
This is not a cynical point about experimental artefacts. It is a positive theoretical claim about what the self-efficacy questionnaire is doing. It is doing something real and important. It is just not doing what Bandura thought it was doing. It is not measuring an internal cognitive property. It is performing a social act that has real consequences for the recursive ecology within which the participant will subsequently act.
The evidence for this interpretation is not merely theoretical. It is implicit in one of the most well-replicated findings in the self-efficacy literature: that self-efficacy questionnaires predict performance better when they are completed immediately before the performance than when they are completed well in advance, and better in public contexts than in private ones. If self-efficacy were an enduring internal property, its predictive power should not vary with the timing or social context of its measurement. That it does vary in exactly the ways that the sankalpa account predicts is evidence that what the questionnaire is doing is not measurement but participation.
VII. The Sankalpa and Interrecursive Self-Recursivity
In the South Asian ritual tradition, a sankalpa is a formal declaration of future intention. Before undertaking a pilgrimage, a fast, a rite of passage, a period of study, or a significant work, the practitioner declares: I shall do this. The declaration is made in a specific form, at a specific moment, typically in the presence of witnesses or in a ritual context that lends the declaration social and cosmic weight. It is not a private wish or a hopeful intention. It is a public commitment that enters the social world as a fact and from which the practitioner cannot easily retreat without consequences for their social identity and relational standing.
The sankalpa does not work through magic. Its logic is entirely transparent. By making a formal public commitment, the practitioner reorganizes their future coordinative field. Future versions of themselves will inhabit a world in which this commitment was made. Others who heard or know of the commitment will hold expectations based on it. The practitioner's own self-understanding will be partly organized around the commitment. Actions that support the commitment will be easier; actions that undermine it will create friction. The declaration does not guarantee success, but it restructures the recursive ecology within which success or failure will occur.
This is the structure that LVT calls interrecursive self-recursivity. It is recursive because it involves a loop between the self's declaration and the self's subsequent states and actions. It is self-recursive because the self is one of the parties in this loop, the future self is committed by the present self's declaration. It is interrecursive because the declaration enters the social world and engages the recursive processes of others, their expectations, their responses, their willingness to collaborate or support or hold the practitioner accountable.
National Novel Writing Month works through exactly this structure. Participants publicly declare in November that they will write fifty thousand words in thirty days. The declaration is made in an online community in which others know about it, where progress is tracked, where participants encourage each other, and where completing the commitment carries social recognition and failing to complete it carries mild social cost. The writing that results is not primarily explained by any change in participants' internal beliefs about their capacity. It is explained by the coordinative restructuring that the public declaration has performed: the future has been organised differently by the act of declaring, and the recursive field of social expectation and mutual accountability has been brought to bear on what would otherwise be a purely private struggle.
Recovery programmes, doctoral deadlines, marriage vows, sporting commitments made publicly before significant competitions: these all work through the same structure. The declaration does not merely express a state; it constitutes a new social and temporal fact. The person who has publicly committed to sobriety inhabits a different recursive ecology than one who has merely decided privately to stop drinking. The doctoral student who has submitted a written schedule to her supervisor inhabits a different recursive ecology than one who intends privately to finish by the same date. The difference is not in the content of the intention but in its social and temporal reality: the declaration has made the commitment part of the interrecursive field within which subsequent action will occur.
Bandura's self-efficacy questionnaire is, structurally, a miniaturised and stripped-down version of this ancient technology. The participant declares, in a formal institutional context, their orientation toward a future task. The declaration enters a social world, it is read by researchers, it is part of an experimental protocol, it carries the implicit social pressure of having stated something that can be verified or falsified by subsequent events. The participant cannot easily pretend, to themselves or to others, that they did not make the declaration. The declaration reorganizes their recursive ecology in exactly the way that a sankalpa reorganizes the practitioner's. Psychology, interpreting this as measurement of an internal cognitive property, has been describing the shadow of a ritual performance.
VIII. What Is Actually Doing the Work
If self-efficacy is not a cognitive property inside the individual, what is actually doing the work in the cases where high self-efficacy predicts better performance? The LVT account specifies several distinct mechanisms, each of which operates through the mediational structure of the mesocosm rather than through internal cognitive properties.
First, there is the pre-performance declaration effect. A formal declaration of orientation toward a future task, made in a socially significant context, reorganizes the recursive ecology within which the task will be performed. This is the sankalpa mechanism described above. It operates through the interrecursive field, not through internal cognition.
Second, there is the embodied readiness signal. The person who reports high self-efficacy before a performance is often reporting something real about their current mediational configuration: their body is prepared, their equipment is ready, their social supports are in place, their symbolic framework for the task is coherent. The self-efficacy rating, in this case, is a proxy for the overall mediational readiness of the person-in-context, not a measurement of an internal cognitive property. The predictive power of the rating comes from this proxy relationship, not from any direct causal effect of belief on performance.
Third, there is interrecursive entrainment. In social domains, the expressed confidence of one participant restructures the expectations and behaviours of others in ways that alter the subsequent coordinative field. The confident conductor elicits different responses from her orchestra than the uncertain conductor, and these responses affect what is subsequently possible. The confident student elicits different responses from her teacher than the uncertain student. The confident athlete affects the expectations and behaviours of competitors, coaches, and spectators in ways that alter the subsequent competitive ecology. None of this requires any internal cognitive property; it requires only that the declaration enter the interrecursive field and do its coordinative work.
Fourth, there is what might be called the persistence mechanism. Even granting none of the above, there is a domain in which explicit self-belief does seem to operate as Bandura described: the decision to continue trying in the face of difficulty. A person who has declared that they can do something, and who has made this declaration in a socially significant context, faces a higher threshold for giving up than one who has made no such declaration. This is not because of any internal cognitive property but because giving up would require them to revise their public declaration, with consequences for their social identity and relational standing. The persistence effect is real; its mechanism is social and recursive, not cognitive and individualist.
IX. Implications: What Changes
The LVT account of self-efficacy is not merely a theoretical refinement. It has practical consequences for how the phenomenon should be understood, measured, and applied.
The individualist assumption should be abandoned. Interventions designed to boost individual confidence by modifying internal beliefs are addressing the wrong level of analysis. What matters is the mediational configuration and the recursive ecology within which the person acts. Interventions should be designed to restructure these, not to modify internal cognitive states.
The domain distinction should be taken seriously. Self-efficacy interventions in non-recursive domains, where the process being acted upon does not respond to expectations, should be evaluated with appropriate scepticism. The mechanism that makes self-efficacy interventions effective in interrecursive domains is absent in non-recursive ones. Applying the same intervention across domains without attention to this difference is likely to produce inconsistent results, which is exactly what the literature shows.
The timing question should be taken seriously. Explicit confidence assessments should be understood as pre-performance declarations rather than neutral measurements of ongoing cognitive states. Their effects are strongest when they are made immediately before performance, in socially significant contexts, with real consequences for social identity and relational standing. Designing interventions that create these conditions is likely to be more effective than designing interventions that attempt to modify internal beliefs in the abstract.
The ritual dimension should be acknowledged rather than suppressed. The insight that Bandura's questionnaire functions as a miniature sankalpa is not a reason to dismiss the tradition's findings. It is a reason to take seriously the ritual and social infrastructure that makes the declaration effective. Recovery programmes that use public commitment rituals, educational programmes that build in formal declaration moments, athletic training regimes that incorporate public commitment structures: these are not irrational additions to evidence-based practice. They are the actual mechanism through which the phenomenon operates.
Most importantly, the finding that declared commitment reorganizes recursive ecologies should prompt attention to the social conditions that make such declarations real. A self-efficacy questionnaire administered to an isolated individual in a context where the declaration carries no social weight will have limited effects. The same declaration made in a community of practice, with real social consequences for follow-through, will have substantial effects. The difference is not in the individual's internal cognitive state but in the recursive ecology within which the declaration is made.
X. What Bandura Actually Discovered
Albert Bandura discovered that explicit declarations of future orientation, made in socially significant contexts, predict subsequent performance differences across a wide range of domains. This is a genuine and important finding. It has been replicated thousands of times. It has practical applications that have helped people. It deserves the attention it has received.
What Bandura did not discover is a general internal cognitive mechanism called self-efficacy. He did not discover that people perform better because they have a property inside them called confidence. He did not discover that this property can be measured by asking people to rate their confidence. He did not discover that interventions designed to modify this internal property will reliably improve performance. The evidence for these claims is much weaker than the evidence for the basic finding, and the theoretical framework built around them is, as this paper has argued, ontologically mistaken.
What he discovered, without recognising what he had found, is the contemporary Western form of a coordinative technology that human societies have been using for as long as there is evidence of ritual practice: the formal declaration of future orientation in a socially significant context, which reorganizes the recursive ecology within which subsequent action occurs. This technology is ancient, widespread, and effective. It works not through internal cognitive mechanisms but through the interrecursive restructuring of social fields, temporal horizons, and identity commitments.
Modern psychology found this technology in the laboratory, stripped of its ritual context, its communal infrastructure, and its temporal depth, administered through questionnaires to anonymous participants in exchange for course credit, and interpreted it as evidence of an internal cognitive property. The interpretation was wrong. The finding was real.
Living Value Theory offers that theory. Interrecursive self-recursivity, the process by which a public declaration of future orientation reorganizes the recursive ecology within which action subsequently occurs, is the mechanism that Bandura's research has been tracking without being able to name. The self-efficacy literature, reinterpreted through this framework, becomes a rich body of evidence about how interrecursive commitment structures work, when they are effective, and what conditions are required for them to produce the coordinative effects that Bandura documented.
Bandura's lasting contribution is not the cognitive theory he built around his finding. It is the finding itself: that something about explicit forward-oriented declarations matters for what comes after. Living Value Theory can explain why.