A Synthesis of Attractor Dynamics, Predictive Processing, and Agency Theory

Prepared for submission to a journal of cognitive neuroscience and philosophy of mind

Daryl Costello: Independent Researcher

Rosendale / High Falls, New York, USA

Correspondence: Daryl.costello@outlook.com

July 2026

Abstract

The question of how a coherent sense of self persists across time (despite ceaseless neural, bodily, and environmental change) remains one of the most consequential unsolved problems in cognitive neuroscience and philosophy of mind. Static models of identity, whether trait-based, narratively constituted, or socially constructed, capture important features of selfhood but fail to account for its dynamic, embodied, and time-varying character. This paper introduces the Identity Attractor Theory (IAT), a formal theoretical framework proposing that the self is best understood as a dynamic attractor state; a high-dimensional basin in the space of behavioral, affective, cognitive, and relational configurations to which the nervous system characteristically returns after perturbation.

IAT draws on three converging scientific traditions: dynamical systems theory, which provides the attractor metaphor and the mathematics of basin geometry; predictive processing and the free energy principle, which explain the neural mechanisms of attractor maintenance through hierarchical generative modeling; and agency theory, which grounds the self’s capacity to act in ways that preserve its own characteristic configuration. The theory maps specific neural systems onto components of the attractor model: the default mode network (DMN) as the core geometry-maintaining substrate; predictive processing as the implementation of the return function; interoception as the body-level anchoring mechanism; hippocampal memory binding as the temporal coherence constraint; and hemispheric specialization as a dual-channel monitoring architecture.

A central theoretical contribution of IAT is its articulation of a hierarchical calibration architecture comprising three nested layers: mood (fast, labile), personality (intermediate, dispositional), and morality (slow, constitutive); each operating at distinct timescales and with distinct degrees of plasticity. The theory further introduces a typology of self-disruption and repair dynamics, characterizing micro-, meso-, and macro-perturbations and the corresponding mechanisms of identity reconsolidation. Pathological attractor configurations (rigidity, fragility, and false fixation) are mapped onto recognized clinical phenomena with direct treatment implications. IAT is positioned as compatible with Friston’s free energy principle while extending it through explicit multi-layer calibration and disruption-repair typology. Implications for clinical psychology, philosophy of personal identity, and empirical neuroscience research programs are discussed.

Keywords: identity, attractor dynamics, self-model, predictive processing, default mode network, interoception, agency, calibration

1. Introduction

Who are you, and how do you remain that person from one moment to the next? The question sounds almost trivially familiar, yet it conceals one of the deepest puzzles in the sciences of mind and brain. The neurons that compose your cortex today are not the same neurons you were born with; many have died and been replaced by glia, synaptic weights have shifted billions of times, your body has been rebuilt molecule by molecule, your beliefs have evolved, your relationships have transformed, and yet something persists. Something recognizable. Something that can be embarrassed by what it did at twenty-two, proud of what it accomplished at forty, and anxious about what it might become. That something is the self; and its persistence across radical flux constitutes what we shall call the problem of identity continuity.

Scientific and philosophical attempts to solve this problem have proliferated across disciplines. Trait-based accounts in personality psychology propose that the self is constituted by stable dispositions (the Big Five dimensions, for instance) that remain relatively constant across situations and time (Roberts et al., 2006). These accounts have considerable empirical support but tend to treat identity as a static profile, a kind of fixed inventory, and struggle to explain how the self navigates genuine transformation without ceasing to be itself. Narrative accounts, most influentially developed by Ricoeur (1992) and Bruner (1991), hold that identity is constituted through the stories we tell about ourselves; a temporal integration of past, present, and projected future into a coherent arc. These accounts capture something essential about the phenomenology of selfhood but are cognitively top-heavy: they privilege linguistic and reflective processes at the expense of the subcortical, bodily, and largely prereflective dimensions of self-experience. Social constructivist accounts rightly emphasize that identity is co-constructed in relational fields and cultural contexts, but they frequently leave unexplained how a self that is socially constituted nevertheless resists certain social pressures; how one can experience the demand “be someone different” as an attack rather than a simple invitation.

What is missing from all these accounts is a framework that is simultaneously dynamic (treating the self as a process rather than an object), multi-level (encompassing subcortical, cortical, cognitive, and social dimensions), neurologically grounded (mapping onto what is now known about brain organization), and formally tractable (offering the possibility of precise, testable predictions). The present paper proposes that dynamical systems theory, specifically the mathematics of attractor dynamics, provides exactly such a framework.

Core Claim The self is not an object: a fixed entity located somewhere in the brain or in the narrative archive. It is a process with a characteristic signature: a dynamic attractor state in a high-dimensional space of neural, bodily, behavioral, and relational configurations, to which the organism reliably returns after perturbation.

In the vocabulary of dynamical systems theory, an attractor is a region of state-space toward which a system tends to evolve from a wide variety of initial conditions. The self, on this view, is not defined by any particular state (any particular mood, any particular thought, any particular behavior) but by the shape of the basin that draws the system back toward a characteristic region of its state-space. Identity is the topology of that basin. To be yourself is to be in (or to be returning toward) your attractor basin.

This paper develops this proposal into a comprehensive theoretical framework: the Identity Attractor Theory (IAT). The paper proceeds as follows. Section 2 develops the theoretical foundations in dynamical systems theory and positions IAT relative to prior models of the self. Section 3 maps the neural substrates of identity (the default mode network, hemispheric specialization, predictive processing, interoception, and memory binding) onto components of the attractor model. Section 4 addresses agency as the organism’s capacity to act in ways that preserve its attractor configuration. Section 5 presents the central theoretical innovation of IAT: a hierarchical calibration architecture comprising three nested layers (mood, personality, and morality) operating at different timescales. Section 6 analyzes the dynamics of self-disruption and repair. Section 7 offers a formal conceptual model statement. Section 8 discusses theoretical, clinical, and philosophical implications. Section 9 concludes with a research agenda.

2. Theoretical Foundations: Attractor Dynamics and the Self

2.1 Dynamical Systems Theory and Attractors

Dynamical systems theory (DST) is the mathematical study of how systems evolve over time. A dynamical system is described by a set of state variables (quantities that characterize the system’s configuration at any moment) and a set of rules (differential or difference equations) governing how those variables change. The totality of possible configurations of the state variables constitutes the system’s state-space, and the system’s evolution traces a trajectory through that space.

An attractor is a subset of state-space toward which trajectories in a surrounding region (the basin of attraction) converge over time. There are several canonical attractor types. A fixed-point attractor is a single state toward which all nearby trajectories converge: the system settles and stays. A limit cycle is a closed loop in state-space; the system oscillates indefinitely between a set of states. A strange attractor (chaotic attractor) is an intricate fractal structure toward which trajectories converge but on which the system exhibits sensitive dependence on initial conditions; bounded but unpredictable. Each type has biological analogs: the fixed-point attractor resembles homeostatic regulation (body temperature, blood glucose); the limit cycle is exemplified by the cardiac rhythm, the sleep-wake cycle, and many neural oscillations; the strange attractor may characterize the rich, unpredictable variability of moment-to-moment neural activity at rest (Deco et al., 2011).

Definition: Basin of Attraction The basin of attraction of an attractor A is the set of all initial states from which the system will evolve toward A over time. The depth of the basin measures how strongly the system resists perturbation; a deeper basin requires a larger perturbation to exit. The breadth of the basin determines how much variation the system can exhibit while remaining within the attractor’s gravitational field.

Crucially for our purposes, attractors have two properties that make them ideal candidates for modeling identity: resilience and characteristic return. A system in a deep attractor basin will return to its characteristic configuration after perturbation. The perturbation may be large (the system may be displaced far from its resting region) but as long as it remains within the basin, it will find its way back. This is precisely the phenomenology we associate with identity persistence: the self survives illness, grief, failure, and transformation, and recognizes itself afterward.

IAT proposes that identity is a high-dimensional attractor basin in the space of an organism’s behavioral, affective, cognitive, and relational states. We use “high-dimensional” deliberately: the state-space is not a simple two- or three-dimensional landscape but a vastly complex manifold whose axes include neural firing patterns, hormonal states, postural configurations, emotional valences, active beliefs, social roles, and more. The attractor is not a single point but a structured region of this manifold; a region the organism inhabits characteristically and returns to reliably. Perturbations within the basin (everyday stressors, minor surprises) are absorbed by homeostatic mechanisms without lasting structural change. Perturbations that push the system to the edge of the basin (or beyond it) constitute identity-threatening events, and their resolution requires more substantive attractor reconsolidation.

2.2 Prior Models and Their Limitations

IAT does not emerge in isolation; it engages directly with several influential prior frameworks, each of which captures important features of selfhood while leaving others underdetermined.

Narrative identity theory, most fully developed by Paul Ricoeur in his concept of narrative identity (Ricoeur, 1992) and extended by Jerome Bruner (1991) and Marya Schechtman (1996), holds that the self is constituted by the stories a person constructs about their own life; stories that integrate disparate episodes into a coherent trajectory with temporal extension, thematic continuity, and evaluative perspective. This is a genuinely powerful account of how human beings experience identity, and IAT does not dismiss it. However, narrative theory has a fundamental limitation: it operates almost exclusively at the level of explicit, linguistically mediated self-representation. It cannot easily accommodate the fact that identity has subcortical, visceral, and embodied dimensions that operate below narrative access; the sense of self that persists through dreamless sleep, the bodily identity that grounds an amnesiac’s social functioning, the pre-narrative identity of the infant. Narrative is, on the IAT account, the left-hemisphere articulation of attractor geometry; important, but not constitutive.

Metzinger’s self-model theory of subjectivity (Metzinger, 2003) offers a phenomenologically rigorous account of the self as a transparent self-model; a representational construct that the organism mistakes for an entity rather than recognizing as a model. This account achieves impressive precision in describing the phenomenal structure of self-experience and accounts elegantly for dissociative phenomena. Its limitation, from the IAT perspective, is that it remains underdetermined at the dynamical level: it does not provide a model of how the self-model is maintained over time, what makes it stable or unstable, or how it recovers from disruption. The concept of a “self-model” is not the same as the concept of a dynamic attractor maintained by specific neural and agentic mechanisms.

Friston’s free energy principle (FEP) and self-evidencing (Friston, 2010; Hohwy, 2013) are perhaps the closest kin to IAT among existing frameworks. The FEP proposes that biological agents minimize the free energy of their sensory states; equivalently, they minimize surprise by acting on the world to confirm their generative models and by updating those models in the light of prediction error. The concept of self-evidencing captures the way organisms act to confirm their model of what-kind-of-thing-they-are. This maps directly onto the attractor metaphor: minimizing free energy is equivalent to returning to the attractor center. IAT extends the FEP framework by adding (i) a multi-layer calibration architecture that distinguishes mood, personality, and morality as operating at different timescales and with different degrees of plasticity; (ii) an explicit disruption-repair typology; and (iii) a more direct integration of clinical and phenomenological evidence.

Social identity and relational self theories (Tajfel & Turner, 1979; Cross & Markus, 1991) make the important point that the self is not a purely intrapsychic phenomenon but is constituted in part by group membership, relational histories, and social mirroring. IAT incorporates this insight by treating relational inputs as significant sources of attractor-shaping information; the social field is one of the high-dimensional axes of the identity state-space. However, social theories are insufficient as complete accounts of identity persistence because they cannot explain how the self resists social pressure; how it experiences certain social demands as attacks rather than as instructions. The attractor model provides the missing mechanism: social demands that would push the system outside its basin are experienced as aversive and are resisted by the return function.

IAT’s synthesis is thus multi-level (subcortical to narrative), embodied (interoceptive and somatic as well as cortical), neurologically grounded (mapped onto specific neural circuits), and dynamical (formalized in attractor geometry). It is compatible with the preceding accounts while correcting their characteristic blind spots.

3. The Neural Substrate of the Identity Attractor

3.1 The Default Mode Network as the Attractor’s Core

The default mode network (DMN) is a set of interconnected cortical regions that exhibits robust activation during rest (when the brain is not engaged in externally directed tasks) and reliable deactivation during demanding cognitive tasks (Raichle et al., 2001; Raichle, 2015). Its core nodes include the medial prefrontal cortex (mPFC), the posterior cingulate cortex (PCC) and precuneus, the angular gyrus and lateral temporal cortex, and the hippocampal formation and parahippocampal gyrus. This network was initially described as a “task-negative” network and dismissed as a kind of neural idling. Subsequent research has revealed it to be far more: a network specifically dedicated to self-referential processing, autobiographical memory retrieval, mental time travel (imagining past and future), social cognition (modeling other minds), and default simulation (the ongoing generation of predictive scenarios about the self and its world) (Buckner et al., 2008; Andrews-Hanna et al., 2010).

IAT Interpretation: DMN as Basin Geometry Maintenance In IAT terms, the DMN is the neural implementation of the identity attractor’s resting-state maintenance function. During periods of non-directed cognition, the DMN continuously runs predictive simulations of the self in possible contexts; maintaining the attractor’s shape in the absence of external demands. Deactivation of the DMN during external task engagement reflects the attractor “releasing” its hold temporarily to allocate resources to environmental processing, then reasserting itself during the next idle period.

The mPFC is particularly critical: it is the site of self-relevant evaluation, the integration of self-referential information with emotional valence, and the generation of predictions about one’s own future states (Northoff et al., 2006). The PCC and precuneus are involved in the continuous monitoring of the self’s state relative to its environment; a kind of internal compass. The hippocampal formation links current experience to autobiographical history, ensuring that DMN simulations are anchored in actual memory rather than pure confabulation.

IAT predicts that individual differences in DMN topology (its connectivity strength, its hubness, its coupling with the hippocampus and mPFC) should correlate with individual differences in identity stability. Persons with more coherent, strongly connected DMNs should exhibit greater identity resilience across life transitions. This is an empirically testable prediction; longitudinal neuroimaging studies tracking DMN structure across major life transitions (bereavement, migration, religious conversion) represent a key proposed research program.

3.2 Hemispheric Specialization and Identity Processing

One of the most theoretically productive resources for understanding the neural architecture of identity is the research on hemispheric specialization in self-related processing. The left and right cerebral hemispheres, while extensively interconnected, exhibit systematic differences in their contributions to self-representation that have profound implications for understanding attractor maintenance.

The left hemisphere is the dominant hemisphere for language, sequential logical processing, categorical classification, and explicit narrative construction. Its contribution to identity is the generation of the explicit self-story: the verbally articulated, categorically structured account of who one is; “I am a physician; I value scientific rigor; I am the eldest of three children; I am politically progressive.” This is the hemisphere that can answer “Who are you?” in complete sentences, that generates the autobiographical narrative, that maintains categorical self-labels and uses them to filter incoming information. The left hemisphere is, in Iain McGilchrist’s (2009) striking terminology, the Emissary: precise, articulate, categorical, and confident, but operating on a simplified map of the territory rather than the territory itself.

The right hemisphere, by contrast, is dominant for holistic and contextual processing, embodied self-representation, emotional tone and affective nuance, spatial self-location, recognition of incongruity, and the integration of the self with its relational and environmental context. Its contribution to identity is the maintenance of the broader identity gestalt; the felt sense of being a self of a certain kind, in a certain relational field, with a certain emotional coloring. Critically, the right hemisphere appears to be particularly specialized for detecting discrepancy; for registering when the organism’s current state deviates from what is expected or familiar. This makes it, in IAT terms, the primary deviation-detection mechanism of the attractor system.

Definition: Dual-Channel Attractor Monitoring IAT proposes that the two cerebral hemispheres implement complementary attractor maintenance functions. The left hemisphere is the “narrator” (generating the explicit linguistic description of the attractor’s center (the identity narrative). The right hemisphere is the “sensor” (detecting deviations from the attractor and generating the affective signal of deviation (the feeling of “not being oneself”). Together they constitute a dual-channel monitoring architecture for basin maintenance.

Clinical evidence supports this architecture. Left hemisphere stroke patients frequently lose the capacity to generate verbal self-narrative but may retain a visceral sense of self-continuity; they cannot tell you who they are but they act in recognizably continuous ways. Right hemisphere strokes, by contrast, can produce the remarkable phenomenon of anosognosia; the inability to recognize one’s own deficits or changes, including dramatic alterations in personality and behavior (Ramachandran & Blakeslee, 1998). This is precisely what IAT would predict if the right hemisphere is the primary deviation-detection mechanism: damage to it removes the system’s ability to register that it has departed from its attractor, producing a subjective experience of seamless continuity that is objectively false.

McGilchrist’s master-emissary framework offers an extended philosophical analysis of these dynamics, arguing that the dominant culture of modernity has progressively privileged left-hemisphere processing at the expense of right-hemisphere integration; a pathology of identity at the civilizational scale, if IAT’s mapping is correct.

3.3 Predictive Processing and Identity as a Generative Model

The predictive processing (PP) framework, most comprehensively developed by Karl Friston (2010) and extended by Andy Clark (2016) and Anil Seth (2021), proposes that the brain is fundamentally a hierarchical generative model; a system that generates predictions about its sensory inputs at multiple levels of abstraction and continuously updates those predictions in the light of prediction errors (mismatches between expected and actual inputs). Perception, on this account, is not the passive reception of sensory data but the active construction of a best-guess interpretation of the causes of sensory signals, constrained by the generative model’s predictions. Action is the process of changing the sensory data to fit the model; rather than updating the model to fit the data.

Key Concept: Active Inference In active inference, the organism acts to minimize prediction error not only by updating its beliefs (perceptual inference) but also by acting on the world to make it conform to its prior predictions (active inference proper). Identity-maintaining behavior is, on this account, active inference under self-related priors: the organism acts to keep the world (and its own states) consistent with its self-model.

The integration of PP with IAT is direct and productive. In PP terms, the identity is the brain’s highest-level, most persistent, and most influential set of priors; the self-model is a prior distribution over expected perceptual, affective, social, and bodily states that propagates downward through the processing hierarchy, shaping everything from low-level sensory processing to high-level social judgment. When the organism’s current state is consistent with its self-model, prediction errors are low and the sense of identity is untroubled. When current states generate high prediction errors relative to the self-model (when experience is markedly identity-inconsistent) the system faces a choice: assimilate the error (update the prior, accept a small degree of identity change) or suppress the error (discount the discrepant information, maintain the prior at the cost of accuracy).

This maps elegantly onto the attractor model. The attractor basin in PP terms corresponds to the region of state-space in which prediction errors are low relative to the self-model’s priors. The depth of the basin corresponds to the precision assigned to the self-model’s priors; a deep, narrow basin corresponds to high-precision self-model priors, where prediction errors are heavily suppressed (a rigid self-model with low error tolerance). A broad, shallow basin corresponds to low-precision self-model priors; a flexible, open identity that can assimilate substantial discrepant information without existential distress (Clark, 2016; Seth, 2021).

Friston’s concept of self-evidencing (the organism’s continuous gathering of evidence to confirm its model of what-kind-of-thing-it-is) maps directly onto the attractor’s return function: self-evidencing behavior is the mechanism by which the organism navigates back toward its basin center. Hohwy (2013) has provided a rigorous philosophical analysis of the PP framework’s implications for the self, emphasizing that the self that does the predicting is itself a construct of the generative model; a point entirely consistent with IAT’s process-based ontology.

3.4 Interoception: The Body as Attractor Anchor

If the DMN and predictive processing provide the cortical architecture of the identity attractor, interoception provides its subcortical anchor; the most fundamental and high-fidelity signal of self-continuity available to the organism. Interoception refers to the sensory processing of the body’s internal states: the beating of the heart, visceral pressure and distension, respiratory dynamics, the thermal and chemical environment of the body’s interior, pain and itch, hunger and satiety, arousal and calm (Craig, 2009).

A.D. Craig’s landmark research on the insular cortex has established the posterior and mid-insula as the primary cortical recipients of interoceptive signals, with the anterior insula integrating these signals with emotional and cognitive information to produce the subjective sense of the “material me”; the felt sense of inhabiting a body of a specific kind, in a specific condition, right now (Craig, 2009). This felt bodily sense is not peripheral to identity; it is, on Craig’s account and on the IAT account, the most basic layer of self-representation; the one that is continuous through states in which higher-level narrative and cognitive identity is disrupted.

Antonio Damasio’s somatic marker hypothesis (Damasio, 1994) provides a complementary account. On Damasio’s view, the body does not merely provide raw physiological data; it generates somatic markers (affective tags attached to representations of situations, outcomes, and social signals) that function as rapid, pre-deliberative evaluations of relevance to the organism’s wellbeing. These somatic markers constitute a biological record of the organism’s evaluative history, a body-memory of what has mattered and how. IAT integrates this insight by treating somatic markers as part of the attractor’s information archive; the body carries the identity in its affective responses, even when the cognitive narrative is disrupted.

Definition: Interoceptive Anchoring IAT proposes that interoceptive signals are the lowest-level, highest-fidelity anchoring mechanism of the identity attractor. The body maintains attractor continuity even when narrative identity is disrupted; as in amnesia, intoxication, acute stress, or dissociation. Chronic mismatch between expected and actual interoceptive signals (as in chronic illness, trauma, or severe dissociation) constitutes a low-level attractor erosion that can destabilize higher calibration layers.

Critically, interoceptive processing is predictive: the brain generates predictions about expected bodily states and processes interoceptive signals as prediction errors relative to those predictions (Seth, 2013; Critchley & Garfinkel, 2017). Interoceptive prediction error (the mismatch between expected and actual bodily states) is a powerful destabilizing signal for the identity attractor. When it is chronically elevated, as in conditions of persistent physiological dysregulation (trauma, chronic pain, autonomic dysregulation), the body-level anchor of the attractor is eroded, contributing to the phenomenology of dissociation, depersonalization, and “not feeling real” that characterizes severe identity disruption.

3.5 Memory Binding and Temporal Continuity

The persistence of identity across time requires more than the maintenance of a characteristic attractor configuration at any given moment; it requires the binding of present experience to the remembered past in a way that preserves attractor-relevant information across temporal gaps. This function is performed, critically, by the hippocampus and its interactions with the mPFC.

The hippocampus is the brain’s primary indexing mechanism for episodic memory; the record of specific autobiographical events with their temporal, spatial, and emotional context (Schacter et al., 2012). Each episode is encoded as a pattern that references back to the self-schema active at the time of encoding. When such memories are later retrieved, they do not merely reproduce the past episode; they reactivate it in the context of the current self-model, allowing comparison and updating. Memory retrieval in the hippocampus is not a playback but a reconstruction; one that is systematically biased toward consistency with the current identity (Conway & Pleydell-Pearce, 2000).

The process of memory reconsolidation (the observation that retrieved memories become temporarily labile before being re-stabilized) is of particular theoretical significance for IAT (Moscovitch et al., 2016). It means that the neural mechanism of identity revision is literally built into the architecture of episodic memory: every act of autobiographical recall is an opportunity for the attractor to be updated. This is not a bug but a feature: it is the mechanism by which the identity attractor incorporates new experience without losing its characteristic shape.

Key Concept: Pattern Completion vs. Pattern Separation The hippocampus mediates two complementary functions relevant to identity maintenance. Pattern completion: the reconstruction of a stored memory pattern from a partial cue; is the neural mechanism of identity recognition (“This is who I am; I have been here before”). Pattern separation: the encoding of similar experiences as distinct representations; is the neural mechanism of registering genuine novelty and change. The balance between these two functions determines how conservatively or liberally the attractor updates in response to new experience.

Schema theory, in its modern neuroscientific formulation by van Kesteren and colleagues (van Kesteren et al., 2012), provides a complementary account: new information is preferentially processed and integrated when it is compatible with existing cognitive schemas (including self-schemas), while highly discrepant information triggers more extensive encoding processes. This gives the self-schema (the identity attractor’s cognitive representation) a privileged role in determining what gets remembered, how it gets remembered, and how it modifies the attractor upon reconsolidation.

The temporal continuity of identity is thus, on the IAT account, a hippocampal-mPFC dialogue: the hippocampus provides episodic anchors (specific, temporally located memories that tie the present self to its history) while the mPFC integrates these anchors into the ongoing self-model, ensuring that the attractor does not drift randomly but maintains its characteristic shape across time (Moscovitch et al., 2016).

4. Agency and the Agentic Layer of Identity

4.1 Agency as Attractor-Maintaining Behavior

The neural mechanisms described in Section 3 are primarily reactive; they describe how the system returns to its attractor in response to perturbation. But identity is not merely maintained by passive homeostatic processes; it is actively sustained through the organism’s behavior in the world. This is the domain of agency, and it constitutes a distinct layer of the IAT architecture.

In IAT terms, agency is defined as the capacity of a system to act in ways that preserve or restore its attractor state. This definition distinguishes agency from mere reactivity: a thermostat that adjusts temperature in response to deviation is reactive; an agent who, upon sensing that a social environment is pulling them away from their values, actively seeks different environments, reasserts their commitments, or works to change the environment is exercising genuine agency. Agency, on this account, is forward-looking and selective: the agent anticipates deviation, selects from a repertoire of actions those most likely to maintain attractor position, and executes those actions in the face of competing demands.

Definition: Sense of Agency (SoA) The sense of agency is the phenomenal feeling of being the author of one’s own actions; the experience that “I am doing this, from my own center.” In IAT terms, high SoA corresponds to actions that fall within the attractor basin; actions that are identity-consistent and experienced as expressions of the self. Low SoA (alienation, compulsion, coercion, “acting out of character”) corresponds to being moved outside the basin; to actions that diverge from the identity attractor’s characteristic region.

The neural mechanisms of the sense of agency are well characterized (Haggard, 2017). The brain generates predictions about the expected sensory consequences of self-initiated actions; an efference copy of the motor command is used to generate a predicted sensory feedback signal. When actual feedback matches the prediction, the action is attributed to the self (agency is felt); when feedback diverges substantially from the prediction, the action may be experienced as externally caused or alien; a mechanism that Frith (1992) proposed as the basis of passivity phenomena in schizophrenia. IAT interprets this mechanism as the neural implementation of attractor-consistency checking: predicted sensory consequences of self-generated, identity-consistent actions are subtracted from the incoming signal, producing low residual error and high SoA. Identity-inconsistent actions (forced, coerced, or otherwise incongruent) generate high residual error and low SoA (Gallagher, 2000).

Agency thus has a bidirectional relationship with the attractor: agency is the expression of the attractor (acting from one’s identity center generates high SoA and maintains the attractor) and also its maintenance mechanism (choosing identity-consistent actions over time deepens and stabilizes the basin).

4.2 Coherence vs. Continuity

A central conceptual distinction in IAT concerns two dimensions of identity persistence that are often conflated but are theoretically and clinically separable.

Continuity is the temporal dimension of identity persistence: the thread of memory and narrative connecting past and present selves into a recognizable arc. Continuity answers the question: “Is this the same person as the one who existed ten years ago?” It is primarily hippocampal-dependent, grounded in autobiographical memory binding, and articulated through the narrative resources of the left hemisphere. Continuity can be disrupted by amnesia, radical environmental change, or experiences so transformative that the prior self seems to belong to a different person.

Coherence is the synchronic dimension of identity persistence: the internal consistency of values, dispositions, and responses at any given moment. Coherence answers the question: “Does this person’s behavior, at this moment, constitute an integrated, mutually consistent expression of a unified self?” It is more dependent on the midline structures of the DMN (the mPFC and PCC) and on the prefrontal orchestration of competing motivational systems. Coherence can be disrupted by severe internal conflict, post-traumatic fragmentation, or states of dissociation in which different self-states fail to communicate.

Theoretical Insight: Coherence-Continuity Dissociation Continuity and coherence can dissociate in instructive ways. A person may retain rich autobiographical continuity while exhibiting severe incoherence; as in post-traumatic fragmentation, where past and present selves are connected by narrative but the current self-state is internally conflicted and contradictory. Conversely, certain amnesiac conditions or radical identity transformations (cult exit, religious conversion) may produce high coherence in the present self while severing continuity with the remembered past. IAT predicts that these dissociations have distinct neural signatures: continuity disruption is primarily hippocampal; coherence disruption is primarily mPFC/PCC.

The coherence-continuity distinction has productive implications for understanding identity growth. A degree of temporary incoherence (the willingness to hold apparently contradictory values, to not yet have resolved the tension between who one was and who one is becoming) is not a sign of identity pathology but of identity development. The creative tension between continuity and coherence, in IAT terms, is the zone in which the attractor basin is capable of expanding. The system tolerates a wider range of states temporarily in order to integrate new experience into a more capacious basin.

5. Calibration Layers: The Hierarchical Architecture of Identity

The central theoretical innovation of IAT is its articulation of a hierarchical calibration architecture; a nested set of layers that modulate the identity attractor at different timescales and with different degrees of plasticity. Identity is not monolithic; it is organized into layers that operate semi-independently, interact bidirectionally, and respond to different categories of perturbation. We propose three primary calibration layers: mood, personality, and morality.

Figure 1: Conceptual Diagram: The Identity Attractor Landscape This figure would depict the identity state-space as a three-dimensional energy landscape, analogous to a topographic map with a central attractor basin. The basin has three concentric zones representing the calibration layers: the innermost zone (deepest, steepest walls) represents the moral layer; the constitutional constraints on permissible attractor configurations; the middle zone represents the personality layer: the structural parameters of the basin’s shape, depth, and orientation; the outermost zone, closest to the basin rim, represents the mood layer: the dynamic gain modulation that temporarily shifts the basin’s effective center without restructuring it. The basin center (μ_I, the identity setpoint) is marked. Multiple trajectories are shown: micro-disruption trajectories that oscillate within the outermost zone before returning to center; meso-disruption arcs that reach the personality layer before returning; and macro-disruption trajectories that exit the basin entirely. Repair arcs show return trajectories converging on a new basin center (μ’_I), slightly displaced from the original, representing post-disruption identity reconsolidation. The surrounding landscape shows neighboring basins representing alternative identity configurations (pathological attractors: rigid basin with steep, narrow walls; fragile basin with shallow walls and poorly defined center; false fixation basin representing a local minimum distant from the organism’s authentic attractor center).

5.1 Layer 1: Mood (Fast, Labile, Reversible)

Mood is the most rapid and reversible calibration layer of the identity attractor; a fast-acting modulation of the system’s operating parameters that adjusts its moment-to-moment configuration without altering the underlying basin structure. A person in a depressed mood is still recognizably themselves; their characteristic values, traits, and commitments are not erased; but the system operates differently, perceives differently, and behaves differently. Mood is the attractor’s dynamic gain control.

In neuroscientific terms, mood states are implemented through the interaction of subcortical affective systems (the amygdala, the ventral tegmental area (VTA), the nucleus accumbens, and the hypothalamic-pituitary-adrenal (HPA) axis) with the prefrontal systems that regulate them (Phillips et al., 2003). These systems modulate perceptual sensitivity (a depressed mood lowers the signal-to-noise ratio for positive information), action thresholds (elevated mood lowers the threshold for approach behavior; depression raises it), and social engagement (mood affects the precision of interoceptive and social cues).

IAT proposes that affective valence shifts the identity attractor’s effective position temporarily without changing its structural parameters. Depression, in dynamical systems terms, flattens the basin; it reduces the return force, making it harder for the system to navigate back to its characteristic center. The depressed person reports feeling “not like themselves”; this is the phenomenological signature of reduced basin depth. Mania, conversely, may shift the basin center to a region of heightened activation and approach motivation, producing behavior that feels authentic and energized but is actually an artifact of altered gain rather than genuine attractor change.

Definition: Mood Layer Mood is the dynamic gain control of the identity attractor; the fastest-acting calibration layer, operating on timescales of minutes to weeks, implemented through subcortical affective systems (amygdala, VTA, nucleus accumbens), and modulating the effective position and depth of the attractor without restructuring its underlying geometry. Mood disorders correspond to chronic dysregulation of this gain control mechanism.

The therapeutic implication is important: effective mood interventions (pharmacological, behavioral, or psychological) operate by restoring normal gain control; returning the basin to its characteristic depth and allowing the return function to operate normally. They do not, in themselves, restructure the basin. This is why mood treatment alone, without personality-level or morality-level work, often results in symptom remission without fundamental identity change.

5.2 Layer 2: Personality (Slow, Dispositional, Moderately Stable)

Personality is the structural layer of the identity attractor; the set of parameters that determine the basin’s depth, shape, orientation, and coupling to the social environment. Whereas mood is a transient modulation of operating parameters, personality is the basin geometry itself, operating at timescales of months to years and resistant to transient perturbation.

IAT proposes a mapping between the five-factor model of personality (the “Big Five”: Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism; see McCrae & Costa, 1999) and the structural parameters of the attractor basin. This mapping is not arbitrary but reflects the functional logic of each trait domain:

Big Five Dimension	IAT Attractor Parameter	Basin Interpretation
Openness to Experience	Basin breadth	How wide a range of states can be assimilated without destabilization; high Openness = broad basin
Conscientiousness	Basin depth	Resistance to disruption; capacity to maintain attractor under load; high Conscientiousness = deep basin
Neuroticism	Frequency of excursions toward basin edge	High Neuroticism = frequent near-exits from basin; high emotional reactivity to prediction errors
Agreeableness	Social attractor coupling	Degree to which the identity attractor is coupled to, and shaped by, the attractors of significant others
Extraversion	Energetic return force	The strength of the approach drive that moves the system back toward social and environmental engagement after withdrawal; high Extraversion = strong return force toward social stimulation

Personality change over the lifespan (the now well-documented tendency for adults to become more conscientious, agreeable, and emotionally stable with age (Roberts et al., 2006)) is interpreted in IAT terms as a slow drift in basin parameters: progressive deepening (greater Conscientiousness), reduced edge-excursion frequency (lower Neuroticism), and expanded breadth (greater Openness, in some trajectories). These changes are driven by accumulated experience, sustained relational engagement, and the progressive integration of self-discrepant information over developmental time.

The neural correlates of personality dimensions map onto this architecture: frontostriatal connectivity is associated with Conscientiousness (reflecting the capacity for top-down attractor maintenance); amygdala reactivity predicts Neuroticism (reflecting the frequency of error signals that push the system toward basin edges); dopaminergic systems in the ventral striatum contribute to Extraversion (the energetic approach drive) (DeYoung et al., 2010).

5.3 Layer 3: Morality (Slow, Deep, Highest Resistance to Change)

The deepest and most resistant calibration layer in IAT is morality: the set of constitutive moral commitments (values, principles, and prohibitions) that define the outer boundary of the identity attractor. Moral commitments are not merely preferences or beliefs that happen to be held with high confidence; they are, on the IAT account, the structural walls of the attractor basin, the configurations that the system will resist most powerfully, because they define the limit of what the organism can do and remain itself.

Theoretical Claim: Morality as Constitutional Layer Morality is the constitutional layer of the identity attractor; the meta-constraint on which identity configurations are permissible. Moral commitments define the outer boundary of the basin: actions that would violate them are experienced as self-betrayal, a visceral signal of deep basin exit. Violations of deep moral commitments feel existential precisely because they are attractor-structural; they do not merely displace the system within the basin but threaten the basin’s integrity itself.

Jonathan Haidt’s (2012) moral foundations theory identifies several candidate attractor anchors that vary across individuals and cultures: care/harm, fairness/reciprocity, loyalty/betrayal, authority/subversion, and purity/degradation. IAT does not presuppose any particular moral foundation as universal; instead, it proposes that each individual’s and each culture’s distinctive configuration of moral foundations constitutes the distinctive outer boundary of their identity attractor. What functions as a deep attractor wall for one person (purity norms, in certain cultural contexts) may be a superficial constraint for another.

The phenomenology of moral injury (the profound sense of irreversible self-change experienced when one crosses, or is forced to cross, one’s fundamental moral boundary) is one of the strongest pieces of phenomenological evidence for the IAT account (Litz et al., 2009). The person who has committed an act they regard as fundamentally wrong does not merely feel bad; they feel that they are no longer the same person; that a qualitative threshold has been crossed. This is precisely the signature of attractor basin exit: the system is no longer in the region to which its return function naturally converges.

Moral development, on the accounts of Kohlberg (1981) and Gilligan (1982), represents the progressive elaboration and deepening of this constitutive layer; not merely the acquisition of more rules but the integration of moral principles into the core geometry of identity, such that they function with increasing automaticity, generality, and self-consistency. IAT frames this as progressive basin elaboration at the deepest level: not mere rule-following but the gradual incorporation of principled commitments into the attractor’s structural walls, where they generate strong return forces and resist revision.

5.4 Cross-Layer Dynamics and Interactions

The three calibration layers are not independent; they interact bidirectionally in ways that have significant theoretical and clinical implications. These interactions constitute some of IAT’s most specific empirical predictions.

The most clinically significant cross-layer interaction is the downward erosion pathway: sustained dysregulation at a faster layer can erode the integrity of slower layers. Chronic mood disorder (particularly sustained depression) does not merely flatten the basin temporarily; over years, it can erode personality-layer parameters, reducing Conscientiousness, increasing Neuroticism, and flattening Openness. Decades of severe depression can, in extreme cases, produce apparent changes in moral sensibility; not because the constitutive moral layer has been directly attacked, but because the sustained degradation of gain control mechanisms has reduced the system’s capacity to act from its moral commitments, creating a functional erosion of the moral layer’s effective influence (even if the underlying commitments formally persist).

The complementary upward anchoring pathway is equally important: deep moral commitments can override personality-layer tendencies and even mood-layer dysregulation. A person with high Neuroticism (frequent basin-edge excursions) who has integrated a deep commitment to care and responsibility may nonetheless function effectively under stress, because their moral commitments provide a powerful return force that compensates for weakened personality-level regulation. This is the neural basis of what Viktor Frankl (1959) described as finding meaning as a survival mechanism.

Therapeutic implications follow directly. Effective interventions must be targeted to the appropriate layer: cognitive-behavioral therapy (CBT) operates primarily at the mood and personality layers, providing tools for gain-control normalization and basin-shape modification. Psychedelic-assisted therapy (as in psilocybin trials for treatment-resistant depression) may temporarily dissolve the gain-control function of all three layers, producing a period of radical openness (a “no-basin” state) before allowing reconsolidation at a healthier basin position (Carhart-Harris et al., 2017). Trauma therapy must address layer-spanning damage: trauma disrupts the interoceptive anchor (body-level), the mood layer (affective dysregulation), and the personality layer (altered trait expression), and in cases of moral injury, the constitutive layer itself.

6. The Dynamics of Self-Disruption and Repair

6.1 Disruption: Perturbations and Basin Exit

IAT provides a typology of identity disruptions organized by scale; specifically, by how deeply into the attractor basin the perturbation penetrates and, correspondingly, by which calibration layers are affected.

Micro-disruptions are everyday perturbations: a frustrating conversation, a minor failure, an unexpected social demand, a period of sleeplessness. These events generate prediction errors and momentarily shift the system’s position within the attractor basin, but the return force is sufficient to restore equilibrium quickly, typically within hours. The mood layer absorbs the perturbation; no lasting structural change occurs. Most of daily life consists of micro-disruption management, and the capacity to handle micro-disruptions efficiently is itself a component of psychological well-being.

Meso-disruptions are more significant perturbations: meaningful loss, professional failure, serious illness, relational rupture, major disappointment of a core aspiration. These events reach the personality layer; they require not merely mood regulation but a degree of self-concept revision, a renegotiation of expectations and self-understanding. The system may experience a period of genuine destabilization, during which the return function is weakened and the person feels “not like themselves.” But with sufficient time, social support, and meaning-making resources, the basin is restored; typically in a slightly modified form that accommodates the new experience.

Macro-disruptions are basin-level events: profound trauma, near-death experience, cataclysmic shame, psychedelic dissolution, radical religious conversion, and exile. These events push the system beyond the basin’s boundaries; the identity attractor as previously configured is no longer available as a return point. The system must find or construct a new viable attractor position. This process is not automatic and may require years and significant support. It may result in a fundamentally different self (not a restored self) which is why survivors of such events so often speak of becoming “a different person.”

Clinical Note: Dissociation as Emergency Basin Preservation When perturbation exceeds the system’s real-time processing capacity (when the distress signal is so intense that integration is impossible in the moment) the system may temporarily “flatten” the attractor as an emergency survival response. Dissociation, in IAT terms, is protective depotentiation of the attractor: the basin is temporarily made shallow and its return force reduced, preventing the catastrophic consequences of being violently expelled from the basin while the threat persists. The cost of this protection is paid later: the flattening persists, contributing to identity fragmentation and the clinical features of trauma.

The neurological signature of severe disruption involves a cascade of interacting systems: the prediction error signal propagates upward through the hierarchy, engaging the salience network (anterior insula, anterior cingulate cortex) in conflict monitoring; the DMN-salience network conflict produces the characteristic subjective sense of unreality and threat; interoceptive noise escalates (elevated heart rate, visceral distress, changed breathing pattern); and hippocampal pattern separation may fail; the system cannot encode the disruptive event as clearly distinct from its prior self-understanding, producing the confused, undifferentiated quality of traumatic memory (van der Kolk, 2014).

6.2 Repair: Return Dynamics and Identity Reconsolidation

Repair, in IAT terms, is not mere return to a prior state; that is only possible for micro-disruptions. For meso- and macro-disruptions, repair is the re-establishment of a viable attractor position, which may be shifted from the prior baseline and which represents a genuinely new self-organization rather than the restoration of an old one. The goal of repair is not the recovery of the prior self but the establishment of a coherent, resilient, and authentic new basin.

IAT identifies four primary mechanisms of identity repair, each corresponding to a different level of the neural and calibration architecture:

Narrative integration is the bringing of disrupted experience into the autobiographical timeline; finding language for what happened, locating it in the story of one’s life, and integrating it as an event that has a place in the ongoing arc rather than an alien intrusion. This is primarily a left-hemisphere operation, engaging the narrative resources of language and sequential logic, and is implemented neurally through hippocampal reconsolidation; the retrieval and re-stabilization of traumatic memory in the context of an updated self-model (Schacter et al., 2012). Narrative exposure therapy and testimony-based trauma therapies operate through this mechanism.

Social mirroring recruits the social field as an external attractor reference. When one’s own return function is weakened (when the internal signal of “who I am” is unclear or absent) the presence of trusted others who know and recognize one’s identity provides an external scaffolding of the attractor. Co-regulation with a trusted other literally lends that person’s nervous system as a stabilizing influence: this is the mechanism behind the powerful restorative effects of secure attachment, therapeutic relationships, and community belonging. The neurobiological substrate involves the oxytocinergic system, the autonomic nervous system’s ventral vagal complex (Porges, 2011), and the social circuits of the mPFC and temporoparietal junction.

Somatic grounding targets the interoceptive anchor of the attractor; the body-level foundation that, when restored to baseline, provides the most immediate signal of self-continuity. Breathing practices, body-based therapies (yoga, somatic experiencing, EMDR), physical exercise, and reliable physical routine all operate by restoring interoceptive predictability; reducing interoceptive prediction error and restabilizing the lowest-level layer of the attractor system. The clinical insight that trauma lives “in the body” (van der Kolk, 2014) is consistent with IAT’s account: when the interoceptive anchor is disrupted, higher-layer restoration efforts are built on unstable ground.

Meaning-making is the generation of a coherent meta-narrative that accommodates the disruption as part of the identity’s trajectory; not merely recording what happened but integrating it as a contribution to who one is becoming. In IAT terms, meaning-making is the process by which a macro-disruption is reconsolidated not merely as a scar (a basin-deforming event) but as a constitutive element of an expanded basin. Post-traumatic growth (the empirically documented phenomenon of positive psychological change following severe adversity (Tedeschi & Calhoun, 2004)) is interpreted in IAT terms as the successful expansion of the attractor basin: the self becomes capable of holding more experience, more complexity, more suffering, and more joy without destabilizing, because the basin has been expanded by the very disruption that threatened it.

6.3 Pathological Attractors: Rigidity, Fragility, and False Fixation

Not all attractor configurations are healthy. IAT identifies three primary pathological variants, each representing a characteristic distortion of attractor geometry.

Rigid attractors are characterized by excessive basin depth and narrowness; a very deep, steep-walled basin that resists integration of any discrepant information. Phenomenologically, the rigid attractor corresponds to the experience of an identity that feels permanently settled and invulnerable; which is not the same as genuine resilience. The system handles discrepant information by chronically suppressing prediction error; by dismissing, distorting, or devaluing information that does not fit the self-model. This is the attractor structure underlying pathological narcissism (grandiose self-model with aggressive prediction-error suppression), rigid ideological identity (information inconsistent with the ideology is systematically rejected), and certain defensive personality structures. The clinical challenge with rigid attractors is that the system will actively resist therapeutic disruption; any attempt to introduce identity-discrepant information will be experienced as an attack and met with defensive counter-pressure.

Fragile attractors are the opposite: shallow basins with poorly defined centers and weak return forces. Small perturbations produce large identity oscillations; the person feels profoundly unsettled by minor criticism, role changes, or relational disruptions because their attractor offers little restoring force. This pattern corresponds to the clinical features of identity diffusion, borderline personality organization, and certain presentations of complex trauma; the absence of a stable center from which to engage with the world. The therapeutic need is basin deepening: the provision of consistent relational experience, structured cognitive frameworks, and graduated exposure to perturbation in a safe context, all of which progressively build the basin’s depth and coherence.

False fixation is a subtler pathology: an attractor that is stable (the system returns to it reliably) but not authentic; the basin the system inhabits is not the one that represents genuine self-organization, but rather a local minimum imposed by conformity, trauma-accommodation, or chronic social coercion. The phenomenological signature of false fixation is the uncanny experience of alienation or inauthenticity; the persistent sense of “not feeling like myself,” of performing a self rather than being it. IAT predicts that false fixation involves a split between the system’s current attractor and a more deeply configured but suppressed attractor; one that represents the organism’s authentic self-organization but has been overlaid or occluded by adaptation to a constraining environment.

“The most common form of despair is not being who you are.” – Søren Kierkegaard (cited in Dreyfus & Kelly, 2011, p. 43)

False fixation is particularly relevant to clinical presentations of long-standing suppression of identity; contexts in which the organism has organized around a self-concept imposed or demanded by an environment that was not safe for authentic self-expression. The therapeutic approach is not basin-deepening of the false attractor but excavation; using the tools of reflective exploration, relational attunement, and embodied inquiry to locate the authentic attractor beneath the false one and create conditions for the system to migrate toward it.

7. Integration: A Formal Model Statement

We now offer a formal conceptual statement of the Identity Attractor Theory, integrating the neural, agentic, and calibration components developed in Sections 3–6 into a unified model. This statement is conceptual rather than mathematically derived; it is intended to provide a precise and internally consistent framework that can guide empirical operationalization and computational modeling.

7.1 The Identity State-Space

Let S denote the identity state-space; the full set of possible configurations of an organism’s neural, bodily, behavioral, affective, cognitive, and relational states. S is a high-dimensional continuous space; each point s ∈ S represents a specific, momentary total configuration of the organism across all relevant dimensions. The organism’s trajectory through S over time is constrained by its biology, its history, and its environment, but is not fully determined by any of these: it retains degrees of freedom that constitute the space of possible selfhood.

Identity State-Space: S = { s = (s_neural, s_bodily, s_affective, s_cognitive, s_relational) : s_i ∈ ℝ^{n_i} } where each subscript denotes a subspace of dimensions corresponding to neural activity patterns, interoceptive signals, affective valence, cognitive content, and relational configurations respectively.

7.2 The Attractor Basin

The identity attractor A ⊂ S is a structured region of the state-space with the following properties:

• Center: μ_I ∈ S: the identity setpoint, the characteristic configuration that best represents the self at its most coherent and integrated. This is not a state the organism is always in; it is a state the organism tends toward.
• Radius: r > 0: the coherence tolerance, measuring how far the organism can be displaced from μ_I while the return function still operates effectively. A large r corresponds to a broadly tolerant identity; a small r corresponds to a narrow, highly specific self-concept.
• Depth: d > 0: the resilience parameter, measuring the strength of the return force as a function of displacement. A deeper basin (large d) returns more powerfully from perturbation; a shallower basin (small d) offers weaker return force.

Basin Definition: A = { s ∈ S : V(s) ≤ V_threshold } where V(s) is a potential function (analogous to a free energy landscape) that is minimized at μ_I and increases with distance from μ_I. The basin boundary is the set of states where V(s) = V_threshold; crossing this boundary constitutes basin exit (disruption).

7.3 The Return Function

The return function R: S → T(S) maps each state s to a vector in the tangent space of S (a direction of change) that moves the system toward μ_I. R(s) is the composite of all neural and agentic mechanisms that implement attractor-maintenance: the DMN’s resting-state simulation, the predictive processing hierarchy’s self-model maintenance, the interoceptive anchoring, the hippocampal memory consolidation, and the organism’s own agency in choosing identity-consistent environments and behaviors.

Return Dynamics: ds/dt = R(s) + η(t) where η(t) represents exogenous perturbations (environmental events, social demands, physiological changes). When ||s – μ_I|| ≤ r, R(s) successfully returns the system to within the basin. When ||s – μ_I|| > r (basin exit), R(s) may be insufficient; repair mechanisms are required to re-establish convergence.

7.4 Neural System Mapping

Each neural system identified in Section 3 maps onto a specific component of the formal model:

Neural System	Model Component	Function
Default Mode Network (DMN)	Basin geometry maintenance	Continuously simulates μ_I and monitors proximity to it during idle periods; maintains the attractor’s shape
Predictive processing hierarchy	Implementation of R(s)	Generates self-consistent predictions and acts to minimize prediction error relative to self-model; the mechanism of return
Insular cortex / Interoception	Low-level state signal input	Provides the real-time bodily position signal that grounds the state variable s in the actual organism; the body anchor
Hippocampus	Temporal coherence constraint	Binds present s to remembered past states; ensures that μ_I does not drift randomly but maintains autobiographical continuity
Hemispheric specialization (L/R)	Dual-channel monitoring	Left hemisphere: narrates μ_I explicitly; Right hemisphere: detects ||s – μ_I|| and generates affective deviation signal

7.5 Calibration Layer Mapping

Calibration Layer	Model Role	Timescale	Plasticity
Mood	Gain modulation of R(s): adjusts the strength and effective direction of the return function without changing basin geometry	Minutes to weeks	High – rapidly reversible
Personality	Parameters of basin shape: determines d (depth), r (breadth), and the orientation of μ_I in state-space	Months to years	Moderate – changed by sustained experience or significant intervention
Morality	Boundary conditions of A: defines the constitutional constraints on which basin configurations are permissible; the outer wall of the attractor	Years to lifetime	Low – highly resistant; violations experienced as existential

7.6 Disruption and Repair

Disruption condition: ||s(t) – μ_I|| > r (the system has exited the basin) Repair condition: Re-establishment of R(s) convergence to a viable μ’_I, where μ’_I may differ from the original μ_I (the new attractor center after successful integration of the disruptive event). Post-traumatic growth condition: r’ > r and d’ ≥ d at μ’_I — the new basin is broader and at least as deep as the prior basin; the self has expanded without losing resilience.

8. Discussion

8.1 Theoretical Implications

The Identity Attractor Theory represents a genuine theoretical synthesis; not merely the application of an existing framework (dynamical systems theory, predictive processing, or neuroscience of the self) to a new domain, but the integration of multiple frameworks into a model that is more explanatorily powerful than any of its constituent parts. Several theoretical implications warrant explicit articulation.

First, IAT generates specific, testable empirical predictions. The claim that basin depth corresponds to psychological resilience implies that individuals with greater DMN connectivity coherence (as measured by resting-state fMRI metrics of within-DMN functional connectivity) should exhibit greater identity stability across life transitions. This is an empirically tractable prediction: longitudinal studies could track DMN topology in participants undergoing major life transitions (bereavement, retirement, immigration, gender transition, religious conversion) and correlate neuroimaging metrics with validated measures of identity stability and resilience. Similarly, the claim that interoceptive precision is a predictor of identity stability implies that individuals with more accurate interoceptive awareness (as measured by heartbeat detection tasks or related paradigms) should exhibit smaller identity oscillations in response to perturbation; another testable prediction.

Second, IAT clarifies the relationship between narrative identity and dynamical identity in a way that respects the genuine insights of narrative theory while correcting its tendency toward cognitive imperialism. IAT does not dismiss narrative; it assigns it a precise structural role: narrative is the left hemisphere’s articulation of attractor geometry; the explicit verbal description of the basin’s center. This articulation is genuinely important: it provides a shareable, culturally transmissible account of who one is; it facilitates identity-maintenance through social mirroring; and it enables the planning of identity-consistent future behavior. But it is not constitutive: the attractor exists and operates whether or not it is narrated. The amnesiac, the pre-verbal infant, the person in deep meditation who has temporarily dissolved narrative self-reference; all retain identity attractors, even in the absence of the narrative that ordinarily accompanies them.

Third, IAT establishes the relationship with Friston’s free energy principle with precision. IAT is compatible with FEP (minimizing free energy is equivalent to returning to the attractor center) but it adds three things that FEP does not explicitly provide: (i) a multi-layer calibration architecture that distinguishes timescales of identity change; (ii) an explicit disruption-repair typology that connects formal theory to clinical phenomena; and (iii) a phenomenological commitment; the claim that the subjective experience of identity (the sense of being oneself, the feeling of authenticity, the distress of inauthenticity) is a direct readout of attractor dynamics rather than an epiphenomenon.

8.2 Clinical Implications

The clinical implications of IAT are extensive, but several deserve particular emphasis.

The IAT account of personality disorders as attractor pathologies offers a theoretically grounded classification that maps naturally onto the DSM-5’s alternative model of personality disorders, which organizes personality pathology along dimensions of self-functioning and interpersonal functioning. From an IAT perspective, the DSM-5’s emphasis on self-clarity, self-direction, and identity coherence as key dimensions of personality pathology is a direct description of attractor geometry: self-clarity corresponds to the definition of μ_I; self-direction corresponds to the efficacy of the return function; and identity coherence corresponds to basin depth. The three pathological attractor configurations described in Section 6.3 (rigid, fragile, false) map onto the familiar Cluster A, B, and C distinctions: Cluster A rigidity corresponds to the rigid attractor with excessive prediction-error suppression; Cluster B fragility corresponds to the shallow basin with high oscillation; Cluster C patterns include elements of both false fixation and fragility in different proportions.

The emergence of psychedelic-assisted therapy (particularly psilocybin for treatment-resistant depression and PTSD) gains theoretical clarity through IAT. Psilocybin’s primary pharmacological action (serotonin 2A receptor agonism) produces a transient and dramatic reduction in the precision of high-level priors, temporarily dissolving the attractor’s characteristic shape and producing the phenomenologically distinctive experience of ego dissolution (Carhart-Harris et al., 2017). IAT interprets this as a temporary depotentiation of all three calibration layers simultaneously (a global flattening of the attractor landscape) that creates a period of radical openness in which reconsolidation can occur at a new, potentially healthier basin position. The therapeutic work during and immediately after the psychedelic experience is, on this account, the work of guiding re-seeding: what meaning-making, social mirroring, and somatic grounding take place in the immediate aftermath will shape the attractor that reconsolidates.

8.3 Philosophical Implications

IAT has significant implications for two long-standing debates in the philosophy of personal identity.

The first concerns personal identity over time. The psychological continuity theory associated with Locke and most influentially developed by Derek Parfit (1984) holds that personal identity consists in the continuity of psychological connections (memories, intentions, beliefs, and desires) between temporal stages of a person. The biological continuity theory holds that personal identity is constituted by the continuity of the biological organism. Both theories have well-known problems: psychological continuity theories struggle with fission cases and with the evident importance of biological continuity; biological continuity theories struggle to explain why identity seems to end with certain radical psychological transformations even when the organism persists.

IAT offers a third option: attractor continuity. Personal identity persists as long as the attractor basin persists in a recognizably continuous form; even when specific memories are incomplete (contra the psychological continuity theory’s emphasis on memory chains) and even when the biological substrate has been substantially renewed (contra biological continuity theory). The attractor can persist through partial memory disruption, biological change, and even moderate psychological transformation, as long as the basin geometry maintains sufficient continuity. Attractor continuity is not an all-or-nothing property but a graded one: identity is more or less continuous as the attractor undergoes more or less dramatic transformation. This is more phenomenologically adequate than either of the classical positions.

The second philosophical implication concerns free will. IAT grounds a compatibilist account of freedom: agency is the system’s capacity to act from within its attractor; to act in ways that are expressions of its own characteristic self-organization rather than being responses to external pressure that would push it outside the basin. Authentic action is attractor-consistent action; coerced or alienated action is action that moves the system outside its basin. On this account, freedom is not the ability to act arbitrarily (to do anything regardless of one’s character) but the ability to act from one’s deepest character, from the most central region of one’s attractor basin. This is a compatibilist position (the self’s actions are determined by its attractor, but the attractor is genuinely the self’s own) and it provides a neuroscientifically grounded account of what it means for an action to be authentically one’s own.

9. Conclusion

The Identity Attractor Theory proposes a comprehensive, multi-level, neurologically grounded, and formally tractable framework for understanding how the self persists, changes, and recovers across time. The theory rests on three core claims.

First: the self is a process, not an object. There is no self-entity located at a particular address in the brain or in the narrative archive. There is an attractor; a structured region of a high-dimensional state-space toward which the organism characteristically returns. Identity is the topology of this basin: its shape, its depth, its breadth, and the characteristic position of its center. To be a self is to be the kind of system that reliably inhabits and returns to this region of its state-space.

Second: the persistence of identity is maintained by nested feedback loops operating at multiple timescales. The mood layer operates in minutes to weeks, providing dynamic gain control. The personality layer operates in months to years, shaping the structural parameters of the basin. The moral layer operates across years and lifetimes, providing constitutional boundary conditions. Each layer is implemented by distinct neural systems operating at different levels of the processing hierarchy, from the interoceptive signals of the insular cortex to the autobiographical simulations of the DMN to the principled reasoning of the prefrontal cortex.

Third: disruption and repair are essential features of healthy identity dynamics, not failures. The identity attractor is not a static sanctuary but a dynamic equilibrium that is continuously challenged, defended, deformed, and restored. Micro-disruptions are the ordinary currency of living; meso-disruptions are the events that shape character; macro-disruptions are the crises that can either break the self or (if met with sufficient resources) expand it beyond what it previously was. Post-traumatic growth, understood through IAT, is attractor expansion: the capacity to hold more experience, more complexity, more ambiguity, without losing the characteristic shape of self.

IAT calls for a new generation of empirical research programs. Longitudinal neuroimaging studies should track DMN topology and connectivity metrics across documented identity transitions, testing the prediction that basin depth (operationalized through DMN coherence and within-network connectivity) predicts identity stability under stress. Interoceptive precision measures (heartbeat detection accuracy, interoceptive awareness questionnaires, autonomic flexibility indices) should be tested as predictors of attractor resilience in clinical and non-clinical populations. Computational models of identity basins in clinical populations should be developed, using reinforcement learning architectures and free-energy models to simulate attractor dynamics in conditions of personality disorder, trauma, and therapeutic intervention.

The identity attractor, we conclude, is not a prison; it is a home. It is the region of the self’s state-space that the organism knows how to return to, that it recognizes as its own, and from which it can venture into the world, knowing that it will find its way back. A healthy identity is not an identity that never changes; it is one that can change, that can be disrupted and reformed, and that knows itself well enough to locate its own center after the storm has passed.

“To be nobody but yourself in a world which is doing its best, night and day, to make you everybody else means to fight the hardest battle which any human being can fight.” – e.e. cummings

The Identity Attractor Theory proposes that this fight has a neuroscience, a dynamics, and a geometry; and that understanding these may be among the most important things a science of mind can offer to the human beings who are, every day, engaged in it.

References

• Andrews-Hanna, J. R., Reidler, J. S., Sepulcre, J., Poulin, R., & Buckner, R. L. (2010). Functional-anatomic fractionation of the brain’s default network. Neuron, 65(4), 550–562. https://doi.org/10.1016/j.neuron.2010.02.005
• Bruner, J. (1991). The narrative construction of reality. Critical Inquiry, 18(1), 1–21. https://doi.org/10.1086/448619
• Buckner, R. L., Andrews-Hanna, J. R., & Schacter, D. L. (2008). The brain’s default network: Anatomy, function, and relevance to disease. Annals of the New York Academy of Sciences, 1124(1), 1–38. https://doi.org/10.1196/annals.1440.011
• Carhart-Harris, R. L., Roseman, L., Bolstridge, M., Demetriou, L., Pannekoek, J. N., Wall, M. B., Tanner, M., Kaelen, M., McGonigle, J., Murphy, K., Leech, R., Curran, H. V., & Nutt, D. J. (2017). Psilocybin for treatment-resistant depression: fMRI-measured brain mechanisms. Scientific Reports, 7(1), 13187. https://doi.org/10.1038/s41598-017-13282-7
• Clark, A. (2016). Surfing uncertainty: Prediction, action, and the embodied mind. Oxford University Press.
• Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107(2), 261–288. https://doi.org/10.1037/0033-295X.107.2.261
• Craig, A. D. (2009). How do you feel – now? The anterior insula and human awareness. Nature Reviews Neuroscience, 10(1), 59–70. https://doi.org/10.1038/nrn2555
• Critchley, H. D., & Garfinkel, S. N. (2017). Interoception and emotion. Current Opinion in Psychology, 17, 7–14. https://doi.org/10.1016/j.copsyc.2017.04.020
• Damasio, A. R. (1994). Descartes’ error: Emotion, reason, and the human brain. Putnam.
• Deco, G., Jirsa, V. K., & McIntosh, A. R. (2011). Emerging concepts for the dynamical organization of resting-state activity in the brain. Nature Reviews Neuroscience, 12(1), 43–56. https://doi.org/10.1038/nrn2961
• DeYoung, C. G., Hirsh, J. B., Shane, M. S., Papademetris, X., Rajeevan, N., & Gray, J. R. (2010). Testing predictions from personality neuroscience: Brain structure and the Big Five. Psychological Science, 21(6), 820–828. https://doi.org/10.1177/0956797610370159
• Frankl, V. E. (1959). Man’s search for meaning. Beacon Press.
• Friston, K. (2010). The free-energy principle: A unified brain theory? Nature Reviews Neuroscience, 11(2), 127–138. https://doi.org/10.1038/nrn2787
• Frith, C. D. (1992). The cognitive neuropsychology of schizophrenia. Lawrence Erlbaum Associates.
• Gallagher, S. (2000). Philosophical conceptions of the self: Implications for cognitive science. Trends in Cognitive Sciences, 4(1), 14–21. https://doi.org/10.1016/S1364-6613(99)01417-5
• Gilligan, C. (1982). In a different voice: Psychological theory and women’s development. Harvard University Press.
• Haggard, P. (2017). Sense of agency in the human brain. Nature Reviews Neuroscience, 18(4), 196–207. https://doi.org/10.1038/nrn.2017.14
• Haidt, J. (2012). The righteous mind: Why good people are divided by politics and religion. Pantheon Books.
• Hohwy, J. (2013). The predictive mind. Oxford University Press.
• Kohlberg, L. (1981). The philosophy of moral development: Moral stages and the idea of justice. Harper & Row.
• Litz, B. T., Stein, N., Delaney, E., Lebowitz, L., Nash, W. P., Silva, C., & Maguen, S. (2009). Moral injury and moral repair in war veterans: A preliminary model and intervention strategy. Clinical Psychology Review, 29(8), 695–706. https://doi.org/10.1016/j.cpr.2009.07.003
• McCrae, R. R., & Costa, P. T., Jr. (1999). A five-factor theory of personality. In L. A. Pervin & O. P. John (Eds.), Handbook of personality: Theory and research (2nd ed., pp. 139–153). Guilford Press.
• McGilchrist, I. (2009). The master and his emissary: The divided brain and the making of the Western world. Yale University Press.
• Metzinger, T. (2003). Being no one: The self-model theory of subjectivity. MIT Press.
• Moscovitch, M., Cabeza, R., Winocur, G., & Nadel, L. (2016). Episodic memory and beyond: The hippocampus and neocortex for encoding, storage, and retrieval. Annual Review of Psychology, 67, 105–134. https://doi.org/10.1146/annurev-psych-113011-143823
• Northoff, G., Heinzel, A., de Greck, M., Bermpohl, F., Dobrowolny, H., & Panksepp, J. (2006). Self-referential processing in our brain – A meta-analysis of imaging studies on the self. NeuroImage, 31(1), 440–457. https://doi.org/10.1016/j.neuroimage.2005.12.002
• Parfit, D. (1984). Reasons and persons. Oxford University Press.
• Porges, S. W. (2011). The polyvagal theory: Neurophysiological foundations of emotions, attachment, communication, and self-regulation. Norton.
• Raichle, M. E. (2015). The brain’s default mode network. Annual Review of Neuroscience, 38, 433–447. https://doi.org/10.1146/annurev-neuro-071013-014030
• Raichle, M. E., MacLeod, A. M., Snyder, A. Z., Powers, W. J., Gusnard, D. A., & Shulman, G. L. (2001). A default mode of brain function. Proceedings of the National Academy of Sciences, 98(2), 676–682. https://doi.org/10.1073/pnas.98.2.676
• Ramachandran, V. S., & Blakeslee, S. (1998). Phantoms in the brain: Probing the mysteries of the human mind. William Morrow.
• Ricoeur, P. (1992). Oneself as another (K. Blamey, Trans.). University of Chicago Press. (Original work published 1990)
• Roberts, B. W., Walton, K. E., & Viechtbauer, W. (2006). Patterns of mean-level change in personality traits across the life course: A meta-analysis of longitudinal studies. Psychological Bulletin, 132(1), 1–25. https://doi.org/10.1037/0033-2909.132.1.1
• Schacter, D. L., Addis, D. R., Hassabis, D., Martin, V. C., Spreng, R. N., & Szpunar, K. K. (2012). The future of memory: Remembering, imagining, and the brain. Neuron, 76(4), 677–694. https://doi.org/10.1016/j.neuron.2012.11.001
• Schechtman, M. (1996). The constitution of selves. Cornell University Press.
• Seth, A. K. (2013). Interoceptive inference, emotion, and the embodied self. Trends in Cognitive Sciences, 17(11), 565–573. https://doi.org/10.1016/j.tics.2013.09.007
• Seth, A. K. (2021). Being you: A new science of consciousness. Dutton.
• Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Brooks/Cole.
• Tedeschi, R. G., & Calhoun, L. G. (2004). Posttraumatic growth: Conceptual foundations and empirical evidence. Psychological Inquiry, 15(1), 1–18. https://doi.org/10.1207/s15327965pli1501_01
• van der Kolk, B. A. (2014). The body keeps the score: Brain, mind, and body in the healing of trauma. Viking.
• van Kesteren, M. T. R., Ruiter, D. J., Fernández, G., & Henson, R. N. (2012). How schema and novelty augment memory formation. Trends in Neurosciences, 35(4), 211–219. https://doi.org/10.1016/j.tins.2012.02.001

End of Paper – Identity Attractor Theory (IAT) | Theoretical Paper | Prepared: July 2026