Arc of Self

1Abstract

Identity is modeled as a closed, re-entrant loop: a fourteen-node architecture in four phases (Hardware, OS Install, Engine, Loop) in which cultural concepts, installed through an innate loader (pre-wired core-knowledge systems), assemble into Personality Properties; the Properties generate a standing demand for validation; the demand shapes attention and situation selection; behavior is executed to confirm the self-model; and the resulting feedback is filtered in the self-model's favor before it can revise the concepts that produced it. Refined concepts re-enter the circuit through a direct recursion from Refinement to the attention gate (node 14 to node 10), and hedonic adaptation attenuates each recognition reward toward baseline, making the recursion an expanding spiral rather than a circle. The framework synthesizes established results (behavioral genetics, core knowledge, predictive processing, self-verification, dissonance, hedonic adaptation); its contributions are the closed architecture itself, the full loop pass as the unit of analysis, an account of emotional concepts as full-loop configurations, each, in the pool mapped to date, carrying an addressable hot node (the point of maximum intervention leverage), and the Collision Layer, a mechanistic description of how two identity loops modify each other through unlabeled behavioral output. Five falsifiable predictions are staked. The framework states its own epistemic status explicitly: it is a concept inside the loop it describes, and prediction error, the loop's single structural opening, grounds both its account of change and its own limits.

2Introduction

Ask what identity is and the literature answers in segments. Behavioral genetics documents heritable dispositions and gene-environment interplay (Plomin & Deary, 2015; Belsky & Pluess, 2009). Developmental psychology traces how caregivers, peers, and culture shape the growing person (Waters, Merrick, Treboux, Crowell, & Albersheim, 2000; Harris, 1998). Self-psychology describes the structure of self-knowledge (Markus, 1977) and the ways people seek confirmation of it (Swann, 1983). Predictive processing recasts the brain as a prediction machine whose perception and action are model-driven (Clark, 2013; Seth, 2013), and constructionist accounts extend this to emotion (Barrett, 2017). Dissonance theory explains why contradiction is aversive (Festinger, 1957), dual-process theory why most behavior runs on automatic (Kahneman, 2011), and biological accounts assemble the causal chain behind a single act (Sapolsky, 2017).

Each of these accounts is well supported, and closed feedback-loop models exist. Self-regulation theory treats behavior as discrepancy reduction toward goals (Carver & Scheier, 1998), and identity control theory, the nearest rival to the present account, models identity maintenance itself as a control loop: an identity standard, perceived reflected appraisals, a comparator, and behavior that counteracts the discrepancy (Burke, 1991; Burke & Stets, 2009). But identity control theory's engine is restoration: behavior moves perception back toward a standard that changes only slowly, drifting under persistent discrepancy. The architecture developed here differs on three counts. Feedback is filtered in the self-model's favor before any comparison can use it. The refinement stage rewrites the concept set itself, and the rewritten lens re-enters as the filter through which the next situation is selected, so the standard is the loop's running product rather than its reference point. And hedonic adaptation strips the reward from achieved congruence, so the circuit tends to escalate where a control loop would settle: a proposed tendency, argued in Sections 4 and 5.14, not a derived necessity.

What remains missing in the literature, then, is a loop whose operating substance is identity itself: how installed concepts assemble into a self, how that self generates a standing demand for validation, how the demand shapes attention, situation choice, and action, how the resulting feedback is filtered before it can teach anything, and how the filtered lessons re-enter the circuit as the next pass's lens: the revised concept set through which the next situation will be perceived and selected. Identity, on the view developed here, is not a structure but a running process, and its most consequential properties (self-reinforcement, escalation, resistance to disconfirmation, the narrow conditions under which it changes) are jointly explained when the loop is analyzed whole.

This paper presents such an account: a fourteen-node model in four phases, connected in a single closed, re-entrant circuit: a developmental lap (nodes 1-14) that builds the machine, and a running circuit (nodes 10-14) that operates it. The lap is not repeated wholesale, but its installation machinery stays available for life, so new concepts and new Properties continue to arrive through the loop's one opening. The model is a synthesis, not a replacement. Nearly every node stands on established results, but the architecture is the contribution: the closed circuit itself, the Loop Pass as the unit of analysis, an account of emotional concepts as complete loop passes with distinct signatures, and the Collision Layer, a mechanistic description of how two identity loops modify each other through behavior neither party labels as feedback. The model also states its own epistemic status without special pleading: it does not claim a position outside the loop it describes.

A word on provenance is owed to the reader. This framework was developed over seventeen years by an independent researcher working outside academia, and it reaches print after, not before, its public form: an interactive visualization at arc-of-self.com where the loop can be explored node by node. The direction of travel, from a lived and iterated model toward the literature rather than from the literature outward, explains the framework's shape: it was built to answer "why do people do what they do, and why do I?", and the academic apparatus was attached to a structure that had been iterated through seventeen years of daily use. The reader is invited to treat that history as neither credential nor disqualification, and to judge the architecture on whether it explains and predicts.

3Key Concepts & Definitions

The framework introduces a small set of proper terms. Each is defined below and positioned against its nearest established counterpart. Where an existing term was not adopted, the divergence is deliberate and the reason is stated: the goal is precision within this architecture, not reinvention.

4Architectural Overview

The model consists of fourteen nodes organized into four phases, connected in a single closed circuit with two readings: the developmental lap (1-14), which dominates while the machine is built, and the running circuit (10-14), the loop that operates for the rest of life. The lap is never re-run wholesale; its machinery of absorption and self-assembly stays available through the opening described below, which is how adults still adopt new concepts and new Properties. The escalator below depends on exactly this. The phase names follow a deliberate computational metaphor (Hardware, OS Install, Engine, Loop) because the metaphor carries real structure: an operating system cannot be installed on bare metal without firmware, and the firmware is not the operating system.

A note on the metaphor's limits is owed at the outset, because the framework uses it as scaffolding, not ontology. In a computer, hardware and software remain physically distinct; in a brain they do not. Installing a concept is a structural event: Hebbian reinforcement physically rewires the substrate, so yesterday's software becomes tomorrow's hardware. The hardware/software boundary in this model is therefore developmental, not physical: it separates what arrives before experience from what experience installs, and claims nothing more. Nor is the framework committed to the thesis that the brain is a computer; each node stands on its own evidence, and the metaphor's job is to hold the architecture in the reader's mind, not to do the arguing. Where the metaphor breaks, the model sides with the brain.

Phase 1: Hardware (nodes 1-3). What is present before experience: the genome's influence on cognitive capacities, temperament, and susceptibility to environmental influence; the modular profile of intelligence; and the baseline state of the newborn mind, empty of cultural content but structurally pre-wired to receive it. Phase 1 supplies the loader that makes Phase 2 possible.

Phase 2: OS Install (nodes 4-7). The installation of culture: the socio-cultural circle whose concepts become the person's needs; the specific Actors of Influence who deliver them; the mechanics of Concept Absorption; and Reality Construction, where the accumulated concepts assemble into the world the person can experience. Different circles install different realities: not primarily because brains differ, but because what surrounds them differs.

Phase 3: Engine (nodes 8-9). The installed material assembles into a self: Personality Properties are adopted from available concepts, and from them the need structure emerges, topped by Personality Properties Recognition, the standing demand that the self-model be validated. This demand is the engine that drives the loop's identity-maintaining rotation; once survival and safety are met, the framework locates the primary driver here, not in pleasure-seeking or pain-avoidance as such.

Phase 4: Loop (nodes 10-14). The running circuit. The Affective Niche filters the environment into Situational Engagement: enrollment, mostly automatic, into situations where Properties can earn recognition. Engagement produces action (inaction included: declining, avoiding, and staying silent are behavioral responses protecting specific Properties from disconfirmation). Action produces feedback, which the brain processes asymmetrically: the framework proposes that Hebbian reinforcement (co-active neurons strengthening their connection; Hebb, 1949) makes confirming feedback cheaper to absorb than disconfirming feedback, so the refinement stage operates on pre-filtered input. Feedback drives Concept Refinement, and the refined concepts re-enter the circuit as the next pass's lens.

The recursion. The circle as drawn is developmental: nodes 1 through 9 build the machine, and the running circuit is nodes 10-14. The loop's re-entry is the arrow from Concept Refinement back to the Affective Niche (14 → 10): refined concepts feed the niche directly, so the lens through which the next situation is selected has already been modified by the last pass. Where the edits land must be stated precisely: refinement writes in place into the standing structure, adjusting concept weights in the constructed reality and revising the Properties it houses (nodes 6-9). The 14 → 10 arrow is not a second write channel; it marks where those edits take effect. The rewritten lens is already in force at the attention gate when the next situation is selected, with no re-run of development.

The escalator. Hedonic adaptation pulls the response to each recognition event back toward baseline; adaptation is partial but relentless (Frederick & Loewenstein, 1999; Diener, Lucas, & Scollon, 2006). What felt like validation becomes the new normal, so the loop tends to escalate, seeking stronger recognition for the same Property or validation of new ones. The recursion is therefore not circular but spiral: an expanding structure of need. This is why achievement so rarely satisfies for long, and why the loop runs from birth to death without a resting state.

The loop has exactly one structural opening: prediction error. When the world fails to match the brain's expectation, the confirmation filter has nothing prepared, and for that moment new content can force itself into experience as a raw mismatch. The definition must be exact, because not every mismatch qualifies: a filter can carry prepared recodings, and a mismatch it can recode is absorbed by the self-model, not admitted past it (the shame configuration of Section 6 runs on precisely such recoding). The opening is the residual class: prediction error for which the filter has no ready recoding. The residual is not defined by its outcome, which would make the claim untestable. A prepared recoding is a standing construction, elicitable before the event (shame's "if they knew" antecedent is exactly such a construction), and Prediction 2 anchors the opening to an observable marker, visible expectation failure, independent of whether revision follows. The claim that it is the only opening stands. The opening's width varies across development: in infancy nearly everything is mismatch, since no predictions are consolidated yet, so the door stands wide open and installation proceeds at full flow; as the map hardens and the confirmation filter grows around the self-model, the same door narrows to moments. Every concept a person will ever absorb or invent enters through this opening; what changes over a lifetime is only how wide it stands. Seeing one's own loop is accordingly not a state to be reached once but a repeated event at these moments of failure: a trainable habit of looking into the cracks instead of sealing them. The framework applies this to itself; it does not claim a position outside the loop it describes.

5The Loop: Node-by-Node Analysis

Each node is presented in three registers: its mechanics within the model, the research it stands on, and, marked explicitly, where the author's synthesis begins. The synthesis markers matter: the reader should always know which claims lean on the literature and which are the framework's own.

6Emotional Concepts as Complete Loop Passes

In this framework an emotion is not a discrete module firing, and not a standalone state that visits the person from outside. An emotional concept is a specific configuration of one full pass of the loop: a signature distribution of work across the fourteen nodes. This is consistent with the constructed-emotion view that emotions are concept-driven constructions rather than triggered essences (Barrett, 2017); what the architecture adds is addressability. Being a configuration does not by itself guarantee a point of concentration; it is an empirical regularity of every configuration mapped to date that the distinctive work clusters at one node. That node is the emotion's Hot Node: the point in the circuit where its signature is most visible and where intervention has the most leverage. Whether single-peak concentration is universal, or some emotional concepts run twin-peaked, is an open question; the claim here extends only to the mapped pool. Three worked examples follow, chosen so that their hot nodes fall at three different points of the circuit.

Envy (hot node: Affective Niche). The envy configuration does its work in attention. To confirm a Property, the system tunes itself to the relevant signals: if the property is "Creative," other people's creative output is noticed; if "Good provider," what other providers deliver. The scan for comparison data is automatic, and it runs only in the domains the person's installed reality marked as meaningful. The felt sting of envy is the output of measurement that already happened. This locates the leverage precisely: not in suppressing the comparison (the niche is upstream of will) but in auditing which Actor of Influence installed the price tag on that domain.

Anger (hot node: Reality Construction). The anger configuration lives in the map: the constructed reality of node 07. Absorbed rules ("fairness exists," "people should behave a certain way") assemble into a working model that does not present itself as a model; it presents itself as reality. A violation is therefore experienced not as a map error but as a world error: the map is right, the world is wrong, and the wrong must be corrected. Behavioral output is enforcement, and the feedback that would help most ("your map might be wrong") is the one signal the configuration is structurally least equipped to receive. Leverage sits at the map's status: the moment a rule is recognized as a rule rather than a law of nature, the enforcement mandate weakens.

Shame (hot node: Feedback Reception). The shame configuration corrupts the return channel. Its behavioral output is concealment (withdrawal, deflection, overcompensation), and the raw social data coming back is often neutral or positive: nothing bad actually happened. But the data does not arrive clean. It passes through a construction that already contains "if they knew, they would reject me," so neutral feedback is recoded: their ease becomes "they did not see it this time," their warmth becomes "they would not be warm if they knew." Successful validation, uniquely, looks like absence: nothing exposed, nothing contradicted. The felt signal is not the feedback itself; it is the recoding construction intercepting the feedback before refinement can use it. Leverage sits in the recoding step, consistent with inhibitory-learning accounts of exposure (Craske et al., 2014): disclosure removes the "if they knew" antecedent, leaving a prediction error the filter has no ready way to recode.

The three configurations use the same fourteen nodes; nothing was added to the architecture to accommodate them. The currently mapped pool covers maladaptive configurations (envy, shame, anger, impostor syndrome, and others); positive-valence passes are unmapped as of this writing and are listed as an open task in Section 9.

7The Collision Layer

Loops do not run in isolation. One person's Behavioral Expression arrives in another person's loop as input, and it arrives unlabeled: behavior does not come stamped with the configuration that produced it. The receiving loop must explain the input using its own map, because that is the only map it has. The explanation it settles on then drives its own behavioral response, which returns to the first loop as equally unlabeled feedback. This coupling of two closed circuits through a channel with no shared labels is the Collision Layer. To our knowledge no single account combines closed two-loop coupling with identity maintenance as the coupling force, though circular-interaction traditions come close (double contingency; participatory sense-making: De Jaegher & Di Paolo, 2007).

The dynamics follow from the architecture. Consider the anger configuration colliding with a receiving loop. Anger's output is correction force: a demand, not a question. The receiver's map converts that force into one of a small number of responses: comply (which hardens the sender's map, since it received compliance instead of counter-evidence), fight (two constructions each presenting itself as reality, each processing the other's resistance as proof), shut down (unreadable silence, which the sender's loop escalates against), or exit (which the sender's refinement records not as "my map was wrong" but as "they could not handle the truth"). The framework's own change mechanism supplies a fifth path: a response the sender's map did not predict, neither compliance nor counter-force (an unexpected question, undefended calm, naming the rule without a verdict). An unpredicted response is a prediction error inside the sender's loop, the one input the correction filter has nothing prepared against, and therefore the only path with a mechanism for opening the sender's map rather than hardening it. In each of the four predicted paths, the receiver's response becomes the sender's next round of feedback, processed through a filter that had already concluded it was right. Neither loop sees the other's mechanics; each is typically certain it is reading the situation correctly.

The nearest existing accounts each capture one arc of this circuit. The looking-glass self describes identity as built from imagined judgments of others (Cooley, 1902). Behavioral-confirmation research shows that one person's expectation measurably reshapes the other's behavior (Snyder, Tanke, & Berscheid, 1977). What these accounts lack is the closed two-loop coupling: a mechanism in which each party's identity maintenance is the other's environment, with no labeled channel between them. From this mechanism follows a distinct prediction (prediction 5, Section 9): in close dyads, changes in one partner's self-descriptions should track the other partner's unlabeled behavioral stream even after controlling for explicit evaluative feedback. The framework adds one boundary condition: the implicit stream does not substitute for explicit recognition indefinitely. Explicit recognition of a Property remains a distinct need, and a relationship configuration in which it starves while only behavioral feedback remains is predicted to be unstable with respect to that Property. The architectural reason: only labeled recognition confirms a Property as seen by another map; unlabeled behavior sustains inference but never delivers that outside confirmation.

8Architectural Overlays

Three phenomena are conspicuously absent from the node list: trauma, faith, and death awareness. This is deliberate. They are not nodes; they are overlays: persistent modifiers of how existing nodes operate, without any change to the loop's structure.

Trauma modifies signal sensitivity. It raises the threat weighting inside the Affective Niche (the niche defaults to narrow), lowers the threshold at which Feedback Reception classifies input as attack, and biases Situational Engagement toward avoidance. The loop runs the same circuit with recalibrated gain: threat learning is among the most conserved and best-documented forms of plasticity (LeDoux, 2000). Trauma enters the ledger twice, and the framework owes the distinction. Its content ("the world is dangerous," "I cannot protect myself") is ordinary Concept Absorption at maximum door width: a traumatic event is a massive prediction error, the filter has nothing prepared, and the appraisals it installs are exactly the post-traumatic cognitions clinical models describe (Ehlers & Clark, 2000). That requires no new machinery. What makes trauma an overlay rather than a node is the second entry: the recalibrated gain that persists after the event, a standing distortion of the channel through which all subsequent content passes.

Faith and belief systems modify Reality Construction. A faith installs a concept package that reinterprets feedback globally: outcomes become messages, suffering becomes test or purpose, and an always-present Actor of Influence is added whose recognition is available in every situation, including situations no human witnesses. In loop terms this is a profound modification: it makes validation of certain Properties independent of social feedback, which can stabilize the loop (recognition never starves) and insulate it (disconfirming feedback is recoded upstream of refinement). The framework takes no position on the truth of any faith; it describes what the overlay does to the machinery. Meaning-system research documents these functions (Park, 2010). One structural question is acknowledged: a standing, always-available recognition source is arguably a permanent collision partner rather than a parameter change; the framework counts it as parametric because no new node is required, only a standing input.

Death awareness modifies the weighting inside the Validation Need. Mortality salience shifts striving toward Properties that promise symbolic continuity: legacy, works, children, causes larger than the self. This is the terror-management claim (Greenberg, Pyszczynski, & Solomon, 1986; Becker, 1973), cited as theory rather than settled result since its flagship mortality-salience effect has faced multi-lab replication failure, restated architecturally: awareness of the loop's terminal boundary reweights which recognitions matter, typically away from immediate social validation and toward validation by imagined future audiences.

The three analyses above are offered as evidence that overlay status suffices for these phenomena: the circuit stays intact while its operating parameters change, so the same fourteen-node analysis applies to a traumatized loop, a believing loop, and a mortality-aware loop. Whether every such modifier reduces to parameter changes is an open question.

9Limitations, Open Questions & Future Directions

Epistemic status. This framework was produced by the machinery it describes. It was built by a brain inside a loop, it served its author's Personality Properties during seventeen years of construction, and it passed through the same property-biased feedback filter it documents. The framework therefore does not claim to stand outside the loop or to describe reality as it is. It claims to be a concept, and by its own account of concepts it asks to be judged the way concepts are judged: by how much context it delivers per word, by what it explains, and by what it predicts. This is a self-consistent position, and the one the framework adopts; it coincides with ordinary theory appraisal, explanatory scope and predictive success, so nothing in what follows asks the reader for a special standard.

The theory predicts its own failure mode. By the model's rules, "I see my loop" is itself a Personality Property. Once adopted, it demands validation like any other: the holder will seek situations confirming their awareness and filter evidence of their own automaticity. The theory thus predicts that its adherents can degrade into dogmatists indistinguishable from adherents of any ideology, and it offers only the mechanism it identifies for everyone else: prediction error, the single moment the filter has nothing prepared. Seeing the loop is a repeated event, not a status; this clause applies to the author and to any adherent of the framework.

Untested architecture. Nearly every individual node stands on established results, cited in Section 5. The architecture is the untested claim: that these mechanisms compose into one closed re-entrant circuit whose full pass is the correct unit of analysis. The framework therefore stakes five falsifiable predictions that follow from the architecture as a whole rather than from any single node:

1. Property-specific confirmation filtering. Biased processing of social feedback is organized around Personality Properties, and the magnitude of the bias scales with a property's centrality in the person's self-inventory (their enumerable set of adopted Properties). Generic confirmation-bias accounts predict bias, and motivated-reasoning accounts predict scaling with motivational stake (Kunda, 1990); the distinct claim is organization over an enumerable set of Properties, with a measurable per-property centrality gradient. In plain terms: the more central a trait is to who you think you are, the deafer you are to evidence against it. Rank a person's traits by importance and you can predict where they stop hearing.
2. The prediction-error window. Revision of self-descriptions concentrates in the moments immediately following visible expectation failure, with a moderator the framework itself supplies: prepared recodings are standing constructions, so their availability is measured before the event, and revision concentrates after failures for which no prepared recoding was elicited beforehand. Failures matched by a pre-elicited recoding are predicted to produce reinterpretation, not revision. The same challenging information delivered at a neutral moment produces measurably less change. Predictive-processing accounts time-lock perceptual updating to error; they have not tested the claim at the level of verbal self-description. In plain terms: people rewrite their self-story right after life visibly breaks their expectations, and only when they had no excuse ready in advance. With a ready excuse the surprise gets explained away instead. The same hard truth on a calm day changes almost nothing.
3. Bundle defense. Properties adopted as pre-packaged bundles are defended as a unit: challenging one never-individually-evaluated component triggers defense of the whole bundle, unlike challenging an individually adopted Property. In plain terms: traits adopted as a package (a religion, an ideology, a role model) are defended as a package. Poke one piece and the whole set fights back. A trait picked up on its own does not do that.
4. The self-aware dogmatist. Holders of "I am self-aware" as a Property (the "I see my loop" Property above) show standard confirmation filtering about their own awareness, with magnitude following the same centrality gradient as any other Property. No rival account known to us makes it a central prediction that awareness of the mechanism becomes fuel for the mechanism. In plain terms: "I am self-aware" is just another trait, so people defend it like any other. The more someone prides themselves on seeing their own biases, the better they get at not seeing them.
5. Collision without evaluation. In close dyads, changes in one partner's self-descriptions track the other partner's unlabeled behavioral stream after controlling for explicit evaluative feedback, with the boundary condition stated in Section 7: implicit input does not indefinitely substitute for explicit recognition. In plain terms: living close to someone reshapes how you describe yourself even when they never praise or criticize you. But silent influence has a limit: sooner or later the self still needs to be recognized out loud.

Known limitations. The model's provenance is a single sustained introspective and observational project, systematized against the literature after the fact; that is a bias source the predictions above are designed to discipline. The mechanism is claimed as universal while installation content is culture-specific, and this claim inherits the WEIRD-sampling problem (Western, Educated, Industrialized, Rich, Democratic; Henrich, Heine, & Norenzayan, 2010) of the underlying literature. Personality Properties currently lack a standardized measurement instrument (a property-inventory method is a prerequisite for testing predictions 1, 3, and 4). The mapped emotion pool covers maladaptive configurations; positive-valence passes (pride, joy, gratitude) remain unmapped, and mapping them is a direct test of the claim that the same architecture suffices. The computational metaphor's limits are stated in Section 4.

Future directions. Three lines follow naturally: formalization (the loop invites an explicit computational model with simulated collision dynamics), dyadic longitudinal studies instrumenting prediction 5, and intervention design (hot-node targeting suggests that interventions matched to an emotion's hot node should outperform generic protocols). The interactive form of the theory at arc-of-self.com, where the loop can be traversed and interrogated conversationally, is offered as an instrument for the dissemination and stress-testing of theoretical models generally.

10References

All 57 entries verified against primary sources, August 2026.

Andersen, S. M., & Chen, S. (2002). The relational self: An interpersonal social-cognitive theory. Psychological Review, 109(4), 619-645.

Ariely, D. (2008). Predictably irrational: The hidden forces that shape our decisions. HarperCollins.

Barrett, L. F. (2017). How emotions are made: The secret life of the brain. Houghton Mifflin Harcourt.

Becker, E. (1973). The denial of death. Free Press.

Belsky, J., & Pluess, M. (2009). Beyond diathesis stress: Differential susceptibility to environmental influences. Psychological Bulletin, 135(6), 885-908.

Berger, P. L., & Luckmann, T. (1966). The social construction of reality: A treatise in the sociology of knowledge. Doubleday.

Blakemore, S.-J. (2008). The social brain in adolescence. Nature Reviews Neuroscience, 9(4), 267-277.

Bless, H., Clore, G. L., Schwarz, N., Golisano, V., Rabe, C., & Wölk, M. (1996). Mood and the use of scripts: Does a happy mood really lead to mindlessness? Journal of Personality and Social Psychology, 71(4), 665-679.

Brickman, P., & Campbell, D. T. (1971). Hedonic relativism and planning the good society. In M. H. Appley (Ed.), Adaptation-level theory: A symposium (pp. 287-302). Academic Press.

Burke, P. J. (1991). Identity processes and social stress. American Sociological Review, 56(6), 836-849.

Burke, P. J., & Stets, J. E. (2009). Identity theory. Oxford University Press.

Carey, S. (2009). The origin of concepts. Oxford University Press.

Carver, C. S., & Scheier, M. F. (1998). On the self-regulation of behavior. Cambridge University Press.

Chomsky, N. (1980). Rules and representations. Columbia University Press.

Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences, 36(3), 181-204.

Coie, J. D., Terry, R., Lenox, K., Lochman, J., & Hyman, C. (1995). Childhood peer rejection and aggression as predictors of stable patterns of adolescent disorder. Development and Psychopathology, 7(4), 697-713.

Cooley, C. H. (1902). Human nature and the social order. Scribner's.

Craske, M. G., Treanor, M., Conway, C. C., Zbozinek, T., & Vervliet, B. (2014). Maximizing exposure therapy: An inhibitory learning approach. Behaviour Research and Therapy, 58, 10-23.

De Jaegher, H., & Di Paolo, E. (2007). Participatory sense-making: An enactive approach to social cognition. Phenomenology and the Cognitive Sciences, 6(4), 485-507.

Diener, E., Lucas, R. E., & Scollon, C. N. (2006). Beyond the hedonic treadmill: Revising the adaptation theory of well-being. American Psychologist, 61(4), 305-314.

Easterbrook, J. A. (1959). The effect of emotion on cue utilization and the organization of behavior. Psychological Review, 66(3), 183-201.

Ehlers, A., & Clark, D. M. (2000). A cognitive model of posttraumatic stress disorder. Behaviour Research and Therapy, 38(4), 319-345.

Festinger, L. (1957). A theory of cognitive dissonance. Row, Peterson.

Fraley, R. C. (2002). Attachment stability from infancy to adulthood: Meta-analysis and dynamic modeling of developmental mechanisms. Personality and Social Psychology Review, 6(2), 123-151.

Frederick, S., & Loewenstein, G. (1999). Hedonic adaptation. In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: The foundations of hedonic psychology (pp. 302-329). Russell Sage Foundation.

Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: The broaden-and-build theory of positive emotions. American Psychologist, 56(3), 218-226.

Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. Basic Books.

Goffman, E. (1959). The presentation of self in everyday life. Doubleday.

Greenberg, J., Pyszczynski, T., & Solomon, S. (1986). The causes and consequences of a need for self-esteem: A terror management theory. In R. F. Baumeister (Ed.), Public self and private self (pp. 189-212). Springer.

Gross, J. J. (1998). The emerging field of emotion regulation: An integrative review. Review of General Psychology, 2(3), 271-299.

Gunnar, M., & Quevedo, K. (2007). The neurobiology of stress and development. Annual Review of Psychology, 58, 145-173.

Harris, J. R. (1998). The nurture assumption: Why children turn out the way they do. Free Press.

Hebb, D. O. (1949). The organization of behavior: A neuropsychological theory. Wiley.

Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2-3), 61-83.

Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux.

Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108(3), 480-498.

LeDoux, J. E. (2000). Emotion circuits in the brain. Annual Review of Neuroscience, 23, 155-184.

Markus, H. (1977). Self-schemata and processing information about the self. Journal of Personality and Social Psychology, 35(2), 63-78.

Maslow, A. H. (1943). A theory of human motivation. Psychological Review, 50(4), 370-396.

Mazar, N., Amir, O., & Ariely, D. (2008). The dishonesty of honest people: A theory of self-concept maintenance. Journal of Marketing Research, 45(6), 633-644.

McAdams, D. P. (2001). The psychology of life stories. Review of General Psychology, 5(2), 100-122.

Mead, G. H. (1934). Mind, self, and society. University of Chicago Press.

Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phenomenon in many guises. Review of General Psychology, 2(2), 175-220.

Park, C. L. (2010). Making sense of the meaning literature: An integrative review of meaning making and its effects on adjustment to stressful life events. Psychological Bulletin, 136(2), 257-301.

Pinker, S. (2002). The blank slate: The modern denial of human nature. Viking.

Plomin, R., & Deary, I. J. (2015). Genetics and intelligence differences: Five special findings. Molecular Psychiatry, 20(1), 98-108.

Sapolsky, R. M. (2017). Behave: The biology of humans at our best and worst. Penguin Press.

Scarr, S., & McCartney, K. (1983). How people make their own environments: A theory of genotype → environment effects. Child Development, 54(2), 424-435.

Schwabe, L., & Wolf, O. T. (2009). Stress prompts habit behavior in humans. Journal of Neuroscience, 29(22), 7191-7198.

Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being: Informative and directive functions of affective states. Journal of Personality and Social Psychology, 45(3), 513-523.

Seth, A. K. (2013). Interoceptive inference, emotion, and the embodied self. Trends in Cognitive Sciences, 17(11), 565-573.

Snyder, M., Tanke, E. D., & Berscheid, E. (1977). Social perception and interpersonal behavior: On the self-fulfilling nature of social stereotypes. Journal of Personality and Social Psychology, 35(9), 656-666.

Spearman, C. (1904). "General intelligence," objectively determined and measured. American Journal of Psychology, 15(2), 201-292.

Spelke, E. S., & Kinzler, K. D. (2007). Core knowledge. Developmental Science, 10(1), 89-96.

Stryker, S. (1980). Symbolic interactionism: A social structural version. Benjamin/Cummings.

Swann, W. B., Jr. (1983). Self-verification: Bringing social reality into harmony with the self. In J. Suls & A. G. Greenwald (Eds.), Psychological perspectives on the self (Vol. 2, pp. 33-66). Erlbaum.

Waters, E., Merrick, S., Treboux, D., Crowell, J., & Albersheim, L. (2000). Attachment security in infancy and early adulthood: A twenty-year longitudinal study. Child Development, 71(3), 684-689.