Korte Samenvatting

Afscheid van de computermetafoor: Ons geheugen werkt niet als een computerarchief, maar is een dynamisch reconstructieproces waarbij herinneringen telkens opnieuw worden gevormd in interactie met de omgeving.

De quaternion-agent: De theorie beschrijft een persoon aan de hand van vier denkwijzen (eenheid, zintuiglijk, sociaal en verstandelijk), die voortdurend op elkaar inwerken.

Vermijden van verrassing: Het principe van ‘vrije energie’ drijft ons geheugen: het systeem streeft naar minimale voorspellingsfout. Een herinnering is een stabiele denkconfiguratie die in een bepaalde context ‘verrassing’ minimaliseert.

Relationeel geheugen: Geheugen ontstaat niet alleen in individuen, maar in de relatie tussen mensen. Dominantie, uitgebreide modaliteit en interferentie bepalen hoe herinneringen in een sociale interactie worden gevormd en doorgegeven.

Nieuwe kijk op Alzheimer: Bij Alzheimer neemt de ‘ruis’ in het systeem toe, waardoor diepe (oude) herinneringen langzamer vervagen dan recente. Een stabiele sociale omgeving kan dit proces deels compenseren.

J.konstapel, Leiden,16-4-2026.

Jump to the english version here.

---

Al meer dan vijftig jaar vergelijken hersenwetenschappers ons geheugen met een computer: een systeem waarin herinneringen worden opgeslagen als bestanden in een archief. Maar steeds meer onderzoek toont aan dat dit beeld niet klopt. Herinneringen zijn geen vaste kopieën van het verleden, maar worden elke keer opnieuw geconstrueerd. Daarom stelt een nieuwe theorie, de ‘Relationele Veldtheorie van het Geheugen’, dat we het geheugen anders moeten zien: niet als een opslagplaats, maar als een dynamisch proces dat ontstaat in de relatie tussen een persoon, zijn omgeving en andere mensen.

De Quaternion-agent: Vier kanten van ons denken

De theorie beschrijft een persoon als een ‘quaternion’, een wiskundig punt op een bol in vier dimensies. Deze vier dimensies staan voor vier manieren van denken, gebaseerd op het werk van McWhinney (1997):

1. Eenheidsdenken (wᵤ): gaat over het scheppen van betekenis, intuïtie en het zien van het grotere geheel.
2. Zintuiglijk denken (wₛ): gaat over ervaringen, gevoelens en je lichaam.
3. Sociaal denken (wₛₒ): gaat over relaties, normen en gemeenschap.
4. Verstandelijk denken (wₘ): gaat over analyse, logica en systemen.

Deze vier manieren staan niet los van elkaar; als de ene manier meer aandacht krijgt, moeten de anderen dat compenseren. Ook de volgorde waarin we deze manieren gebruiken is belangrijk, bijvoorbeeld bij het aangaan van een relatie.

Geheugen als ‘verrassing vermijden’

Het drijvende principe achter ons geheugen is volgens de theorie het Vrije-energieprincipe van Karl Friston. Dit principe stelt dat elk systeem streeft naar een toestand van minimale ‘verrassing’ of voorspellingsfout. Een herinnering is dan een stabiele denkconfiguratie die deze verrassing in een bepaalde context minimaliseert. Hoe sterker een herinnering is, hoe dieper de ‘trechter’ (attractor) in het denklandschap waarin de configuratie valt. Ophalen is dan het opnieuw in die trechter terechtkomen, wat altijd een reconstructief proces is (zoals Bartlett al in 1932 beschreef). Encoderen, consolideren en ophalen zijn dus geen aparte processen meer.

Het relationele veld: Samen herinneren

De meest opvallende bewering van de theorie is dat geheugen geen individuele eigenschap is, maar ontstaat in relaties. Een interactie tussen twee mensen (bijvoorbeeld een leraar en een leerling) wordt gezien als één systeem. In deze relatie zorgen drie mechanismen voor het vormen van geheugen:

• Dominantie (λᴅ): de meer ervaren persoon (referent) structureert het ‘verrassingslandschap’ van de ander.
• Uitgebreide modaliteit (γ): de leerling kan in de relatie tijdelijk configuraties bereiken die normaal buiten zijn bereik liggen (de ‘Zone van Naaste Ontwikkeling’ van Vygotsky).
• Interferentie (κ): als de relatie uit balans is, kost dat energie en ontstaan er onstabiele toestanden, wat een verklaring kan zijn voor angstige hechting (Bowlby, 1969).

Wat wij als ‘cultuur’ of ‘gedeelde normen’ ervaren, zijn volgens de theorie stabiele configuraties in het collectieve veld. Ze zijn niet opgeslagen in individuen, maar blijven bestaan als ‘trechters’ in het interactieveld tussen mensen. Dit verklaart ook hoe trauma’s van generatie op generatie kunnen worden doorgegeven (Yehuda et al., 2016), zonder dat er sprake is van directe genetische overdracht.

Een nieuwe kijk op de ziekte van Alzheimer

De theorie biedt ook een nieuw model voor de ziekte van Alzheimer. Neurodegeneratie wordt gezien als een toename van ‘ruis’ in het systeem, waardoor de trechters (attractoren) ondieper worden. Diepe trechters (oude herinneringen) eroderen langzamer dan ondiepe (recente herinneringen), wat het kenmerkende patroon van geheugenverlies bij Alzheimer verklaart. De theorie voorspelt ook dat een stabiele sociale omgeving (een consistent relationeel veld) het verlies van attractoren gedeeltelijk kan compenseren.

Conclusie

De Relationele Veldtheorie van het Geheugen nodigt uit tot een fundamentele verschuiving in hoe we naar herinneringen kijken. Het gaat niet langer om de vraag waar een herinnering is opgeslagen, maar onder welke omstandigheden een bepaalde denkconfiguratie stabiliseert. Het antwoord is onlosmakelijk relationeel en dynamisch, en staat haaks op de computer-metafoor die ons vakgebied zo lang heeft gedomineerd.

---

Uitgebreide Referentielijst

Hieronder volgt de volledige, geannoteerde bibliografie zoals gepresenteerd in de originele blog. De bronnen zijn gerangschikt per thema.

Fundamenteel: De crisis van het opslagparadigma

• Bartlett, F. C. (1932). Remembering: A Study in Experimental and Social Psychology. Cambridge University Press.
  • Toelichting: Bartletts klassieke werk toonde aan dat herinneren geen getrouwe reproductie is, maar een reconstructief proces, beïnvloed door culturele schema‘s. Zijn concept van ‘inspanning voor betekenis’ loopt vooruit op het voorspellende verwerkingskader. In de Relationele Veldtheorie worden zijn bevindingen geformaliseerd als aantrekkingsdynamiek: herinneren is een traject naar een stabiele configuratie, niet het uitlezen van een opgeslagen spoor.
• Schacter, D. L. (1996). Searching for Memory: The Brain, the Mind, and the Past. Basic Books.
  • Toelichting: Schacter’s toegankelijke synthese introduceert de ‘zeven zonden van het geheugen’. De Relationele Veldtheorie beweert dat al deze zonden kunnen worden afgeleid als specifieke gevallen van het dynamische raamwerk. Bijvoorbeeld ‘blokkade’ als tijdelijke gevangenschap in een concurrerende aantrekkingsput, en ‘misattributie’ als stroming naar een contextueel ongepaste aantrekker.
• van Gelder, T. (1998). The dynamical hypothesis in cognitive science. Behavioral and Brain Sciences, 21(5), 615-628.
  • Toelichting: Een baanbrekend theoretisch artikel dat stelt dat cognitieve systemen beter als dynamische systemen dan als digitale computers kunnen worden begrepen. Dit artikel vormt de filosofische en methodologische basis voor de afwijzing van opslag-en-ophaal in de Relationele Veldtheorie, ten gunste van aantrekkingsdynamiek.

Kern van het wiskundige en computationele raamwerk

• Friston, K. (2010). The free-energy principle: A unified brain theory? Nature Reviews Neuroscience, 11(2), 127-138.
  • Toelichting: Het vrije-energieprincipe is de belangrijkste wiskundige basis van de theorie. Friston stelt dat elk zelforganiserend systeem zijn vrije energie moet minimaliseren (een bovengrens voor verrassing of voorspellingsfout). De Relationele Veldtheorie past dit variationele raamwerk direct toe op geheugenvorming als gradiëntstroom in een verrassingslandschap.
• McWhinney, W. (1997). Paths of Change: Strategic Choices for Organizations and Society. Sage Publications.
  • Toelichting: McWhinney’s vier-operator model is de empirische en conceptuele basis voor de quaternion-toestandsruimte. De Relationele Veldtheorie neemt McWhinney’s taxonomie over als de vier dimensies van de cognitieve toestandsruimte, en formaliseert ze als quaternion-componenten.
• Kelso, J. A. S. (1995). Dynamic Patterns: The Self-Organization of Brain and Behavior. MIT Press.
  • Toelichting: Kelso’s fundamentele tekst over coördinatiedynamica verschaft de wiskundige taal van aantrekkers, bifurcaties en faseovergangen die de theorie gebruikt. De concepten van aantrekkingsbekkens, bifurcaties en het meten van geheugensterkte als bekken diepte zijn rechtstreeks ontleend aan Kelso’s raamwerk.

Relationele en ontwikkelingsfundamenten

• Bowlby, J. (1969). Attachment and Loss, Vol. 1: Attachment. Basic Books.
  • Toelichting: Bowlby’s gehechtheidstheorie levert het psychologische fenomeen dat de theorie formaliseert als ‘geïnternaliseerde samengestelde aantrekkers’. De theorie herinterpreteert het ‘intern werkend model’ als een configuratie die oorspronkelijk gestabiliseerd is door dominantiekoppeling tijdens een gevoelige periode, en vervolgens geconsolideerd is tot een autonome aantrekker.
• Vygotsky, L. S. (1978). Mind in Society: The Development of Higher Psychological Processes. Harvard University Press.
  • Toelichting: Vygotsky’s concept van de Zone van Naaste Ontwikkeling wordt in de theorie geformaliseerd als activering van uitgebreide modaliteit. Herhaalde scaffolding-interacties consolideren deze kortstondige toestanden tot stabiele aantrekkers.
• Porges, S. W. (2011). The Polyvagal Theory: Neurophysiological Foundations of Emotions, Attachment, Communication, and Self-Regulation. Norton.
  • Toelichting: Porges’ polyvagaal theorie levert het neurobiologische mechanisme voor de bewering dat co-regulatie de codeerefficiëntie vormgeeft. Co-regulatie van affect moduleert de coderingsdrempel in de Relationele Veldtheorie.

Collectief en cultureel geheugen

• Halbwachs, M. (1992). On Collective Memory. University of Chicago Press. (Origineel werk gepubliceerd 1950)
  • Toelichting: Halbwachs’ sociologische klassieker stelt dat geheugen inherent sociaal is. De theorie formaliseert zijn inzichten als ‘veldherinneringen’: stabiele configuraties van het multi-agent veld die de verwachte verrassing in de populatie minimaliseren. Culturele schema‘s, verhalen en normen zijn niet opgeslagen in individuen, maar blijven bestaan als aantrekkers in de collectieve dynamiek.
• Hutchins, E. (1995). Cognition in the Wild. MIT Press.
  • Toelichting: Hutchins’ etnografische studie toont aan dat cognitieve processen verdeeld zijn over personen, artefacten en omgevingen. Er is geen enkele ‘plaats’ waar de positie van het schip is opgeslagen; de positie ontstaat uit de coördinatie van meerdere agenten en hulpmiddelen. Dit ondersteunt direct de bewering dat geheugen niet gelokaliseerd is in individuele hersenen, maar een eigenschap is van het multi-agent veld.
• Yehuda, R., Daskalakis, N. P., Bierer, L. M., et al. (2016). Holocaust exposure induced intergenerational effects on FKBP5 methylation. Biological Psychiatry, 80(5), 372-380.
  • Toelichting: Deze empirische studie levert bewijs voor transgenerationele overdracht van trauma-gerelateerde biologische markers. De theorie interpreteert dergelijke bevindingen als overdracht van veldgeheugen: de traumatische gebeurtenis veranderde de veldtoestand van de oudergeneratie, en het kind ontwikkelde zich binnen dat gewijzigde veld, wat leidde tot consolidatie van een gewijzigde configuratie.

Klinische en neurodegeneratieve toepassingen

• Lupien, S. J., McEwen, B. S., Gunnar, M. R., & Heim, C. (2009). Effects of stress throughout the lifespan on the brain, behaviour and cognition. Nature Reviews Neuroscience, 10(6), 434-445.
  • Toelichting: Dit uitgebreide overzicht laat zien hoe acute en chronische stress de geheugencode moduleren. De theorie formaliseert dit als een toestandsafhankelijke coderingsdrempel.
• Stern, Y. (2009). Cognitive reserve. Neuropsychologia, 47(10), 2015-2028.
  • Toelichting: Sterns cognitieve reservetheorie stelt dat individuele verschillen in hersenstructuur en levenservaring de klinische expressie van Alzheimer pathologie moduleren. De theorie biedt een mechanistische verklaring: relationele veldkoppeling compenseert gedeeltelijk voor aantrekkererosie.

Slaap en consolidatie

• Stickgold, R. (2005). Sleep-dependent memory consolidation. Nature, 437, 1272-1278.
  • Toelichting: Stickgold’s overzicht synthetiseert bewijs dat slaap noodzakelijk is voor de consolidatie van herinneringen. De theorie biedt een mechanistische verklaring: tijdens de slaap wordt externe koppeling verwijderd, waardoor de dynamiek wordt gereduceerd tot pure gradiëntstroom – een formele verklaring waarom slaap het geheugen verbetert zonder een apart ‘consolidatiemechanisme’.
• Walker, M. P. (2017). Why We Sleep: Unlocking the Power of Sleep and Dreams. Scribner.
  • Toelichting: Walker’s toegankelijke synthese van de slaapwetenschap biedt een breed empirisch fundament voor de slaap-als-gradiëntstroom hypothese. Het boek is een uitstekend startpunt voor lezers die niet vertrouwd zijn met de slaap-literatuur.

Coderingsspecificiteit en context

• Tulving, E., & Thomson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80(5), 352-373.
  • Toelichting: Het coderingsspecificiteitsprincipe – dat succesvolle herinnering afhangt van de overlap tussen codeercontext en ophaalcontext – wordt door de theorie formeel afgeleid.
• Godden, D. R., & Baddeley, A. D. (1975). Context-dependent memory in two natural environments. British Journal of Psychology, 66(3), 325-331.
  • Toelichting: De klassieke demonstratie dat duikers die lijsten op het land leerden ze beter onthielden op het land dan onder water, en vice versa. Dit levert het empirische anker voor de definitie van contextueel geheugen in de Relationele Veldtheorie.

Synthese en toekomstige richtingen

• Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences, 36(3), 181-204.
  • Toelichting: Clark’s doelartikel biedt de meest toegankelijke introductie tot voorspellende verwerking – het computationele raamwerk dat de verrassings-minimalisatie dynamiek van de theorie ondersteunt. De theorie breidt dit uit van individuele hersenen naar dyadische en collectieve systemen.
• Siegel, D. J. (1999). The Developing Mind: How Relationships and the Brain Interact to Shape Who We Are. Guilford Press.
  • Toelichting: Siegel’s synthese van ontwikkelingspsychologie, gehechtheidstheorie en neurowetenschap introduceert het concept van ‘integratie’ als het kernmechanisme van geestelijke gezondheid. De theorie operationaliseert Siegel’s integratie als coherentie in het multi-agent veld.

---

Beyond the Archive: Memory as a Relational Field

For over half a century, the cognitive sciences have been guided by a powerful metaphor: the mind as a computer, and memory as a storage system. From the multi-store models of Atkinson and Shiffrin to the working memory framework of Baddeley and Hitch, this paradigm has yielded a rich taxonomy of memory systems and their neural correlates. Yet, as the authors of A Relational Field Theory of Memory argue with compelling force, a mounting accumulation of anomalies—context-dependent recall, reconstructive memory, relational modulation, distributed cognition, and the peculiar patterns of neurodegenerative disease—suggests that the storage metaphor has reached its explanatory limits.

This essay synthesises the core argument of the paper: that memory is not a repository of stored representations but a multi-scale dynamical process emerging from the stabilisation of cognitive configurations under the continuous imperative to minimise surprise. The framework is mathematically explicit, drawing on quaternion algebra and the free energy principle, yet its implications extend directly to developmental psychology, clinical practice, and the study of collective and cultural memory.

1. The Quaternion Agent: Four Operators of Cognition

The theory begins with a reconceptualisation of the individual cognitive system. Rather than a collection of modules or networks, the agent is represented as a unit quaternion—a point on the surface of a four-dimensional sphere. The four dimensions correspond to irreducible cognitive operators, derived from McWhinney’s (1997) Paths of Change:

• Unitary (wᵤ): Mythopoetic, meaning-making, field-sensing.
• Sensory (wₛ): Experiential, embodied, affective.
• Social (w_{So}): Relational, normative, communal.
• Mental (wₘ): Analytical, logical, systemic.

The constraint that the quaternion has unit length (|q| = 1) embodies a substantive claim: cognitive capacity is conserved. A shift toward one operator necessarily involves a compensatory redistribution across the others. This non-commutative structure—order matters—captures the developmental and clinical reality that the sequence of relational interactions shapes the final cognitive configuration.

For an intellectual audience, the value here is not in the algebra per se but in what it affords: a state space where learning becomes trajectory, memory becomes attractor, and the same equations govern an individual’s recall, a dyadic interaction, and the coherence of an entire community.

2. Memory as Surprisal Minimisation

The engine of memory in this framework is the free energy principle (Friston, 2010). Any adaptive system—biological, cognitive, or social—must minimise the surprise (or equivalently, maximise the evidence) associated with its sensory states. The paper operationalises this by defining a surprisal functional S(q, o) that measures the prediction error between the current cognitive configuration q and an observation o.

A memory trace is formally defined as a stable configuration q* that minimises expected surprisal within a specific context class C. Memory strength is not a property of a stored trace but the depth of the attractor basin surrounding q: the magnitude of the gradient that pulls the system back toward that configuration after perturbation. Contextual memory is precisely this dependency: q is a memory in context C but not necessarily in context C’.

The governing dynamics are captured by a master stochastic differential equation with four forcing terms: the intrinsic surprisal-minimising gradient (learning drive), dominance coupling to a reference agent (social scaffolding), extended modality activation (temporary capacity expansion, as in Vygotsky’s Zone of Proximal Development), and interference from a composite field (relational misattunement).

What makes this framework powerful is that encoding, consolidation, and recall are no longer distinct processes requiring separate mechanisms. Encoding occurs when an observation generates surprisal exceeding a state-dependent threshold, initiating a trajectory toward a new attractor. Consolidation is the stabilisation of that attractor against perturbation—achieved most efficiently during sleep, when external coupling is removed and the system performs pure stochastic gradient descent. Recall is re-entry into an existing attractor’s basin, triggered by a partial match between current observation and the context class. Because the system flows toward the attractor along the gradient, and because noise and contextual changes modify that flow, recall is inherently reconstructive (Bartlett, 1932)—a dynamical completion, not a readout.

3. The Relational Field: Dyadic and Collective Memory

Perhaps the most radical departure from standard cognitive neuroscience is the claim that memory is not an individual property. The paper formalises dyadic interaction as the formation of a composite system. Two agents—a more differentiated ‘reference’ and a less differentiated ‘respondent’—couple to form a composite state q_comp. Three mechanisms shape memory within this composite:

• Dominance (λᴅ): The reference agent structures the surprisal landscape of the respondent. In healthy development, this coupling starts high (complete dependence) and decreases with differentiation—a formalisation of Vygotskian scaffolding.
• Extended Modality (γ): Within the relationship, the respondent can temporarily access configurations beyond their stable range (the Zone of Proximal Development). Repeated interaction consolidates these transient states into autonomous attractors.
• Interference (κ): Divergence between the individual and the composite consumes cognitive energy. Chronic misattunement increases κ without consolidation, producing functionally unstable configurations—a formal model of anxious attachment (Bowlby, 1969).

The core claim here is that internal working models (Main, Kaplan & Cassidy, 1985)—the stable relational schemas of attachment theory—are internalised composite attractors. They were encoded during periods of maximal dominance coupling (infancy and early childhood), when the relational field entirely structured the child’s surprisal landscape. This explains the otherwise puzzling finding that early relational experiences have disproportionate, lifelong effects on memory and affect regulation.

Extending this to the collective level, the theory defines a field state q_field as the weighted mean of individual configurations across a population, with coherence C(t) measuring the degree of alignment. Field memories—cultural schemas, institutional norms, shared narratives—are stable configurations of this collective dynamics. They are not stored in any individual but persist as attractors in the interaction field. Most strikingly, memory can be transmitted without direct experience: an individual developing within a stable field is biased toward its configuration via dominance coupling, and over time this external pressure consolidates as an intrinsic attractor. This provides a formal mechanism for phenomena ranging from cultural transmission to transgenerational trauma (Yehuda et al., 2016).

4. A Temporal Hierarchy

The framework explicitly recognises that memory operates across nested temporal scales, from milliseconds to generations. Crucially, the same dynamical equations govern each scale; only the parameters differ:

Scale	Time Range	Dominant Mechanism	Example
Micro	ms–sec	EM activation, working memory	Holding a phone number
Meso	min–days	Surprise-driven encoding, sleep consolidation	Learning a procedure
Developmental	months–years	Composite coupling, internalisation	Attachment patterns
Cultural	decades–generations	Field coherence, institutional coupling	Shared historical schemas

Developmental transitions correspond to bifurcations—qualitative changes in the attractor landscape. The age-dependence of memory is captured by a decaying time constant: early memories are encoded more slowly but consolidated more deeply. This explains the paradox of infantile amnesia (very early experiences have lasting effects but are not consciously recalled) and directly predicts the preservation of remote memory in Alzheimer’s disease.

5. Clinical Application: Alzheimer’s as Attractor Erosion

The theory generates a coherent, mechanistically explicit account of Alzheimer’s disease. Neurodegeneration is modelled as a progressive increase in noise amplitude σ² combined with a decrease in attractor depth M(q*). The degradation follows a simple differential equation: dM/dt ∝ –σ²·M. Deep attractors (remote memories) erode more slowly than shallow attractors (recent memories).

This yields five falsifiable predictions:

1. New memory failure: Increased noise raises the effective encoding threshold, preventing new attractor formation.
2. Remote memory preservation: Deep attractors erode slowly, explaining the characteristic temporal gradient of memory loss.
3. Context dependence: As attractor basins narrow, recall becomes possible only in environments closely matching the original encoding context.
4. Regression: As recent attractors erode, the system regresses to earlier, deeper stable configurations (e.g., childhood language, early relationship patterns).
5. Relational stabilisation: Consistent social environments (stable q_comp, high field coherence C) partially compensate for attractor erosion, consistent with the cognitive reserve literature (Stern, 2009).

The intervention architecture follows directly: attractor maintenance (reminiscence therapy, familiar music), noise reduction (stress reduction, sleep, anti-inflammatory interventions), and relational field stabilisation (consistent care environments, reduced context variability).

6. Computational Implementation and Empirical Agenda

The framework has been operationalised in SWARP (Shared Wisdom And Relational Platform), a multi-agent computational system that implements the field dynamics described above. Agents are represented as quaternion configurations derived from empirically validated assessments (Human Design, RIASEC/O*NET, Enneagram typology). Memory is not stored as discrete records but as the accumulated trajectory of state-space evolution. Retrieval is pattern-completion: partial matches to past contexts trigger flow toward the corresponding attractor.

The research programme outlined in the paper is extensive and falsifiable:

• Neuroscience: BOLD signal synchrony at encoding should predict recall strength better than activation magnitude at any single locus (cf. Paller & Wagner, 2002). EEG coherence should serve as a proxy for field coherence C(t).
• Developmental psychology: Longitudinal measurement of λᴰ trajectories from birth to 24 months using interaction coding should predict attachment outcomes.
• Clinical trials: Randomised trials of high-coherence versus low-coherence care environments for Alzheimer’s patients should test the relational stabilisation hypothesis.
• Computational simulation: Agent-based modelling should map the phase diagram of coherence emergence as a function of N, λ_field, and noise.

Conclusion: Memory as Trajectory, Not Archive

The Relational Field Theory of Memory invites a fundamental shift in perspective. Memory is not a library of stored traces but a set of stable trajectories in a cognitive state space—attractor basins shaped by the history of surprisal-minimising interactions with the environment and with other agents. This is not merely a mathematical reformulation of existing concepts. It provides a unified mechanism for encoding, consolidation, and recall; a formal account of relational and cultural transmission; a predictive model of Alzheimer’s disease; and an empirically falsifiable research programme spanning multiple scales and disciplines.

For the intellectual reader, the central takeaway is this: the next generation of memory science will likely not ask “where is this memory stored?” but rather “under what conditions does this configuration stabilise?” The answer, as this framework makes clear, is irreducibly relational, deeply dynamical, and fundamentally at odds with the computational metaphor that has served us so well—and now, perhaps, outlived its usefulness.

---

Extended Annotated Bibliography

The following annotated bibliography is designed to equip the reader with the foundational and contemporary sources necessary to explore the concepts of the Relational Field Theory of Memory in depth. Each entry explains the source’s core contribution, its relevance to the framework, and how it can be used for further investigation.

Foundational: The Crisis of the Storage Paradigm

Bartlett, F. C. (1932). Remembering: A Study in Experimental and Social Psychology. Cambridge University Press.
Annotation: Bartlett’s classic work demonstrated that memory is not a faithful reproduction of past events but a reconstructive process shaped by cultural schemas and personal interests. His concept of ‘effort after meaning’ prefigures the predictive processing framework. In the Relational Field Theory, Bartlett’s findings are formalised as attractor dynamics: recall is a trajectory toward a stable configuration under noisy and contextual conditions, not a readout of a stored trace. For further exploration: Read Bartlett’s original ‘War of the Ghosts’ experiment to understand the empirical basis for reconstructive memory.

Schacter, D. L. (1996). Searching for Memory: The Brain, the Mind, and the Past. Basic Books.
Annotation: Schacter’s accessible synthesis of memory research introduces the ‘seven sins of memory’ (transience, absent-mindedness, blocking, misattribution, suggestibility, bias, persistence). The Relational Field Theory claims that all seven sins can be derived as limiting cases of the dynamical framework—for example, ‘blocking’ as temporary entrapment in a competing attractor basin, ‘misattribution’ as flow toward a contextually inappropriate attractor. For further exploration: Use Schacter’s taxonomy as a checklist to test whether a dynamical model can reproduce each ‘sin’ under specific parameter regimes.

van Gelder, T. (1998). The dynamical hypothesis in cognitive science. Behavioral and Brain Sciences, 21(5), 615-628.
Annotation: A landmark theoretical paper arguing that cognitive systems are best understood as dynamical systems rather than digital computers. Van Gelder contrasts the ‘computational metaphor’ with the ‘dynamical hypothesis’ using the example of a Watt governor. This paper provides the philosophical and methodological foundation for the Relational Field Theory’s rejection of storage-and-retrieval in favour of attractor dynamics. For further exploration: Read van Gelder alongside Fodor’s critique of dynamical systems to understand the ongoing debate.

Core Mathematical and Computational Framework

Friston, K. (2010). The free-energy principle: A unified brain theory? Nature Reviews Neuroscience, 11(2), 127-138.
Annotation: The free energy principle is the single most important mathematical foundation for the Relational Field Theory. Friston argues that any self-organising system that maintains its integrity over time must minimise its free energy (an upper bound on surprisal or prediction error). The paper derives the dynamics of perception, learning, and action from this single imperative. The Relational Field Theory adopts Friston’s variational framework directly, applying it to memory formation as gradient flow on a surprisal landscape. For further exploration: Start with Friston’s more accessible ‘tutorial’ papers on active inference before tackling the mathematical derivations in this target article.

McWhinney, W. (1997). Paths of Change: Strategic Choices for Organizations and Society. Sage Publications.
Annotation: McWhinney’s four-operator model (Unitary, Sensory, Social, Mental) is the empirical and conceptual basis for the quaternion state space. Originally developed for organisational change and conflict resolution, the model posits that human cognition operates through four irreducible modes, each with its own logic, language, and criteria for truth. The Relational Field Theory adopts McWhinney’s taxonomy as the four dimensions of the cognitive state space, formalising them as quaternion components. For further exploration: Apply McWhinney’s four operators to analyse a familiar organisational conflict; note how each operator frames problems and solutions differently.

Kelso, J. A. S. (1995). Dynamic Patterns: The Self-Organization of Brain and Behavior. MIT Press.
Annotation: Kelso’s foundational text on coordination dynamics provides the mathematical language of attractors, bifurcations, and phase transitions that the Relational Field Theory employs. Empirical phenomena such as finger-tapping coordination (the Haken–Kelso–Bunz model) demonstrate how stable patterns emerge from non-linear dynamics. The theory’s concept of attractor basins, bifurcations during developmental transitions, and the measurement of memory strength as basin depth are directly borrowed from Kelso’s framework. For further exploration: Reproduce the classic ‘phase transition’ experiment using the freely available coordination dynamics software to see attractor switching in real time.

Relational and Developmental Foundations

Bowlby, J. (1969). Attachment and Loss, Vol. 1: Attachment. Basic Books.
Annotation: Bowlby’s attachment theory provides the psychological phenomenon that the Relational Field Theory formalises as ‘internalised composite attractors.’ Bowlby argued that the infant’s early relationship with the primary caregiver is internalised as an ‘internal working model’—a stable cognitive-affective schema that shapes all subsequent relationships. The Relational Field Theory reinterprets this model as a configuration q* originally stabilised by dominance coupling (λᴅ ≈ 1) during a sensitive period, then consolidated into an autonomous attractor. For further exploration: Read Bowlby alongside Main et al. (1985) to understand how the Adult Attachment Interview measures internal working models via narrative coherence.

Vygotsky, L. S. (1978). Mind in Society: The Development of Higher Psychological Processes. Harvard University Press.
Annotation: Vygotsky’s concept of the Zone of Proximal Development (ZPD)—the difference between what a learner can do independently and with guidance—is formally modelled in the Relational Field Theory as Extended Modality (EM) activation. Within the composite (teacher–learner), the learner can temporarily access configurations q that are outside their stable attractor basin. Repeated scaffolding interactions consolidate these transient states into stable attractors. For further exploration: Design a simple teaching intervention and map the predicted EM activation parameters (γ_X) for each operator dimension.

Porges, S. W. (2011). The Polyvagal Theory: Neurophysiological Foundations of Emotions, Attachment, Communication, and Self-Regulation. Norton.
Annotation: Porges’ polyvagal theory provides the neurobiological mechanism for the Relational Field Theory’s claim that co-regulation shapes encoding efficacy. The theory distinguishes three neural circuits (ventral vagal, sympathetic, dorsal vagal) that map onto states of safety, mobilisation, and shutdown. Co-regulation of affect—facilitated by prosodic vocalisation, facial expression, and gesture—modulates the encoding threshold θ_e(q) in the Relational Field Theory. For further exploration: Use Porges’ concept of ‘neuroception’ (the nervous system’s automatic detection of safety or threat) to understand why the dominance coupling parameter λᴅ varies with relational context.

Collective and Cultural Memory

Halbwachs, M. (1992). On Collective Memory. University of Chicago Press. (Original work published 1950)
Annotation: Halbwachs’ sociological classic argues that memory is inherently social: individuals remember as members of groups, and the frameworks of collective memory shape what is retained and forgotten. The Relational Field Theory formalises Halbwachs’ insights as field memories—stable configurations q_field* of the multi-agent field that minimise expected surprisal across the population. Cultural schemas, narratives, and norms are not stored in any individual but persist as attractors in the collective dynamics. For further exploration: Apply Halbwachs’ method of analysing ‘les cadres sociaux de la mémoire’ (the social frameworks of memory) to a contemporary case, such as how different national communities remember a shared historical event.

Hutchins, E. (1995). Cognition in the Wild. MIT Press.
Annotation: Hutchins’ ethnographic study of navigation aboard naval vessels demonstrates that cognitive processes are distributed across persons, artefacts, and environments. There is no single ‘place’ where the ship’s position is stored; rather, position emerges from the coordination of multiple agents and tools. This directly supports the Relational Field Theory’s claim that memory is not localised in individual brains but is a property of the multi-agent field. For further exploration: Observe a team performing a complex task (e.g., a hospital emergency room, a software development team) and map the distributed cognitive processes using Hutchins’ methodology.

Yehuda, R., Daskalakis, N. P., Bierer, L. M., et al. (2016). Holocaust exposure induced intergenerational effects on FKBP5 methylation. Biological Psychiatry, 80(5), 372-380.
Annotation: This empirical study provides evidence for transgenerational transmission of trauma-related biological markers. Offspring of Holocaust survivors showed methylation changes in the FKBP5 gene, which regulates glucocorticoid receptor sensitivity. The Relational Field Theory interprets such findings as field memory transmission: the traumatic event altered the field state q_field of the parent generation, and the child developed within that altered field, leading to consolidation of a modified configuration q* via dominance coupling. For further exploration: Distinguish between epigenetic inheritance (direct germline transmission) and field transmission (social/relational) as alternative or complementary mechanisms.

Clinical and Neurodegenerative Applications

Lupien, S. J., McEwen, B. S., Gunnar, M. R., & Heim, C. (2009). Effects of stress throughout the lifespan on the brain, behaviour and cognition. Nature Reviews Neuroscience, 10(6), 434-445.
Annotation: This comprehensive review documents how acute and chronic stress modulate memory encoding. Stress hormones (glucocorticoids) enhance or impair encoding depending on timing and intensity. The Relational Field Theory formalises this as a state-dependent encoding threshold θ_e(q) = θ₀ + α·w_M(t) – β·w_S(t). Dysregulation (extreme w_S) raises the threshold, impairing encoding—consistent with the finding that acute stress impairs hippocampal-dependent memory formation. For further exploration: Use this review to derive additional predictions: e.g., chronic stress should systematically shift the attractor landscape toward configurations with low w_U (meaning-making) and high w_S (sensory vigilance).

Stern, Y. (2009). Cognitive reserve. Neuropsychologia, 47(10), 2015-2028.
Annotation: Stern’s cognitive reserve theory posits that individual differences in brain structure and life experience (education, occupation, social engagement) modulate the clinical expression of Alzheimer’s pathology. The Relational Field Theory provides a mechanistic account: relational field coupling r_comp(t) partially compensates for attractor erosion dM/dt. Consistent social environments (stable q_comp, high field coherence C) maintain functional memory longer than socially impoverished environments, independent of neural pathology. For further exploration: Design a clinical trial that manipulates field coherence C (e.g., structured group reminiscence therapy vs. individual cognitive training) and measures the rate of attractor erosion M(q*) over time.

Sleep and Consolidation

Stickgold, R. (2005). Sleep-dependent memory consolidation. Nature, 437, 1272-1278.
Annotation: Stickgold’s review synthesises evidence that sleep, particularly slow-wave sleep and REM sleep, is necessary for the consolidation of declarative and procedural memories. The Relational Field Theory provides a mechanistic explanation: during sleep, external coupling λᴅ ≈ 0 and new EM activations γ ≈ 0, reducing the dynamics to pure gradient flow dq/dt = –∇ₑ[S] + η(t). This is precisely the condition for stochastic gradient descent toward a fixed point—a formal account of why sleep enhances memory without requiring a separate ‘consolidation mechanism.’ For further exploration: Use this paper to generate predictions about the effect of sleep deprivation on attractor depth M(q*), operationalised as resistance to interference from subsequent learning.

Walker, M. P. (2017). Why We Sleep: Unlocking the Power of Sleep and Dreams. Scribner.
Annotation: Walker’s accessible synthesis of sleep science provides a broad empirical foundation for the sleep-as-gradient-flow hypothesis. The book reviews evidence for sleep-dependent memory consolidation, the role of sleep spindles and slow oscillations, and the effects of sleep deprivation on cognitive function. While not technical, it is an excellent entry point for readers unfamiliar with the sleep literature. For further exploration: Use Walker’s practical recommendations (e.g., consistent sleep schedule, temperature reduction) as experimental manipulations to test the predicted effects on attractor consolidation rates.

Encoding Specificity and Context

Tulving, E., & Thomson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80(5), 352-373.
Annotation: The encoding specificity principle—that successful recall depends on the overlap between encoding context and retrieval context—is one of the most robust findings in memory research. The Relational Field Theory derives this principle formally: the probability of recall is P(recall | o) ∝ exp(–|o – ō|² / 2σ²_context) · M(q), a Gaussian proximity function around the centroid of the context class C*. *For further exploration:* Design an experiment that parametrically varies context similarity (e.g., using virtual reality environments) and fits the Gaussian decay constant σ_context to test the model’s quantitative prediction.

Godden, D. R., & Baddeley, A. D. (1975). Context-dependent memory in two natural environments. British Journal of Psychology, 66(3), 325-331.
Annotation: The classic demonstration that divers who learned lists on land recalled them better on land than underwater, and vice versa. This provides the empirical anchor for Definition 3.3 (Contextual Memory) in the Relational Field Theory: a configuration q* is a memory in the context in which it was formed, not universally. For further exploration: Replicate the essential finding in a controlled laboratory setting using immersive virtual reality to obtain the precise parameters for the Gaussian proximity function.

Synthesis and Future Directions

Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences, 36(3), 181-204.
Annotation: Clark’s target article and the accompanying peer commentary provide the most accessible introduction to predictive processing—the computational framework that underpins the Relational Field Theory’s surprisal-minimisation dynamics. Clark argues that the brain is a prediction engine that continuously generates and updates a generative model of its sensory inputs. The Relational Field Theory extends this from individual brains to dyadic and collective systems. For further exploration: Work through Clark’s ‘hierarchical prediction’ diagrams to understand how surprisal minimisation can be implemented in neural networks.

Siegel, D. J. (1999). The Developing Mind: How Relationships and the Brain Interact to Shape Who We Are. Guilford Press.
Annotation: Siegel’s synthesis of developmental psychology, attachment theory, and neuroscience introduces the concept of ‘integration’ as the core mechanism of mental health. The Relational Field Theory operationalises Siegel’s integration as coherence C(t) in the multi-agent field. High coherence—the alignment of individual configurations around a shared attractor—corresponds to mental health and relational well-being; low coherence corresponds to disorganisation and pathology. For further exploration: Map Siegel’s nine domains of integration (e.g., consciousness, memory, state, relational) onto the four operator dimensions and the coherence metric C(t).