Introduction: The Elusive Object
Intelligence research has been cycling through the same pattern: a paradigm is established, a new generation finds it limited, develops a more adequate framework, and then finds that the new framework has reproduced the mistake it was trying to escape. The psychometric tradition measured one kind of performance and called it intelligence. The cognitive science tradition identified the mechanisms behind that performance and discovered they were unstable across contexts. The evolutionary psychology tradition recovered ecological validity and found itself fragmenting intelligence into a list of modules without a unifying account. Howard Gardner insisted on plurality and produced eight or nine intelligences whose boundaries are contested and whose ontological basis remains obscure. Emotional and social intelligence traditions added further domains. Distributed cognition showed that intelligence extends beyond individual skulls into tools, environments, and communities. Ecological psychology showed that perception and action are already intelligent before any symbolic processing occurs.
None of these discoveries was wrong. Each captured something real. But each also ended in the same place: a more complex picture of the same wrong object. The wrong object is intelligence understood as a property that organisms have, a capacity, a faculty, a quantity, or a set of cognitive operations that can be identified, measured, and compared independently of the conditions in which it operates.
Living Value Theory proposes a different starting point. Intelligence is not a property of an organism. It is a quality of an organism's ongoing relationship with the stakes of its situation across the five irreducible mediations of human life: embodiment, being-with, dwelling, multimateriality, and multisymbolism. Every human being develops intelligence across all five mediations. The interesting question is not which mediations a person possesses, all five are always present, but what mediational profile develops: which forms of stakes-responsiveness become richly developed, which remain developing, and which combinations produce the distinctive hybrid excellences we call expertise or talent. Each mediation has its own form of intelligence, its own temporality, its own characteristic stakes. Intelligence in the full sense is distributed across all five, operates across five levels of recursivity from absorbed pre-reflective coping to explicit meta-reflection, and requires what LVT calls recursive fluidity: the capacity to move between levels as the situation demands, escalating when misalignment requires attention and descending back into seamless coordination when it does not.
The argument of this article is that the history of intelligence research has repeatedly discovered fragments of this larger phenomenon without recognising their common structure. Psychometrics discovered symbolic intelligence, specifically in its abstract L4 register. Gardner discovered that intelligence is plural. Anthropology documented dwelling, embodied, and relational expertise without calling it intelligence. Ecological psychology recovered situated responsiveness at L1 and L2. Distributed cognition recovered collective mediation. LVT's contribution is to show that these are not disconnected findings but expressions of a single underlying architecture of living mediation, and to provide the theoretical stopping rule that Gardner's account, for all its empirical insight, was unable to supply.
I. From One Intelligence to Many: The Incomplete Revolution
The modern history of intelligence research begins with a convergence and ends with an impasse. Charles Spearman's early twentieth-century discovery that performance across diverse cognitive tests is positively correlated, that people who perform well on verbal tasks tend to perform well on spatial and numerical tasks too, seemed to establish that a single general factor, g, underlies all cognitive performance. This was a key empirical finding. The positive manifold is real and replicable. The question is what it means.
The psychometric tradition concluded that positive correlation across tasks reflects a unitary underlying capacity, a latent trait that causes performance across domains. But correlation does not establish a unitary cause. The positive manifold can equally be explained by a common feature of all the tasks selected for measurement: they are school-like tasks, requiring sustained performance on decontextualised symbolic problems under conditions of managed calm. The g factor may not reflect general intelligence. It may reflect a general tendency to perform well on a specific kind of measurable symbolic task that industrial education systems found useful to assess. What psychometrics discovered was not general intelligence but measurable intelligence, that subset of intelligence most hospitable to symbolic quantification.
Howard Gardner's theory of multiple intelligences was the most ambitious attempt to correct this. His 1983 Frames of Mind correctly identified that the abilities measured by standard tests are a small subset of the capacities that matter for human life. The bodily intelligence of a dancer, the interpersonal intelligence of a therapist, the musical intelligence of a composer, and the spatial intelligence of a navigator are real forms of human excellence that standard tests ignore entirely. Gardner was right about plurality and wrong about the stopping rule. By what principle does the list of intelligences end? Gardner added an eighth intelligence (naturalistic) in 1999 and considered further candidates without a principled criterion for inclusion or exclusion. The framework cannot distinguish between a form of intelligence and a sophisticated form of skill.
Here lies the most important distinction between Gardner's project and LVT's. Gardner discovered plurality empirically: he looked at the distribution of human capacities, identified clusters that resisted reduction to a single factor, and proposed that each cluster represented a distinct intelligence. LVT derives plurality theoretically. The five mediational intelligences are not discovered by mapping human talents. They are derived from the irreducible structure of human existence itself, from the five mediations through which all human life is constituted. There are five mediational forms of intelligence because there are five irreducible mediations, and the number is not arbitrary. This gives LVT the stopping rule Gardner lacked: new candidates for intelligence are assessed not by empirical clustering but by asking whether they represent a distinct mediation or a combination of existing ones.
Gardner's further mistake, shared with the entire tradition he challenged, was to treat intelligences as properties of individuals that some possess more or less of. LVT replaces this with a different claim: every normal adult human develops intelligence across all five mediations. The question is what profile develops. No mediation is reducible to another. All five are always present. What varies is the richness of development in each, shaped by developmental environment, sustained engagement, cultural emphasis, and the specific fields of activity in which someone has invested sustained attention. Intelligence is constituted in the relationship between organism and field, not deposited in the organism as a stable property.
II. Intelligence as Stakes-Responsiveness Across the Mediations
The foundational claim of a Living Value Theory of intelligence is this: intelligence is successful stakes-responsiveness across the five mediations of living. It is the capacity to be appropriately oriented by what actually matters in a situation, to feel the real pressures across all five mediational dimensions, respond to what is actually pressing, escalate when escalation is warranted, maintain coordination when it is not, invent when invention is necessary, and return to baseline when the pressure has resolved.
This definition has several precise implications.
First, intelligence is relational and situational, not a fixed property of a cognitive system. The same organism may be highly responsive to stakes in one mediational configuration and in an earlier stage of development in another. The master surgeon with extraordinary embodied and material intelligence may still be developing dwelling intelligence. This is not because the surgeon possesses two different amounts of two different intelligences. It is because intelligence is a quality of the organism-in-situation, and different situations involve different mediations with different stakes. All five mediations are always present in any situation, but their relative salience, and the depth of the practitioner's attunement to them, varies enormously.
Second, stakes-responsiveness operates before explicit analysis. The most powerful forms of intelligence are often those that register what matters before a propositional account of what is happening has been formulated. The experienced nurse who senses that a patient is deteriorating before the monitors show it, the negotiator who feels the shift in a room before anyone has spoken, the craftsperson whose hands know the material is ready before conscious inspection confirms it, all are exercising intelligence at L2, in the space before the analytical layer has been reached. This is not inferior intelligence. It is intelligence closer to the stakes.
Third, there is no domain-general intelligence independent of mediational activity. What appears to be general intelligence is better understood as a general capacity for L4 symbolic manipulation that appears general because the tasks used to measure it all come from the same narrow mediational domain. This does not mean there is no such thing as sophisticated hybrid coordination across mediations, there is, and it can be highly developed. But hybrid coordination is not a meta-intelligence floating above the five mediations. It is the achievement of integrating them, and that integration is itself a form of development that requires exposure to real stakes across all five.
Fourth, intelligence includes the capacity to return to baseline. The mouse in the children’s book The Gruffalo is the paradigm case of this. Once the predators were bested with superior recursive range and fluidity, the mouse does not remain at higher recursive levels. It returns to L1, it eats its nut, calmly, without vigilance. This is not the absence of intelligence. It is intelligence completing its cycle. The capacity to escalate when stakes demand escalation and to release when they do not is itself a form of stakes-responsiveness. It reflects accurate reading of the field: the stakes have resolved; the appropriate response now is absorption in ordinary life ()
III. The Three Recursive Domains
A second dimension of the LVT framework cuts across all five mediations and determines what kind of stakes-responsiveness a situation requires. Living Value Theory distinguishes three recursive domains: the non-recursive, the self-recursive, and the inter-recursive. Recognising which recursive domain a situation belongs to is itself a form of intelligence.
The non-recursive domain encompasses phenomena that do not respond to being observed, theorised, or categorised. Gravity does not change because someone believes differently about it. The load-bearing properties of a beam are indifferent to the engineer's emotional state. Weather, soil chemistry, the progression of disease through a population: these are non-recursive processes. Intelligence in the non-recursive domain requires accurate, disciplined attention to what is actually happening, the capacity to read the system's actual behaviour with fidelity. Prediction is relatively tractable here: the same antecedent conditions reliably produce the same outcomes.
The self-recursive domain encompasses processes in which the agent is itself an object of its own influence. Learning, habit formation, skill development, self-discipline, and identity formation are self-recursive: what I believe about my own capacities affects those capacities. Intelligence here requires the ability to model oneself clearly and to use that self-model to develop new responses. Prediction is moderate: self-recursive processes are partially predictable because habits and dispositions have inertia, but change creates uncertainty.
The inter-recursive domain encompasses situations where two or more agents are mutually modelling each other, and where each agent's behaviour is partly constituted by their model of the other's model of them. Friendship, therapy, teaching, parenting, negotiation, and politics are all inter-recursive. Intelligence here requires the capacity to sustain mutual modelling, to hold one's own perspective while tracking the other's. Prediction is poor: inter-recursive domains resist prediction systematically, because symbolic projections into their future alter the very coordination they claim to forecast. The therapist who tells a patient what they will feel next creates the feeling. The politician who announces what voters will do changes what voters do.
This temporal asymmetry of prediction is one of the most important but least discussed features of intelligence. The engineer's dwelling and material intelligence can generate reliable predictions about load and stress. The physician's knowledge can predict the progression of a disease with moderate accuracy. The therapist's relational intelligence cannot predict the course of a therapeutic relationship, because the relationship is itself inter-recursively coupled to any predictions made within it. This does not make relational intelligence less intelligent. It makes it differently intelligent, responsive to a domain that is constitutively unpredictable. Many of the most consequential forms of inter-recursive intelligence operate below the level of explicit awareness: the parent who regulates their child's developing emotional life through embodied cues and timing, the therapist whose presence creates conditions of safety, the teacher who holds a classroom in a state of engaged curiosity. These are L1 and L2 inter-recursive intelligence, genuine and high-performing, carrying more of the weight of the interaction than any explicit analytical capacity does.
IV. Recursivity Levels and Recursive Fluidity
One of the most consequential omissions in existing intelligence theory is the failure to distinguish between different levels of recursive engagement. Most frameworks implicitly collapse L1 through L5 into a single dimension, treating the explicit, reflective, analytical performance of intelligence tests as the paradigm case. LVT shows that this collapses fundamentally different forms of intelligent engagement.
An essential clarification before proceeding: Levels L1 and L2 are not restricted to the first four mediations. They characterise any form of coordination operating at those recursivity levels. L3, L4, and L5 are forms of multisymbolism applied reflexively to the other four. When a surgeon reflects explicitly on what their hands are doing, they are applying symbolic articulation (L3) to embodied coordination (L1). When a farmer develops abstract principles of soil management from years of practice, they are applying symbolic abstraction (L4) to dwelling intelligence. When an anthropologist questions whether their theoretical framework is distorting their understanding of a relational field, they are applying meta-reflective symbolism (L5) to relational experience. The levels are not separate intelligence faculties. They are degrees of symbolic recursion applied to the living mediational field.
L1 intelligence is absorbed coping: the seamless, pre-reflective exercise of competence in a domain where coordination is flowing without friction. The skilled musician playing without deliberation, the experienced driver navigating complex traffic without conscious analysis, the practised cook moving through a preparation sequence with the body knowing what to do, all are at L1. L1 intelligence is the form that psychometrics most systematically misses because it produces no reportable output. It is also the telos of learning. The goal of developing any form of intelligence is not to remain at L4, explicitly analysing what one is doing. It is to develop L1 competence: to descend from deliberate practice back into absorbed, effortless coordination. The surgeon who still has to think about each suture is not yet fully expert. The expert has descended: the knowledge has become embodied, and the body knows what to do. This principle, that genuine learning culminates in effortless coordination rather than explicit knowledge, is one of the places where current psychology has it backwards.
L2 intelligence is the capacity for felt misalignment: the ability to detect that something is off before it can be named or analysed. L2 is not pre-intelligent uncertainty. It is one of the most acute and valuable forms of intelligence available to a practitioner. The experienced clinician who feels that a patient's recovery is not proceeding as expected, before any specific indicator has changed; the teacher who senses that a student's apparent compliance is covering disengagement; the fieldworker who registers that the social dynamics of a situation have shifted, all are exercising L2 intelligence. L2 is the threshold at which coordination becomes available for change. Without the capacity for felt misalignment, no learning, no adaptation, and no repair is possible.
L3 intelligence is the capacity for accurate symbolic articulation: the ability to name, describe, share, and act on what has been felt at L2. It is not the same as verbal fluency or analytical sophistication. It is the capacity to find the right articulation, the description that captures the felt misalignment and opens it to response. The nurse who can tell the senior clinician exactly what feels wrong with a patient in terms that allow effective response; the teacher who can name for a student the specific nature of their difficulty in a way that makes it recognisable and addressable; the parent who can articulate what is happening in a difficult interaction: all are exercising L3 intelligence. L3 is where inter-recursive intelligence most often operates, because communicating a felt sense to another person in a way they can receive is itself a demanding inter-recursive achievement.
L4 intelligence is the capacity for abstraction, generalisation, and systematic thinking: the ability to move beyond the specific case to the pattern, the principle, the concept. Standard intelligence tests measure this most directly, and educational systems most explicitly develop it. It is important. But L4 is symbolism applied reflexively to the other mediations, not a separate faculty. There is L4 applied to embodied coordination: the master teacher of a physical discipline who can articulate principles of movement transmissible across students. There is L4 applied to dwelling: the expert farmer who has developed abstract principles of soil management applicable across varying conditions. There is L4 applied to relations: the experienced therapist who has developed understandings of relational dynamics that can be brought to bear on specific cases. L4 is a level of reflexive symbolic engagement that can be applied across all five mediational domains.
L5 intelligence is meta-recursive repositioning: the capacity to step outside the framework one is using and ask what kind of thinking is occurring, what assumptions structure the approach, and whether a different framing would serve better. L5 is what allows intellectual innovation: not the application of an existing framework to a new problem but the recognition that the existing frameworks are generating false problems. It is also what allows the highest forms of intellectual honesty: the capacity to notice when one's own theoretical commitments are producing systematic blind spots and to work against them in the service of better understanding.
Recursive fluidity is the capacity most completely missing from existing intelligence accounts. It is the capacity to move between these levels in ways that preserve and repair coordination. Some people excel at detecting misalignment (L2). Some at articulating (L3). Some at generalising (L4). Some at critique and repositioning (L5). Some at descending back into practice (L1). These are different forms of intelligence, and they can be developed to different degrees independently. The person who can detect misalignment exquisitely at L2 but cannot articulate it at L3 is in a different situation from the person who can articulate brilliantly at L3 but cannot see the abstract pattern at L4, and different again from the person who produces elegant L4 theory but cannot feel when the theory has diverged from the mesocosm it describes. Recursive fluidity, the capacity to move through the full arc from absorbed L1 coordination through L2 sensing, L3 articulation, L4 analysis, and L5 repositioning, and back down to L1, is the intelligence of the whole cycle, and it is the form that existing tests are least equipped to measure.
V. Development, Talent, and Expertise
The LVT framework generates a specific hypothesis about human development: children develop intelligence across all five mediations and at all five recursivity levels, but not at the same pace, not with the same facility, and rarely to the same extent. It is this differential developmental trajectory, different rates of development across different mediational-recursive combinations, that produces what we call talent.
This hypothesis cuts against two prevailing views. The first is that talent is primarily genetic, an innate endowment that unfolds under suitable conditions. The second is that talent is primarily the product of accumulated practice, with no significant innate differences. LVT's position is different from both. It does not deny that there are individual differences in the rate and facility of development across mediational-recursive domains. But these differences are not best understood as differences in a single underlying capacity. They are differences in the specific mediational-recursive combinations that develop most readily for a particular person, in a particular developmental environment, under particular conditions of engagement.
Consider a child who grows up in a household where music is constantly present, where adults demonstrate close attention to sound, rhythm, and the emotional texture of musical experience. This child develops, through sustained exposure to a musically saturated environment, forms of embodied, relational, and symbolic intelligence specifically tuned to the musical domain. The body learns to feel rhythm at L1. L2 sensitivity to harmonic misalignment develops. L3 capacity to name and share musical experience grows. Symbolic intelligence at L4 that can later develop theoretical musical understanding is built on this foundation. The child who develops in this environment is not simply accumulating practice hours. They are building a mediational-recursive profile specifically well-suited to musical intelligence, and this profile, once established in early development, shapes every subsequent encounter with musical material. That is talent: not a fixed gift, but a developmental trajectory with deep roots.
Children at different stages are not less intelligent than adults. They are differently distributed across the mediational-recursive matrix. The two-year-old who cannot perform the Sally-Anne false belief task, the standard test of theory of mind, may nevertheless be exercising extraordinary inter-recursive intelligence at L1 and L2: reading caregiver emotional states with remarkable accuracy, calibrating expressions to the inter-recursive responses received, managing complex social negotiations with siblings and peers. The failure of the explicit false-belief task does not indicate the absence of inter-recursive intelligence. It indicates that L4 symbolic representation of others' mental states has not yet developed. The intelligence is present at lower recursive levels. Theory of mind, on this account, is not the acquisition of inter-recursive intelligence. It is the development of L4 symbolic representation of inter-recursive intelligence that was already operating. This reframes developmental psychology's central finding: children are not gradually acquiring the capacity for social cognition. They are gradually developing the symbolic apparatus to represent and reflect on social cognition they already possess.
The developmental implications of the LVT framework extend beyond theory of mind. They reach every domain of mediational development. A child raised in a household with rich material culture, surrounded by objects that are used, maintained, repaired, and explained, develops L1 and L2 material intelligence from very early. A child raised with limited material engagement will have less developed material attunement, not because of any innate difference in material intelligence but because the developmental environment has not provided the sustained exposure to real material stakes that material intelligence requires. A child raised in frequent contact with animals, plants, and ecological cycles develops dwelling intelligence from early childhood: the capacity to read non-human states, to feel seasonal rhythms, to notice ecological signs. This form of development is almost entirely absent from urban educational environments, which means that an entire form of mediational intelligence is routinely underdeveloped by default. The consequences are not merely individual. They are collective: communities whose members have not developed dwelling intelligence are communities that have lost the capacity to feel what is happening in the ecological conditions of their lives before it becomes a crisis requiring symbolic intervention.
Different professions develop characteristic mediational profiles through sustained occupational engagement. An excellent physician develops strong embodied intelligence (reading the body through touch, sound, and observation), relational intelligence (the inter-recursive capacity to understand what a patient is not saying as well as what they are), material intelligence (sensitivity to instruments, medications, and physiological systems), and symbolic intelligence (the capacity to reason with abstract diagnostic categories and research evidence). What marks the great physician from the competent one is often not greater symbolic knowledge but greater L2 sensitivity, the felt sense that something is wrong before the indicators confirm it, combined with recursive fluidity: the ability to move from that felt sense into articulation, analysis, action, and back into absorbed engagement with the patient without losing any of these levels.
A skilled nurse develops a profile weighted toward embodied and relational intelligence, often with remarkable L1 and L2 depth in both: the capacity to read a patient's trajectory through subtle changes in colour, sound, texture, and movement, combined with inter-recursive sensitivity to what a patient's emotional state is signalling about their physical state. The nurse's intelligence is often less visible than the physician's because it operates primarily at L1 and L2, below the threshold of symbolic reportability, and because the institutional apparatus of medicine has been designed to recognise L4 symbolic knowledge more readily than L1 and L2 attunement.
A carpenter or craftsperson develops high material and embodied intelligence across all recursive levels, from the L1 knowledge of how a particular wood grain should feel under the plane to the L4 understanding of structural principles applicable across materials and contexts. Dwelling intelligence is also typically strong: the craftsperson reads the building they are working in as a field of spatial and material relations that shape every decision. Relational intelligence in the inter-recursive sense is less central, though even here the experienced carpenter reads what the client needs often before the client can articulate it.
An experienced anthropologist develops a profile weighted toward relational and dwelling intelligence, with symbolic intelligence playing an integrating and reflective role. The fieldworker develops, through extended engagement, L1 and L2 attunement to the social dynamics and spatial texture of the community they are working with, forms of intelligence that are hard-won and that cannot be transmitted through reading ethnographies. L5 symbolic intelligence is also typically developed: the capacity to notice when one's own theoretical frameworks are generating artefacts, and to reposition the inquiry in response. The combination of deep L1 relational attunement with high L5 reflective capacity, the ability to feel what is happening in a social field while simultaneously examining the frameworks through which one is feeling it, is the distinctive intelligence of good fieldwork.
A farmer develops high dwelling and material intelligence across all recursive levels. At L1, the experienced farmer's body knows the field: the soil underfoot, the quality of light, the weight of the air before rain. At L2, the farmer feels when something is off, when a crop is stressed, when the drainage is failing, when the seasonal rhythm is disrupted. At L3, they can name and communicate what they sense. At L4, they have developed abstract principles of rotation, soil management, and seasonal timing that can be applied across varying conditions. At L5, the reflective farmer can examine whether the frameworks they have inherited from agricultural tradition are adequate to changing conditions, whether the principles learned from their parents still fit a climate that is shifting.
A therapist develops the highest inter-recursive intelligence as their primary mediational domain, combined with embodied intelligence as its immediate vehicle and symbolic intelligence as its reflective apparatus. The exceptional therapist operates primarily at L1 and L2 inter-recursive intelligence in session: feeling the dynamics of the relationship as it unfolds, sensing when the patient is approaching defended material, knowing when to speak and when silence is more responsive. L4 and L5 symbolic intelligence are developed through training and supervision, but the best therapists hold these lightly, using them to orientate without allowing them to override the immediate felt sense of what the relationship needs.
VI. The Five Mediational Intelligences
Having established that all five mediations are always present, that profiles describe development not presence, and that levels L3-L5 are forms of symbolic reflection applied to the other mediations, the five forms of mediational intelligence can be characterised in their own right.
Embodied intelligence is the capacity to be appropriately responsive to the stakes of the body: to read embodied states accurately, manage physiological conditions, navigate the physical requirements of skilled action, and use the body's own signalling as a source of information about the state of the field. The skilled surgeon's hands know things that explicit knowledge does not contain. The experienced dancer's body has internalised structures of timing, balance, and spatial relation that would require pages of notation to approximate propositionally. The midwife's hands, reading the position of a foetus through the abdominal wall, exercise a form of intelligence that cannot be reduced to algorithmic application. The firefighter entering a burning building reads smoke, sound, and the behaviour of flames as a continuous stream of information, not by consciously analysing each signal but by responding with the whole body to the field as a living, dangerous totality.
Embodied intelligence cannot be separated from its stakes without being destroyed. It is not a transferable cognitive capacity that happens to be expressed through the fingers. It is constituted in the relationship between the skilled body, the material being worked with, and the purposes that give every movement its meaning. Its highest development is L1: effortless, absorbed, precise, carrying no overhead of deliberate monitoring.
Relational intelligence, intelligence in being-with, is the capacity to be appropriately responsive to the stakes of inter-recursive situations: to read the configuration of trust, obligation, distress, desire, and possibility in a field of human relationships, and to respond in ways calibrated to what is actually happening rather than to a symbolic representation of it. Relational intelligence in the LVT sense is not the same as emotional intelligence in the psychological sense. Emotional intelligence frames emotions as objects of cognitive processing. Relational intelligence is a form of stakes-responsiveness in an inter-recursive domain: the capacity to be oriented by what is at stake in a relationship, to feel the stakes rather than represent them. Its most powerful form is L1 and L2: the felt sense of what a relationship needs, operating before explicit reflection.
Dwelling intelligence is the capacity to read and respond to the stakes of place: the practical knowledge of how a particular terrain, climate, ecosystem, or built environment works, what it affords, what it demands, what its possibilities are. The Polynesian navigator crossing three thousand kilometres of open ocean without instruments, using the stars, the swell patterns, the flight paths of birds, the colour and temperature of the water, and the behaviour of clouds around distant islands, is reading a living field with the whole body. The same structure appears in the Arctic hunter reading ice, snow, and wind to track game; in the Aboriginal tracker reading the ground for signs of an animal's passage; in the shepherd reading approaching weather; in the farmer reading soil, rain, and season.
What all of these examples share is that the practitioner is not applying a stored knowledge system. They are in continuous, real-time, bidirectional exchange with a living field whose relevant signals are below the threshold of explicit representation. They are maintaining an attunement, a continuous calibration of bodily and perceptual sensitivity to the field, that allows information to be received and acted upon before it has been formulated as information at all. This is L1 and L2 dwelling intelligence: the most direct form of environmental responsiveness, and the form most completely invisible to any measure that operates through explicit task performance.
Material intelligence, intelligence in multimateriality, is the capacity to understand and work effectively with the physical properties of things: materials, tools, technologies, substances, and the mechanical and chemical systems through which human life is organised. The craftsperson who reads the grain of wood and works with rather than against it; the mechanic who hears in the sound of an engine the specific nature of a developing problem; the programmer who feels in the behaviour of a codebase where the architectural tensions are; the chef who senses in a sauce exactly when the balance of flavour has been achieved, all are exercising material intelligence. The common structure: accurate stakes-responsiveness to the real properties of real physical systems, developed through extended engagement and sustained by ongoing attention.
Symbolic intelligence, intelligence in multisymbolism, is the capacity to navigate abstract symbolic systems: language, number, logic, theoretical frameworks, code, and formal argument. Its most important feature is that it can maintain responsiveness to absent stakes: to things not immediately present, to past events and future possibilities, to abstract structures and general patterns. This is what enables mathematics, science, law, history, writing, and institutional memory. Symbolic intelligence can preserve the accumulated responses to stakes that are no longer immediately present, carrying forward what was learned in earlier encounters with difficulty and making it available for future situations that share relevant structure. This capacity for temporal reach is both the distinctive strength of symbolic intelligence and what makes it capable of becoming detached from the living stakes that originally generated it.
VII. Hybrid Intelligence and Mediational Profiles
The five mediational forms of intelligence are conceptually distinct, but no real exercise of intelligence is purely one of them. All real intelligence is hybrid. The difference between practitioners lies not in which mediations they have access to (all five are always present), but in the specific profile of their development across the mediations and across the recursive levels.
Hybrid intelligence is what happens when multiple mediational forms combine in a single domain of activity. Most sophisticated human activities require specific combinations, and the intelligence they require is not simply additive. It is a qualitatively different achievement that emerges from the specific integration.
Surgery requires embodied intelligence (the hands' knowledge of tissue), material intelligence (sensitivity to the properties of instruments and biological material), relational intelligence (the read of the patient's state and the team's coordination), and symbolic intelligence (the procedural and anatomical knowledge that organises all of this). But a practitioner who has all four developed in isolation, without their real-time integration, cannot operate. The intelligence of surgery is constituted in their continuous combination, in the capacity to hold all mediational dimensions simultaneously and let them inform each other. Jazz improvisation requires embodied intelligence (timing and the physical knowledge of the instrument), relational intelligence (live attunement to the other musicians and the audience), symbolic intelligence (harmonic and rhythmic structure), and dwelling intelligence (the feel of the room as a space of resonance). Anthropological fieldwork requires relational intelligence (genuine encounter with another world without either imposing one's own or disappearing into the other's), dwelling intelligence (learning the landscape and its social texture), symbolic intelligence (theoretical and linguistic resources), and embodied intelligence (physical presence, sharing food and work and discomfort).
Mediational profiles are the characteristic distributions of development across the five mediations and across the recursivity levels that individuals develop through the specific conditions of their lives. No person is equally developed across all five mediations or at all five recursivity levels. Everyone has a characteristic profile, areas of rich development and areas still developing, shaped by their developmental history, their ecological situation, and the specific fields of activity in which they have invested sustained attention.
The mediational profile concept replaces Gardner's multiple intelligences with a more precise and theoretically grounded framework. Where Gardner gives a list of types, LVT gives a continuous space of possible distributions. Each individual occupies a position in a multi-dimensional space, with dimensions representing their degree of development in each mediational form and at each recursive level. This allows for indefinitely many distinct profiles, captures the hybrid character of complex competencies, and explains why the apparent number of distinct intelligence types is always contestable, because the space of possible profiles is continuous, not discrete.
The point of the profile concept is distribution, not ranking. A person with richly developed dwelling and embodied intelligence and still-developing symbolic intelligence is not less intelligent than a person with the inverse distribution. They are differently distributed. The question worth asking about any person is not "how intelligent are they?" but "what is the current shape of their mediational-recursive profile, and what configurations of challenge and collaboration would allow that profile to continue developing?"
VIII. Differential Temporalities and the Intelligence of Timing
Each mediation has its own temporality, its own rhythm of change, its own characteristic scale of relevant duration, and intelligence in each mediational domain requires sensitivity to that temporality. One of the most undervalued forms of intelligence is knowing when not to respond.
Embodied temporality is measured in milliseconds, breath-cycles, and the duration of skilled actions. The surgeon's intelligence includes sensitivity to how long an incision takes to bleed, how long a suture takes to hold. The jazz musician's intelligence includes sensitivity to the temporal microstructure of rhythm, to the sub-millisecond timing differences that distinguish swing from mechanical regularity. Embodied intelligence calibrated to this temporality responds to the body's own rhythms rather than imposing an external tempo.
Relational temporality is measured in the duration of relationships and the cycles of trust, rupture, and repair. The therapist's intelligence includes knowing that certain material cannot be approached until the relationship has a certain density and safety. The parent's intelligence includes knowing when to let a child sit with difficulty and when to intervene, and crucially, when to remain silent. The capacity to stay present without acting, to tolerate uncertainty without filling it prematurely, is itself a form of relational intelligence. The parent who waits while a child works through something difficult, without rescuing or intervening, is exercising a form of relational intelligence that premature intervention would suppress. Relational intelligence operating at the right temporal scale is intelligence that can work with how relationships actually unfold.
Dwelling temporality is measured in seasons, years, and generations. The farmer's intelligence includes the capacity to read the seasonal cycle as a dynamic field, to know when the soil is ready, when the risk of late frost has passed, when the harvest window will close. But it also includes the intelligence of delay: knowing when not to plant, when to let a field rest, when the conditions that seem ready are not yet ready. The indigenous tracker includes sensitivity to multi-year cycles of animal behaviour and ecological flux. The architect's intelligence includes the long view: how a building will age, how its social uses will evolve, how the neighbourhood around it will change.
Material temporality is measured in the lifecycles of materials and technologies: how long a structure takes to stress-settle, how long a chemical process takes to stabilise, how long a piece of software takes to accumulate the consequences of early architectural decisions. Material intelligence includes the capacity to read these temporal signatures, to know that a developing condition in a material system will become significant in weeks, to understand the difference between a symptom and a cause. The physician who waits for more information before acting is often exercising higher material intelligence than the physician who acts immediately on incomplete evidence.
Symbolic temporality is the slowest and in some ways the most demanding. Concepts, frameworks, and theories are formed in response to specific historical configurations of stakes, and they carry forward the authority of their founding moment long after the conditions have changed. Intelligence in the symbolic domain includes sensitivity to this lag, the capacity to ask whether a concept is capturing a live stake or preserving a historical one, whether a framework is still adequate to the field it was formed to describe. This temporal self-awareness, holding one's own symbolic resources critically without abandoning them, is one of the highest and most demanding forms of L5 symbolic intelligence.
The intelligence of timing is therefore not reaction speed. It is temporal calibration: the capacity to feel which phase of a process one is in, what rhythm of response the situation calls for, and when the most intelligent act is to wait. The experienced doctor waiting before prescribing; the farmer delaying planting; the parent remaining silent while a child struggles; the therapist tolerating uncertainty, all are demonstrating that intelligence includes knowing when not to respond.
IX. The Social Constitution and the Hard Case
Intelligence is not only exercised in social contexts. In significant part, it is constituted by them. The research tradition on distributed cognition, Hutchins' work on ship navigation, Lave's work on mathematical reasoning in the wild, has established that cognitive performance is systematically different when embedded in social and material systems than when extracted from them. A crew navigating a large vessel does not solve navigational problems by aggregating the individual cognitive performances of its members. The navigation is distributed across people, instruments, charts, and conventions in a way that the collective system can accomplish what no individual could.
Groups and communities also possess mediational profiles. A research team may have developed extraordinary symbolic intelligence collectively while having limited dwelling intelligence: able to theorise the environment brilliantly while unable to read what is actually happening in it. A family may have developed high relational intelligence collectively, a finely tuned capacity for inter-recursive attunement among members, while having underdeveloped material intelligence: unable to manage the physical and practical conditions of the household effectively. A religious community may have rich symbolic and embodied intelligence, liturgical and bodily practices developed over generations, with varying dwelling intelligence depending on whether the community has remained rooted in one place or has been dispersed. These collective profiles are real and consequential, shaped by the community's history, its practices, and the specific stakes it has collectively been exposed to.
The hardest challenge for a stakes-based account remains mathematical and scientific intelligence, which seems to operate independently of mesocosmic stakes. The answer requires two moves. First: the mathematician inhabits a practice, a mesocosm, with its own internal stakes, its own community, its own traditions of correctness and beauty, its own felt tensions. The pressure of a problem that resists solution, the felt sense of a near-miss in a proof attempt, the recognition that an approach is generating real insight, these are real L2 misalignments within mathematical practice. The mathematician's intelligence is the same structure of felt orientation and response that governs all other forms of stakes-responsiveness, operating in a highly formalised symbolic domain with its own ecology.
Second: symbolic intelligence can maintain responsiveness to absent stakes across vast temporal distances. The number theorist developing elliptic curve theory has no application in mind. Decades later, the work is foundational to public-key cryptography. Deep mathematical intelligence tracks the structure of the world at a level where the specific mesocosmic stakes are not yet present, but the connection to future stakes is real. This is symbolic intelligence in its most distinctive mode: preserving responsiveness to situations that do not yet exist.
X. Measurement and Mesocosmic Fit
The history of intelligence testing is not simply a history of better or worse measurements. It is a history of misrecognised mesocosmic fit, and it repeats, with remarkable precision, the central error that the value paper identifies in the history of economics.
IQ testing did not discover intelligence as such. It discovered those forms of intelligence most hospitable to symbolic quantification. This matters because measurement is not merely symbolisation. It is symbolisation under the special constraint of quantification. A test must produce comparable outputs. It must turn performances into scores. It must privilege tasks that can be standardised, repeated, timed, ranked, and aggregated. These requirements are not innocent. They select in advance for forms of intelligence that already fit the measurement apparatus.
Symbolic intelligence, especially in its abstract L4 register, fits this apparatus unusually well. Verbal analogy, numerical reasoning, pattern completion, memory span, and decontextualised problem solving can all be rendered into scored performances. They produce traces that travel cleanly through institutions. By contrast, the most important forms of embodied, relational, dwelling, and material intelligence often do not produce quantifiable outputs without distortion. The nurse who senses deterioration before the monitors show it, the parent who knows when not to intervene, the farmer who delays planting because the soil is not yet ready, the fieldworker who feels a shift in a room before anyone has said anything: all are exercising intelligence. But their intelligence is not naturally scoreable. It lives in timing, attunement, inhibition, repair, and recursive responsiveness. It often becomes visible only retrospectively, when one can say: that intervention was well timed; that silence preserved the relation; that hesitation prevented harm; that misalignment was correctly felt before it could be named.
IQ testing therefore repeats the central error of economics. Economics mistook price, preference, utility, and measurable output for value. Intelligence testing mistook quantifiable symbolic performance for intelligence. In both cases, the measurement system selected the domains most hospitable to quantification and then treated those domains as ontologically primary. What could be measured became real. What resisted measurement became secondary, subjective, informal, or invisible. The measurement system selected its object rather than discovering it.
The parallel with economic value is precise and instructive. The value paper shows that economics progressed by extending measurement from domains that tolerate symbolisation, land area, population, transaction volumes, to domains that do not. The local success of measurement in hospitable domains was mistaken for a general ontology. The same progression occurred in intelligence research. IQ tests succeeded in producing reliable measurements of performance on certain symbolic tasks, and this local success was mistaken for a general theory of intelligence. The measurable subset became the whole, and the unmeasurable remainder was categorised as either irrelevant or as a failure of measurement technique rather than as a feature of the intelligence itself.
What the mesocosm discloses about intelligence measurement, through the felt strain of practitioners who are assessed by tests that do not capture what they actually do, is that measurement is not neutral. It is an intervention. When an assessment system measures only L4 symbolic performance, it does not merely fail to capture other forms of intelligence. It reorganises what is valued, what is taught, what is recognised, and what is rewarded. Over time, it shapes which forms of intelligence are developed by those whose development is organised around the assessment system. The IQ test does not merely discover the intelligence that is there. It contributes to producing the forms of intelligence that it then finds.
A distinction is needed here, drawn from the account of mesocosmic fit in the value paper. Some domains of intelligence tolerate measurement well. Non-recursive symbolic task performance, the performance of verbal and numerical reasoning under standardised conditions, fits measurement. The numbers are real. The correlations are genuine. The problem is not that measurement is never appropriate for intelligence. The problem is when the local success of measurement in hospitable domains is mistaken for a universal account. When measurement of L4 symbolic performance is treated as the measurement of intelligence as a whole, the specific form of intelligence that fits the measurement apparatus is elevated to ontological primacy, and the four other mediational forms are structurally excluded.
A Living Value Theory of intelligence therefore requires a theory of measurement fit. The question is not simply: how much intelligence does this person have? The better questions are: which mediational intelligences are being activated by this task? Which recursive levels does the task require? Which recursive domain does it belong to? Which forms of intelligence are excluded because they cannot be quantified? What forms of mesocosmic intelligence become invisible when intelligence is defined by scoreability?
The answer to these questions should reorient how assessment is understood. It should not abolish measurement, some forms of intelligence can be measured. It should contextualise measurement: make explicit what is being measured and what is not, what mesocosmic conditions the measurement requires, and what forms of intelligence the measurement apparatus structurally cannot see. The IQ score, understood in these terms, is not a measure of intelligence. It is a measure of performance on symbolic tasks under specific conditions, conditions that select for L4 symbolic capacity, that require prior access to the mesocosmic conditions enabling L4 operation, and that systematically exclude embodied, relational, dwelling, and material intelligence from the account. This is a real and useful measurement, within its domain. Its catastrophic failure is the belief that it measures more than it does.
The positive account of intelligence measurement that follows from LVT begins not from the question "how much intelligence does this person have?" but from the set of questions that mesocosmic fit requires. Which mediational intelligences are being activated by this task, and which are being excluded? Which recursive levels does the task require, and which does it suppress? Which recursive domain does it belong to, non-recursive, self-recursive, or inter-recursive? What forms of mesocosmic intelligence become structurally invisible when intelligence is defined by what can be scored? And what conditions does the assessment require, what access to prior symbolic resources, what suspension of competing stakes, what institutional legitimacy in relation to the assessment setting, before a person can perform at all?
These questions do not abolish measurement. They contextualise it. Some forms of intelligence can be assessed with genuine usefulness: the speed and accuracy of L4 symbolic reasoning under standardised conditions is real, measurable, and sometimes relevant to predict performance on tasks that share the same structure. The problem is exclusively the claim to generality. Within its hospitable domain, intelligence measurement works. Beyond that domain, it does not merely fail to capture something. It substitutes a measurable proxy for a living reality, then treats the proxy as primary and the reality as secondary. This is the same movement the value paper identifies in economics: the domain of hospitable symbolisation, transaction volumes, land areas, price indices, successfully measured, then the local success extrapolated into a universal ontology. When the map begins to charge rent to the territory, the measurement system has crossed from useful tool into ontological claim. At that point, as the value paper argues, the reversal of the burden of proof becomes necessary: symbolic systems must justify their intrusion into the mesocosm, not the other way around. The same applies to intelligence measurement. The IQ score does not need to justify itself to the mesocosm of living intelligence. It needs to justify its claim to represent that mesocosm, and on examination, the claim is much narrower than it appears.
XI. Anthropology as the Archive of Intelligence
One of the largest and least recognised consequences of a Living Value Theory of intelligence is what it does to the anthropological record. Over more than a century, anthropology has assembled an extraordinary archive of human competence: detailed documentation of agricultural expertise, navigational intelligence, healing traditions, ritual expertise, ecological knowledge, craft mastery, kinship mediation, animal handling, food preparation, weather reading, architectural intelligence, and dwelling competence across an enormous range of cultures, environments, and historical periods. On the LVT account, this is not merely documentation of cultural variety. It is the world's largest comparative intelligence project, conducted without recognising itself as such.
Ethnographies contain: navigation, hunting, farming, healing, ritual, conflict resolution, kinship management, architecture, weather reading, animal management, craft production, food systems, and trade. This is an astonishing database of mediational intelligence in action, far richer, more varied, and more precisely documented than anything laboratory psychology has produced. The Polynesian navigator, the Aboriginal tracker, the Amazonian healer, the master potter, the seasonal farmer, the ritual specialist, the kinship mediator, all are not merely performing cultural roles. They are exercising real forms of intelligence: high-stakes responsiveness to real conditions, developed through sustained engagement, operating across multiple mediations and multiple recursive levels simultaneously.
Ethnographies are, in particular, documentation of mediational profiles. The Polynesian navigator's profile weights dwelling and embodied intelligence very highly, with symbolic intelligence playing an integrating role through the star-path system and the oral tradition of navigational knowledge. The experienced healer's profile weights relational and embodied intelligence, with symbolic intelligence providing the diagnostic and ritual frameworks that organise what is felt and observed. The farmer's profile weights dwelling and material intelligence across all recursive levels, from L1 attunement to soil and season through L4 principles of cultivation to L5 reflection on whether inherited agricultural knowledge remains adequate to changing conditions.
What anthropology contributes to a Living Value Theory of intelligence, therefore, is not merely examples. It contributes two things of deeper significance. First, it demonstrates that the forms of intelligence that matter most for most of human life, those directly engaged with the management of food, safety, health, relationship, and the conditions of dwelling, have been practised and refined over thousands of years, in thousands of different configurations, by billions of human beings. The history of intelligence is not primarily a history of abstract reasoning. It is primarily a history of the full range of mediational intelligences through which human beings have found their way in an endlessly varied and endlessly demanding world.
Second, anthropological fieldwork is itself a form of inter-recursive intelligence in action: the fieldworker developing relational, dwelling, embodied, and material intelligence in a new field, through sustained exposure to its real stakes. The anthropologist who returns from the field changed, not merely informed but reconfigured, has undergone mediational development. This is why fieldwork cannot be replaced by interviews, surveys, or remote observation. It requires the full presence of a person in a field with real stakes, developing the forms of attunement that only real stakes can produce. The methodology of anthropology is itself an expression of the theory of intelligence that LVT proposes: you cannot understand a world from outside its stakes. The fieldworker's intelligence grows through the field, not through preparation for it.
Conclusion: Intelligence as Living Achievement
Intelligence, on the account developed here, is not a thing but a living accomplishment of successful mediation, a quality of ongoing relation between an organism and the stakes of its situation, distributed across five irreducible mediations that are always present in any human being, operating at five levels of recursivity through which any form of coordination can be engaged with different degrees of symbolic reflection, and requiring recursive fluidity: the capacity to move between levels as the situation demands.
The history of intelligence research, viewed through this lens, is a history of repeated partial discoveries that were unable to integrate their findings because they lacked the theoretical architecture needed to place them in relation to each other. Psychometrics found the positive manifold in symbolic performance and elevated measurable intelligence to the status of intelligence as such. Gardner found plurality but lacked a stopping rule. Ecological psychology found situated responsiveness at L1 and L2. Distributed cognition found collective mediation. Each was right. None was complete. LVT offers the architecture that makes the discoveries intelligible together: they were all finding different expressions of a single underlying structure of living mediation.
The five mediational intelligences, embodied, relational, dwelling, material, symbolic, are not a revised list to replace Gardner's. They are a derived architecture grounded in the irreducible structure of human existence. Every person develops all five. What varies is the profile: the specific distribution of development across the mediations and across the recursivity levels, shaped by developmental history, sustained engagement, and the specific stakes the person has been exposed to. Talent is the name we give to mediational-recursive combinations that have developed with unusual facility and depth. Expertise is the name we give to profiles that have been developed through sustained occupational engagement in specific domains. What both have in common is not a fixed gift but a developmental trajectory: a history of sustained exposure to real stakes in specific mediational-recursive domains, producing forms of attunement that become available for the next generation of stakes encountered.
Intelligence measurement, as currently practiced, is a history of measuring measurable intelligence and confusing it with intelligence as a whole. IQ testing discovered those forms of intelligence most hospitable to symbolic quantification, L4 symbolic capacity under stakes-free conditions, and elevated this to the status of a general measure. The error is structurally identical to the error of economics: the local success of measurement in hospitable domains mistaken for a universal ontology. The remedy is not better measurement but better measurement fit: explicit recognition of what is being measured and what is structurally excluded, what mesocosmic conditions the measurement requires, and what forms of intelligence the apparatus cannot see.
The forms of intelligence most central to most human lives, the L1 attunement of the skilled practitioner, the L2 felt sense of the experienced clinician, the relational and dwelling intelligence of those who maintain the conditions of social and ecological life, are systematically undervalued, underdeveloped institutionally, and invisible to the measurement systems that have come to define intelligence for educational, professional, and policy purposes. This is not merely a measurement problem. It is an ontological one: a mistaking of symbolically quantifiable traces for the reality they were meant to approximate.
Anthropology has spent a century documenting this reality without recognising what it was documenting. Its archive of mediational intelligence, of navigators, farmers, healers, craftspeople, kinship mediators, and ritual specialists, is the richest database of human intelligence in existence. LVT gives these findings the theoretical framework they were waiting for.
The mouse, at the end of The Gruffalo, eats its nut. Not strategising, not theorising, not maintaining vigilance: simply returning to L1 after having done what was needed at the right recursive level, in the right recursive domain, in the right mediational register, at the right moment. That is intelligence: not a thing, but a living accomplishment of successful mediation.
References
Baron-Cohen, S. Mindblindness: An Essay on Autism and Theory of Mind. Cambridge, MA: MIT Press, 1995.
Bourdieu, Pierre. The Logic of Practice. Translated by Richard Nice. Cambridge: Polity Press, 1990.
Carraher, T.N., Carraher, D.W. and Schliemann, A.D. "Mathematics in the streets and in schools." British Journal of Developmental Psychology 3, no. 1 (1985): 21–29.
Csikszentmihalyi, M. Flow: The Psychology of Optimal Experience. New York: Harper & Row, 1990.
Dehaene, S. The Number Sense: How the Mind Creates Mathematics. New York: Oxford University Press, 1997.
Dreyfus, H.L. and Dreyfus, S.E. Mind Over Machine: The Power of Human Intuition and Expertise in the Era of the Computer. New York: Free Press, 1986.
Dweck, C.S. Mindset: The New Psychology of Success. New York: Random House, 2006.
Ecks, Stefan. Living Worth: Value and Values in Global Pharmaceutical Markets. Durham: Duke University Press, 2022.
Flynn, J.R. What Is Intelligence? Beyond the Flynn Effect. Cambridge: Cambridge University Press, 2007.
Fonagy, P., Gergely, G., Jurist, E.L. and Target, M. Affect Regulation, Mentalization, and the Development of the Self. New York: Other Press, 2002.
Gardner, H. Frames of Mind: The Theory of Multiple Intelligences. New York: Basic Books, 1983.
Gardner, H. Intelligence Reframed: Multiple Intelligences for the 21st Century. New York: Basic Books, 1999.
Gibson, J.J. The Ecological Approach to Visual Perception. Boston: Houghton Mifflin, 1979.
Gould, S.J. The Mismeasure of Man. Revised edition. New York: W.W. Norton, 1996.
Hutchins, E. Cognition in the Wild. Cambridge, MA: MIT Press, 1995.
Ingold, T. The Perception of the Environment: Essays on Livelihood, Dwelling and Skill. London: Routledge, 2000.
Ingold, T. Being Alive: Essays on Movement, Knowledge and Description. London: Routledge, 2011.
Jensen, A.R. The g Factor: The Science of Mental Ability. Westport, CT: Praeger, 1998.
Kahneman, D. Thinking, Fast and Slow. New York: Farrar, Straus and Giroux, 2011.
Klein, G. Sources of Power: How People Make Decisions. Cambridge, MA: MIT Press, 1998.
Lave, J. Cognition in Practice: Mind, Mathematics and Culture in Everyday Life. Cambridge: Cambridge University Press, 1988.
Lave, J. and Wenger, E. Situated Learning: Legitimate Peripheral Participation. Cambridge: Cambridge University Press, 1991.
Merleau-Ponty, M. Phenomenology of Perception. Translated by C. Smith. London: Routledge & Kegan Paul, 1962.
Nisbett, R.E. Intelligence and How to Get It: Why Schools and Cultures Count. New York: W.W. Norton, 2009.
Piaget, J. The Origins of Intelligence in Children. Translated by M. Cook. New York: International Universities Press, 1952.
Polanyi, M. The Tacit Dimension. New York: Doubleday, 1966.
Richardson, K. The Making of Intelligence. New York: Columbia University Press, 2000.
Rogoff, B. Apprenticeship in Thinking: Cognitive Development in Social Context. New York: Oxford University Press, 1990.
Salovey, P. and Mayer, J.D. "Emotional intelligence." Imagination, Cognition and Personality 9, no. 3 (1990): 185–211.
Spelke, E.S. "Core knowledge." American Psychologist 55, no. 11 (2000): 1233–1243.
Sternberg, R.J. Beyond IQ: A Triarchic Theory of Human Intelligence. Cambridge: Cambridge University Press, 1985.
Tomasello, M. The Cultural Origins of Human Cognition. Cambridge, MA: Harvard University Press, 1999.
Tomasello, M. Becoming Human: A Theory of Ontogeny. Cambridge, MA: Harvard University Press, 2019.
Varela, F.J., Thompson, E. and Rosch, E. The Embodied Mind: Cognitive Science and Human Experience. Cambridge, MA: MIT Press, 1991.
Vygotsky, L.S. Mind in Society: The Development of Higher Psychological Processes. Cambridge, MA: Harvard University Press, 1978.
Wechsler, D. The Measurement of Adult Intelligence. Third edition. Baltimore: Williams and Wilkins, 1944.