Language, resonance, and the three loops of the civilisational deposit. The deep foundation that explains why it exists at all.

The main argument of the book begins with Alexander, with the civilisational deposit already well advanced, with language and writing in place, and with the memetic revolution underway. But the personal integral did not begin with Alexander, or with Homo sapiens, or with language. It began whenever nervous systems began accumulating experience into parameter sets that shaped subsequent behaviour. That was a very long time ago.

What follows is the deep foundation of the book’s argument: the three-loop model of genetic evolution (slow, open-loop), the neural integral (fast, closed-loop, within a lifetime) and the cultural deposit (the neural loop extended beyond the individual lifetime through language), and the integral’s history from the Cambrian to civilisational accumulation. It explains why human beings have the integration capacity they do, why language proved so decisive, and why the resonance pathways the book describes are biological inheritances refined over hundreds of millions of years rather than cultural inventions. Understanding these evolutionary roots also resolves a puzzle that has troubled memetics since Dawkins introduced it. How could a capacity this metabolically expensive and this slow to develop have been paid for by natural selection alone? The Integral offers a clearer answer.

1. The Three Loops of Evolution

The standard neo-Darwinian account of evolution is, at the level of the individual organism, an open-loop system. Genetic variation arises through mutation and recombination, processes that are random with respect to fitness. Natural selection operates on this variation after the fact, favouring variants that survive and reproduce more successfully in the current environment. The individual organism has no way of feeding what it learns in its lifetime back into the genetic inheritance system. Weismann’s barrier holds. The soma does not write back to the germline. What you accumulate in your lifetime does not change your DNA. The loop is broken at the individual level by design.

The barrier, however, is better understood as heavily guarded than impermeable. Epigenetic research reports evidence that methylation patterns governing gene expression can be partially inherited across one or two generations, in descendants of Holocaust survivors among others. The human findings remain contested and the mechanism is not established. The Dutch Hunger Winter of 1944–45 is often cited here but is not a case of inheritance: those affected were in the womb during the famine, and the methylation differences still measurable sixty years later record their own exposure. That is the stronger result for this account: a developmental window leaving a trace a lifetime did not erase. The DNA sequence is not rewritten by somatic experience. The volume dial on specific genes can be. This partial permeability operates under conditions of extreme and sustained stress, not ordinary accumulation. That is consistent with the three-loop model’s account of the long decay constant. Most inputs fade, but some, arriving at sufficient amplitude during critical developmental windows, leave traces that persist beyond a single lifetime. Far from being a flaw, this is what prevents Lamarckian drift: the accumulation of individually acquired characteristics that would rapidly degrade the genetic signal.

But the open loop creates a problem. Genetic evolution operates at generational timescales, 20 to 30 years minimum in humans and longer in many other species. For organisms navigating complex and rapidly changing environments, this is far too slow. An individual that cannot adapt within its lifetime is at the mercy of whatever its genes happened to encode for the last environment its ancestors inhabited.

Epigenetics occupies the gap between the first and second loops rather than belonging cleanly to either. Its modifications (DNA methylation and histone modification) operate primarily on somatic cells, shaping gene expression continuously in response to developmental environment, diet, and stress. This is not Lamarckian inheritance; the DNA sequence is unchanged. But it is a partial blurring of the clean open-loop account. The volume dial on gene expression can be turned, and in extreme cases that setting can persist across one or two generations through germ cell modification. The three-loop account remains the right working model. Epigenetics is its most important qualification, a secondary channel that confirms the barrier is permeable under sufficient pressure, without collapsing the distinction that the barrier exists to maintain.

Genetic evolution is the first loop (open, slow, variation and selection operating across generations). The neural integral is the second loop, closed and fast, the within-lifetime learning that allows the organism to adapt without waiting for genetic encoding. Together they constitute a two-speed system that could adapt at both timescales simultaneously. But the human case adds a third loop, the cultural one. It runs on the same neural machinery, a special case of the second loop, but its information is symbolic and external and its transmission gain is more than an order of magnitude higher, compounding across generations rather than within a lifetime. That is what makes it a loop in its own right, not merely the neural loop stretched. Language was the phase transition that activated it.

The three loops interact. The second loop, the neural integral, shapes the environments that individuals create and inhabit, which shapes the selection pressures on the first loop. The first loop sets the initial conditions of the second: the starting parameter set, the learning rate, the decay constants, the coupling between different integration systems. The third loop, the cultural deposit, shapes the inputs available to the second loop in every generation, and is shaped in turn by the outputs of the second loop laid down across generations. Over long enough timescales, the second and third loops drive the first. The question is not genes versus culture, nature versus nurture. The question is how the three loops interact, and how those interactions have changed over evolutionary time.

Sources: Boyd & Richerson — Culture and the Evolutionary Process (1985); Not By Genes Alone(2005). Henrich — The Secret of Our Success (2015).

2. The Deep History of the Integral

Integration does not begin with Homo sapiens. It begins whenever a nervous system first accumulated experience into a parameter set that shaped subsequent behaviour, which is to say with the first learning organisms. The Cambrian explosion, approximately 540 million years ago, saw the first complex animal nervous systems diversify and elaborate. Simpler nervous systems already reached back into the Precambrian, and it remains debated whether neurons arose once or independently more than once. It is plausible that the integral, in its most primitive form, was part of what the Cambrian explosion was about. Evolution had discovered that learning within a lifetime was a powerful enough adaptation to justify the metabolic expense of a nervous system.

Eric Kandel’s Nobel Prize-winning work on Aplysia californica (a sea slug with approximately 20,000 neurons) demonstrated that the cellular pathways of learning and memory, including synaptic strengthening through repeated activation, are present in nervous systems of extraordinary simplicity. The basic operation of the integral, experience modifying a parameter set through synaptic weight adjustment, is ancient. It predates vertebrates, complex brains, and anything that would recognisably be called intelligence. It is a fundamental feature of nervous system anatomy, conserved across hundreds of millions of years of evolution because it works.

What has evolved is the anatomy that runs it. More neurons. More layers of processing. Longer time constants, the ability to accumulate experience over longer periods without the signal decaying. Richer coupling between different integration systems (visual, auditory, spatial, social, emotional, linguistic). And increasingly sophisticated pathways for coupling individual integrals to each other and to shared cultural stores that persist beyond any individual lifetime.

The question of what it is like to be a crow, a dog, an elephant, or an octopus is, in the integral’s terms, the question of what kind of parameter set those animals have built, what they have integrated, across what modalities, with what time constants and coupling. The richness of animal cognition documented by researchers like Frans de Waal in primates, by Irene Pepperberg in Alex the parrot, and by Peter Godfrey-Smith in cephalopods, is not surprising here. These animals have been integrating for a very long time. Their integrals are tuned to different problems and operating across different modalities, but the underlying wiring is the same.

A cultural selection loop is not ours alone. Orca pods carry distinct foraging cultures transmitted by imitation across generations; chimpanzee tool use varies by group in ways environment does not explain; corvids maintain regional tool traditions. In each case something is accumulating above the genetic loop and below language. What language did was not start that process. It raised the bandwidth and the fidelity until the accumulation could compound.

The human brain is a quantitative extreme on a very long continuum rather than a discontinuity in this history, the most sophisticated integration machinery evolution has so far produced, distinguished above all by the pathways that allow individual integrals to couple to a shared cultural store that persists across lifetimes. The measurement supports it. Counted cell by cell, the human brain has almost exactly the number of neurons expected for a primate brain of its mass. There is no special human scaling rule to be found. What is unusual is the size, and what the size costs: a fifth of the body’s energy budget for a fiftieth of its mass.

Sources: Kandel — In Search of Memory (2006). Godfrey-Smith — Other Minds (2016); Metazoa (2020).

3. What Evolution Was Selecting For

If the neural integral is the second loop of evolution, then variations in the quality of the integral, in the sophistication of the integration machinery, are subject to natural selection. An organism with a tighter feedback loop, a longer time constant, a richer coupling between integration systems, and a more sophisticated way of updating, is a better learner. Better learners survive and reproduce more successfully in complex and changing environments. Selection thus favours better integrators.

This reframes what was being selected for across the long history of nervous system evolution. It was not just brain size, though larger brains generally support more sophisticated integration, nor just specific cognitive capacities in isolation, though those matter too. What was under selection was the quality of the integration machinery as a whole: the ability to accumulate experience faithfully, to update the parameter set appropriately, to couple different integration systems productively, and to couple the individual integral to other integrals and to shared cultural stores.

Joseph Henrich has argued, with substantial empirical support, that the distinctive human cognitive capacity is not raw intelligence but cultural learning, the ability to learn from other individuals with high fidelity, to accumulate socially transmitted knowledge, and to improve on what was learned. The human brain was shaped by selection for coupling to a cultural inheritance system. Selection was building a better receiver for cultural transmission, not just a better individual processor.

In the integral’s terms, what was under selection was the Jacobian between individual integrals and the cultural deposit, the susceptibility of individual parameter sets to inputs from the shared cultural store. Individuals with higher Jacobian entries for cultural inputs (who could absorb, integrate and build on culturally transmitted patterns more effectively) had higher fitness in the social environments that characterised human evolution. The brain evolved to be maximally receptive to the culture.

Evolution was selecting for the ability to connect individual integrals to each other and to the shared cultural store with increasing fidelity and bandwidth. The human brain is, above all, a cultural learning organ, optimised for resonance with other integrals. The pressure need not run only one way. As culture increasingly buffers the individual from the environment, it can also relax selection on the nervous system itself. The third loop both raises the value of a good receiver and shelters it, so the two effects run together rather than in a single direction.

One further consequence connects the evolutionary account to individual development. If what evolution was selecting for was the capacity to update the parameter set productively, to integrate new patterns rather than merely accumulate them, then it was also, implicitly, selecting for the capacity to recognise when the parameter set needed updating. This is the distinction between two modes of disequilibrium set out in the book’s developmental account: imposed disequilibrium, in which external events force the parameter set to update, and chosen disequilibrium, in which the organism or person recognises that its current shape has stabilised into a plateau and deliberately seeks the condition that will extend it. The great lives in Appendix B are distinguished, in part, by an unusually high capacity for chosen disequilibrium. The deliberate traveller, the voluntary exile, the thinker who seeks out the argument most likely to break their current model. That capacity is a biological inheritance: the evolved sensitivity to one’s own consolidation, and the drive to move beyond it.

Sources: Henrich — The Secret of Our Success (2015). Donald — Origins of the Modern Mind (1991).

4. The Symbolic Threshold: Art, Music, Dance, and the Pre-Linguistic Integral

Before language, there was resonance. The first coupling pathways between individual integrals were bodily, rhythmic, and aesthetic, not linguistic. Music, dance, and visual art appear in the archaeological record tens of thousands of years before anything that could be called written language, and the evidence suggests they were not decoration but functional coupling technologies, ways of synchronising the parameter sets of multiple individuals into shared resonance states.

Rhythm is the most primitive of these. The ability to entrain (to synchronise one’s bodily rhythms to an external pattern) is widespread in the animal kingdom, but the human capacity for rhythmic entrainment is extraordinary in its precision and flexibility. A human being can lock onto a rhythm, maintain it across changing contexts, and synchronise with other humans doing the same. This is not a trivial capacity. It requires a sophisticated integration system that can model the rhythm, predict its future course, and adjust motor output to match. But its significance is deeper than the mechanics. Rhythmic entrainment is a form of parameter set synchronisation. When people move together to a shared rhythm, their neural states converge and their integrals begin to resonate.

Music extends this from rhythm into the full frequency spectrum of human hearing. The emotional power of music (its ability to reliably induce specific affective states across diverse listeners) is not culturally arbitrary. Certain harmonic structures, rhythmic patterns, and timbral qualities produce convergent responses across human nervous systems because those nervous systems share an evolutionary history. Music exploits the resonance structure of the human parameter set. That structure likely did not evolve for music. It more likely evolved for auditory scene analysis (the work of picking a sound source out of a noisy world), and music is the cultural technology that found it and tuned it. A melody that moves you is one that activates what your integral has built through a lifetime of experience, and through the evolutionary inheritance of hundreds of thousands of years of human musical practice. The shiver down the spine at a particular chord progression is the felt experience of deep resonance: a pattern in the incoming signal that aligns with a shape built into the parameter set at multiple levels simultaneously.

Music may be the oldest resonance technology in the human lineage. Steven Mithen’s The Singing Neanderthals (2005) argues that a musical proto-language (holistic, manipulative, multi-modal, musical and mimetic) preceded the evolution of referential language. If correct, the coupling function of music predates the content function of language. Music was synchronising integrals across individual brains before propositional speech could update them. Alan Harvey’s neuroscientific account provides the anatomical grounding. Auditory pathways make multiple synaptic relays through the brainstem before reaching the cortex, giving music direct access to autonomic and arousal circuitry that other sensory channels bypass. That privileged route is not metaphor. It is anatomy.

Dance does the same through movement. The body is not a vehicle for the brain; it is, in Damasio’s account and in the framework, part of the integration machinery itself. The somatic parameter set (the settled history of bodily experience encoded in motor patterns, postural habits and affective-muscular responses) is as much part of the integral as the cognitive. Dance synchronises somatic parameter sets. When people dance together, they are inducing convergent states in each other’s somatic integrals. Ritual dance, from the most ancient shamanistic practices to contemporary club culture, functions as a technology for connecting somatic parameter sets across individuals, creating a shared state for a time.

Visual art adds the spatial and representational dimension. The cave paintings at Lascaux and Chauvet (some dating to more than 30,000 years ago) are sophisticated symbolic interventions in the visual parameter set, not primitive attempts at realistic depiction. Images designed to activate specific resonance states in viewers who share the relevant cultural context. The bison on the cave wall is not a bison. It is a pattern capable of inducing, in a prepared parameter set, the full resonance complex associated with the hunt, the animal, the sacred, and the communal. It is an early meme in the most direct sense, an image propagated through a medium to achieve resonance across multiple integrals.

What all of these pre-linguistic coupling technologies share is that they operate through resonance induction rather than propositional content. They do not transmit information in the linguistic sense (facts, rules, instructions). They induce states. They synchronise parameter sets. They create shared resonance states in multiple integrals simultaneously without requiring those integrals to share a common language. This is both their power and their limitation. They can achieve extraordinary depth of coupling. The shared resonance of a religious ceremony, a great musical performance or a collective dance can produce states of integration that language rarely matches. But they cannot transmit fine-grained content, accumulate propositional knowledge across generations, or build the kind of explicit, revisable, combinable cultural store that produces civilisation. For that, language is required.

Sources: Mithen — The Singing Neanderthals (2005). Dissanayake — Homo Aestheticus (1992). Tomasello — The Cultural Origins of Human Cognition (1999).

5. Language: The Decisive Upgrade

Language did not replace music, dance, and visual art as coupling technologies. It added a new dimension to them, one that transformed the integral from a powerful but limited resonance pathway into the engine of civilisational accumulation.

The puzzle that Blackmore and others have identified (“How could language, with its extraordinary metabolic cost and developmental complexity, have evolved through natural selection alone?”) dissolves once the question is reframed correctly. The question is not what fitness advantage individual words or grammatical structures conferred. The question is what language did to the integration machinery as a whole. And the answer is that language upgraded f(·).

Deacon’s account of co-evolutionary feedback between language and the symbolic capacity of the brain captures the same picture from the biological side. The upgrade was not language alone but the joint reshaping of neural anatomy and the symbolic capacity it could support, each enabling the other across hundreds of thousands of years.

In the integral formulation, f(·) is the moment function, the function that maps the inputs of each moment (choices and experiences) into their contribution to the parameter set. Before language, f(·) operated primarily on direct sensory and somatic experience: what was seen, heard, touched, felt, or done. Language extended f(·) to operate on symbolic representations of experience, on descriptions of things not present, events not witnessed, and possibilities not yet encountered. It allowed the integral to accumulate not just direct experience but reported experience, imagined experience, and the distilled experience of people long dead.

Language did not just give humans a new communication tool. It gave the integral a new class of input. Symbolic, transmissible, combinable, persistent across time. The upgrade was qualitative, and it landed not on the integrating function itself but on the class of input it could take and the coupling it could sustain: the ability to integrate experience at one remove, through symbolic representation, rather than only through direct sensory encounter.

This had three consequences that compounded each other.

The first was temporal extension. An integral before culture is bounded by the individual lifetime and the direct social group. What can be accumulated is what can be experienced or observed. Language allowed it to extend backward in time, through narrative, through oral tradition, through the settled wisdom of elders and ancestors transmitted in story form. The personal integral, connected to a linguistic cultural store, could integrate experience that predated its own existence. The effective time constant of the integration system expanded from a single lifetime to the depth of the oral tradition.

The second was combinability. Pre-linguistic coupling technologies (music, dance, ritual) synchronise whole states. They are holistic. Language is compositional. It can isolate specific elements of experience, name them, recombine them in novel ways, and transmit the recombination. This makes linguistic cultural accumulation productive in a way that pre-linguistic coupling is not. You cannot improve on a shared dance in the same way you can improve on a shared argument. The combinatorial structure of language (the fact that a finite vocabulary and grammar can generate an unbounded range of novel propositions) made the cultural set not just a store of past experience but a generative system capable of producing combinations that no individual integral had ever encountered directly.

The third was explicit revisability. Pre-linguistic culture is transmitted through resonance induction. It is absorbed whole or not at all. Linguistic propositions can be examined, questioned, revised, and improved before transmission. Once the cultural set became linguistic, it became a system that could criticise itself. This is the foundation of philosophy, science and law. All three are, in the integral’s terms, explicit revision applied to the cultural deposit. Athens was possible because the Greeks had not just language but a cultural practice of treating linguistic propositions as objects to be examined, challenged, and improved. The Socratic method is explicit revisability applied systematically to the most fundamental patterns a culture holds.

Sources: Deacon — The Symbolic Species (1997). Harari — Sapiens (2011). Christiansen & Chater — ‘Language as shaped by the brain’ (Behavioral and Brain Sciences, 2008).

6. Education as Resonance Transmission

Once language existed, the question of how the cultural deposit was transmitted became decisive. Genetic transmission is high-fidelity but zero-enrichment. What arrives is identical to what was sent, slightly shuffled. Pre-linguistic cultural transmission through resonance induction is high-depth but low-precision. The shared state is rich but the specific content is difficult to isolate and transmit reliably. Linguistic transmission added precision. But precision alone is not sufficient. The question is whether the transmitted pattern achieves resonance in the receiving integral, or merely sits in it as inert information.

This is the distinction between education in its deepest sense and mere instruction. Instruction transmits content. Education induces resonance. A student who has been instructed knows that Pythagoras’ theorem relates the squares of the sides of a right-angled triangle. A student who has been educated feels the theorem, understands why it must be so, recognises its presence in other contexts, and finds it resonating with other mathematical shapes already in the parameter set. The theorem has been integrated, not merely stored.

The difference is consequential for cultural transmission. A pattern that has merely been stored (memorised without integration) has a short effective time constant. It decays unless repeatedly rehearsed. A pattern that has achieved resonance, one that has activated related shapes in the parameter set and been taken into what is already there, has a long time constant. It becomes part of the anatomy rather than an addition to it. And when that student later transmits it to the next generation, they transmit not the inert fact but the resonance itself, the felt understanding that makes the next integration possible.

Education is the induction of resonance, the deliberate activation of shapes in a young integral that a mature integral has found stable and generative. The goal is that the student’s integral resonates at the shapes the teacher’s integral has found, not that the student knows what the teacher knows.

This reframes what the great educational institutions of history were actually doing. The Athenian gymnasium, the Platonic Academy, the Islamic madrasa, the medieval university, the scientific apprenticeship. These were resonance induction systems, not primarily information transmission systems. Environments carefully constructed to bring young integrals into extended contact with mature ones, to create the conditions under which deep resonance could develop across many dimensions at once. The physical environment mattered. The architecture of the school, the rituals of the seminar, the shared practices of the discipline. All of these were ways of creating the resonance conditions that made educational transmission possible.

The compounding consequence is this. When a mature integral transmits a pattern to a young one through resonance induction, the young integral does not merely receive it unchanged. It integrates it against its own settled experience, finds connections the teacher had not seen, and transmits a refined version to the next generation. Each transmission potentially enriches it. The cultural deposit does not merely persist across generations. It compounds. Each generation inherits the previous generation’s refined understanding as its starting point. This is how cumulative cultural evolution works. Not copying fidelity but resonance enrichment, generation after generation, producing understanding of increasing depth and sophistication.

The great intellectual traditions are not information archives. They are living resonance lineages, chains of transmission in which each generation integrates the inherited pattern against new experience and transmits a deepened version. The line from Socrates to Plato to Aristotle is a chain of integrations rather than of copies, each one finding new resonances in what it inherited, each one transmitting something richer than it received.

This also explains why some cultural inheritances are extraordinarily stable across millennia while others decay rapidly. Stability is a function of how reliably it achieves resonance across diverse parameter sets, not of copying fidelity. Mathematical truths, certain ethical principles, certain narrative structures, and certain musical forms persist across thousands of years and across cultural boundaries because they reliably resonate with the human parameter set, regardless of the specific shapes laid down by individual experience. The conditions for their resonance are reproduced by human experience itself, in every generation and every culture. They are preserved not by the transmission system but by the anatomy of the receiver. The sharpest demonstration is recent. When deaf children in Nicaragua were brought together in the 1980s with no shared language between them, they built one, and it acquired grammatical structure across the first two or three cohorts, each intake regularising what the one before had left loose. There was no deposit to inherit. The structure came out of the receivers.

Sources: Ericsson — Peak (2016). Whitehead — The Aims of Education (1929).

7. The Memetic Puzzle Resolved

The puzzle that has haunted memetics since Dawkins introduced the meme concept is a cost problem. Language requires a larynx. It requires the neural anatomy for symbolic reference, syntax, and working memory capable of holding a sentence structure while generating its content. Syntax is what makes the input combinable: a finite store of symbols yielding an unbounded set of structures, so that what is transmitted is not a list of things already thought but the means to assemble things never thought before. It requires years of intensive developmental learning before functional use. The cost is enormous. What was the selective advantage sufficient to pay it?

The standard answers (better social coordination, more effective mate-selection signalling, richer instruction for tool use) are plausible but feel underpowered. Groups with slightly better communication could cooperate slightly more effectively, but the magnitude of the advantage seems insufficient to explain the scale of the investment. Something more is required. Transmission on this scale needs time, and human childhood is extraordinarily long. A decade and more at maximum plasticity, producing nothing, is affordable to a group that raises children together, and to grandmothers who outlive their own fertility, but not to a mother alone. A large part of the genetic bill for the third loop was paid here, not in a language organ but in a schedule. The cost of the machinery was paid in genes. The cost of the content was not. Language changes far faster than genes can track, so the fit is better explained by language having been shaped to the receiver than by the receiver having been rebuilt for language. Meaning found its way through whether the genes adapted or not, which moves part of the bill to the third loop, and makes language the deposit’s first product as well as its cause.

The answer is clear. Language was selected for because it was a qualitative upgrade to the class of input it could take and the coupling it could sustain, the upgrade that activated the third loop. A phase transition in the sense of a threshold crossed by a gradually assembled, many-component system, not a single switch thrown at one moment. And the third loop, once started, is a runaway process. Each generation inherits a richer cultural deposit than the previous one, which means each generation starts from a higher base, which means the deposit grows faster with each generation. The advantage of being part of a culturally accumulating group rather than a group without this capacity is, over generational timescales, overwhelming.

Henrich’s account of cumulative cultural evolution supports this. Human groups with effective cultural transmission systems (groups that could accumulate and refine technological and social knowledge across generations) vastly outcompeted groups without this capacity. The advantage was in the accumulation process itself. A group that could reliably transmit and improve on inherited knowledge had a compounding advantage that grew with each generation.

Language did not evolve because any particular thing it allowed people to say was adaptive. Language evolved because it enabled the third loop, the compounding of the civilisational deposit across generations. The selective advantage was not the content but the anatomy. The ability to build a shared store that persisted beyond individual lifetimes and improved with each transmission.

The memetic puzzle dissolves completely in this frame. Memes (in the integral’s account, resonant patterns within neural parameter sets capable of transmission and replication) are its product and its vehicle at the same time, not the cause of cumulative cultural evolution. Language created the conditions under which memes could achieve the kind of purposeful, revisable, combinable transmission that makes cultural accumulation possible. And cultural accumulation created the selection pressure for better language, for greater precision, greater combinability, greater capacity for abstraction and symbolic reference. The two processes drove each other in a tight feedback loop across hundreds of thousands of years of human evolution.

Akbari and Reich’s ancient-DNA survey measures what the first loop was selecting on across exactly this period, and the answer is instructive: pigmentation, immunity and disease resistance, metabolism and body composition. Density, pathogen load, dairying, farming: the conditions the third loop had created. Selection did not slow as culture took over. It accelerated once farming began, and it moved on what the third loop exposed it to rather than on what the third loop carried.

Sources: Blackmore — The Meme Machine (1999). Henrich — The Secret of Our Success (2015). Richerson & Boyd — Not By Genes Alone (2005).

8. Writing and the External Integral

Language extended the temporal reach of the integral from a single lifetime to the depth of the oral tradition (perhaps 10 to 20 generations at most, with significant degradation of fidelity over that distance). Writing extended it indefinitely. A pattern committed to writing does not decay with the death of its carrier. It persists in the medium, available for integration by any integral that can read, across any distance of time and space.

Merlin Donald identifies external symbolic storage (the externalisation of memory into the environment) as the decisive transition in human cognitive evolution. Writing is the most powerful form of external symbolic storage. It creates a persistent, reviewable, revisable record of what would otherwise exist only in individual parameter sets and decay with them. The cultural deposit, once writing existed, became in principle unbounded in its temporal reach. What Aristotle produced in the fourth century BCE is available for integration by any literate human being in any century and any culture. The time constant of the deposit extended from decades to millennia.

The effect on the civilisational deposit was transformative. Before writing, it was limited by the capacity of human memory and the fidelity of oral transmission. Each generation could only inherit what could be held in living minds and transmitted through living relationships. Writing broke both constraints. The cultural store could grow without limit (no longer bounded by what any group of living people could remember) and could be transmitted with precision across distances of time and space that would have made oral transmission impossible.

This is what Alexander carried east. Not just personal military genius but the written deposit of Greek civilisation, the philosophical, scientific, literary and political inheritance that had been built up across several centuries and committed to texts that could travel. And it is what the Islamic scholars of the Golden Age preserved and extended when Greek texts arrived in Arabic translation, the written record of a civilisation, available for integration by parameter sets a thousand years and several thousand miles removed from the original authors.

The resonance between Harari’s account in Sapiens and the integral framework is direct and useful here. Harari’s cognitive revolution (the sudden explosion of human cultural complexity, long before writing but apparently associated with the development of full modern language) is, in the integral’s terms, the moment when the resonance pathways between individual integrals became powerful enough to produce a cumulative civilisation. There is a demographic version of the same explanation, and it needs nothing sudden to have happened in the head. Cumulative culture scales with the number of connected minds: raise the density and the interconnection and the deposit compounds; lower them and it is lost. The shared fictions that Harari identifies as the key human innovation (the ability to collectively believe in things that do not exist as physical objects, like gods, nations, money and human rights) are, in the integral’s terms, highly resonant memes capable of connecting vast numbers of individual integrals into shared states. They are the cultural equivalent of the musical and ritual resonance technologies that preceded language, but with the precision and combinability that language alone makes possible.

Harari’s shared fictions are the culture’s most powerful technologies. Patterns that achieve resonance across millions of individual parameter sets simultaneously, coordinating behaviour at scales that would be impossible through direct personal interaction. A nation, a religion, a currency. These are memetic patterns so broadly resonant with the human parameter set that they can organise the behaviour of people who have never met and will never meet. Civilisation is built from them.

Sources: Donald — Origins of the Modern Mind (1991). Harari — Sapiens (2011). Goody — The Logic of Writing and the Organisation of Society (1986).

9. The Compounding of the Civilisational Deposit

The picture that emerges from this analysis is of a process of compounding that has been underway for hundreds of millions of years, with three decisive accelerations, each corresponding to the activation of one of the three loops.

The first acceleration was the Cambrian. Complex nervous systems emerged and, with them, the first integrals, the second loop added to the first. Within-lifetime learning systems running in parallel with the slow loop of genetic evolution. From this point, variation in integration machinery was itself subject to natural selection. Better integrators out-competed worse ones.

The second acceleration was the pre-linguistic symbolic threshold. Music, dance, ritual and visual art emerged as technologies for connecting individual integrals into shared resonance states. This created the first cultural inheritance system, an inheritance that persisted in the shared practices of a group beyond any individual lifetime. The second loop began to operate collectively, before language made it fully compound. Civilisation began, in primitive form, with the first shared ritual around a fire, the first communal dance, or the first image scratched into a cave wall. Lynne Kelly’s cross-cultural documentation of oral knowledge technologies suggests the integrated package (music, rhythm, spatial anchoring, embodied performance) is universal across unconnected traditions because it rests on a shared genetic substrate.

The third and most powerful acceleration was language, the upgrade to f(·) that activated the third loop and made symbolic, compositional, revisable cultural accumulation possible. Language transformed the resonance connection of music and ritual into the propositional, combinable, explicitly revisable cultural store that could compound across generations without limit. Writing extended the time constant of this store from generations to millennia. The printing press extended its bandwidth from hundreds to millions. The internet extended it to billions simultaneously.

Merlin Donald reached this model from the other direction. Working from cognitive archaeology rather than mathematics, he identifies four representational stages in the evolution of the modern mind: the episodic mind of apes, which stores events but cannot deliberately recall or recombine them; the mimetic mind of Homo erectus, the first capacity to re-enact experience through voluntary bodily representation (gesture, rhythm, mime); the mythic mind of early Homo sapiens, with symbolic language and narrative; and the theoretic mind of the literate period, with external symbolic storage making the cultural record persistent and reviewable. His sequencing argument is important. Each stage is a qualitative break in representational capacity, not more content within an unchanged one.

He argues that symbols followed the cognitive advance rather than causing it. The capacity is rebuilt first, then its outputs become visible in the cultural record. The four stages map closely onto the three accelerations stated above. Mimetic is the pre-linguistic coupling of the second acceleration. Mythic is the third loop activating. Theoretic is writing and everything downstream. Donald could not have anticipated that the theoretic stage would extend into the LLM (he was writing in 1991), but his framing of external symbolic storage as a non-biological hardware change applies directly to that extension. The convergence between his account, developed from evolutionary evidence, and the formal model developed here from the mathematics of the integrand is one of the model’s strongest pieces of abductive validation. Two independent lines of inquiry arrived at the same shape.

Each acceleration tightened the feedback loops between individual integrals and the cultural deposit, increased the bandwidth of transmission, extended the time constant of accumulation, and made the compounding faster. The compounding has been accelerating for half a billion years. We are currently living through the latest and most dramatic acceleration. Artificial systems have integrated the whole cultural store and made it queryable at any point.

The gymnosophists on the banks of the Beas were not primitive. They were the product of a transmission lineage that had been refining human integrals for thousands of years, inheriting the settled philosophical and contemplative wisdom of the Indian subcontinent, integrating it against individual experience, and transmitting a refined version to students who would do the same. The difference between them was not genetic. They had a more deeply resonant shape of the human integral, transmitted and refined across many generations of deliberate education in how to run a human life well.

What passed between them and Alexander, whatever it was, triggered a moment of resonance, representing a different tradition of wisdom about how to be a person. That moment of resonance generated something. A grain that would travel west and continue to resonate across two and a half millennia. It is still running. This is what the integral of life may look like at civilisational scale: not a metaphor, not a philosophical conceit, but a possible account of the process by which human wisdom, once laid down, propagates through time, finding resonance in new parameter sets in every generation, compounding as it goes.

The Framework Illustrating Itself

The argument of this appendix is that knowledge compounds across generations through resonance transmission, with each generation integrating what it inherits and passing it on enriched. It can be demonstrated as an accumulation.

Darwin establishes the deep process: variation, selection and accumulation operating across generational time. That is the first loop. Thomas Henry Huxley receives Darwin’s pattern, integrates it against his own formidable scientific parameter set, and drives it into the civilisational deposit, championing evolution publicly, fighting the battles Darwin could not, ensuring it achieves the breadth of resonance a single scientist’s work rarely reaches alone. Without Huxley, Darwin’s integral would have remained a scientific deposit; with him, it became a civilisational one.

Julian Huxley (Thomas Henry’s grandson) inherited it and extended it. Writing in the mid-twentieth century, he applied the evolutionary framework to human civilisation and culture, anticipating by decades the biogeographical and cultural arguments that would later be developed with greater rigour and empirical depth. The transmission is literal. Grandfather to grandson, the inheritance passing through a family as well as through a field.

Jared Diamond receives it in extended form and adds the empirical grounding it lacked. Guns, Germs and Steel is a rigorous account of why the first loop produced different initial conditions in different places, of how geography, domesticable species and disease gradients determined which civilisations could compound fastest from the starting gun. Diamond does not cite Julian Huxley as a direct predecessor, but the line is recognisably continuous. Evolution applied to civilisational divergence, now with the data to sustain the argument.

Harari receives Diamond and extends him upward. Sapiens adds the cognitive and cultural layer, the role of shared fiction, collective belief and the civilisational deposit in enabling the coordination that produced modernity. Harari is describing the third loop operating at full scale, without the formal machinery to specify what it is or how it works.

Maynard Smith and Szathmáry, writing in 1995, had already seen the same sequence from a different angle. Their major transitions framework recast evolutionary history as a story of how information is stored, transmitted and interpreted, from replicating molecules through to human language and culture. They named the door. What they could not yet specify was what happens inside the individual mind after the language threshold is crossed. How information accumulates within a single life, how it couples through resonance, how it compounds across a lifetime and passes enriched into the civilisational deposit. That interior remained unmapped. They also stressed the first and third loops, while seeming to absorb the second loop into the transition to language. Jablonka and Lamb (2006) made the same objection from within evolutionary biology, arguing that the nervous system deserves to be counted a major transition in its own right, the enabler of behavioural and cultural heredity. They named the missing transition. The interior it opens is what this book has tried to model.

10. Key Sources and Further Reading

On gene-culture coevolution and dual inheritance

Boyd & Richerson — Culture and the Evolutionary Process (University of Chicago Press, 1985). Not By Genes Alone (University of Chicago Press, 2005). Henrich — The Secret of Our Success (Princeton University Press, 2015).

Akbari, Reich et al. — ‘Ancient DNA reveals pervasive directional selection across West Eurasia’ (Nature, 2026). The first direct measurement of first-loop change across the period in which the third loop compounded.

On the evolution of nervous systems and animal minds

Kandel — In Search of Memory (Norton, 2006). Godfrey-Smith — Other Minds (Farrar, Straus and Giroux, 2016) and Metazoa (Farrar, Straus and Giroux, 2020). De Waal — Are We Smart Enough to Know How Smart Animals Are? (Norton, 2016).

On the evolution of language and symbolic cognition

Deacon — The Symbolic Species (Norton, 1997). Mithen — The Singing Neanderthals (Harvard University Press, 2005). Donald — Origins of the Modern Mind (Harvard University Press, 1991). Kelly — The Knowledge Gene (Allen & Unwin, 2024).

On the civilisational consequences

Harari — Sapiens (Harvill Secker, 2011). Tomasello — The Cultural Origins of Human Cognition (Harvard University Press, 1999).