After the End of Language: Walden 3.1

(Skinner–Chomsky Debate Series, 3 part)

Abstract The history of Walden can be read as a sequence of increasingly radical attempts to escape what civilisation has made appear inevitable. Henry David Thoreau’s Walden asked what remained of human life when the individual withdrew from the unnecessary material and social complexity of nineteenth-century civilisation. A century later, B. F. Skinner’s Walden Two reversed the experiment: instead of escaping from an imperfect society, why not redesign the social environment itself? Walden 3.1 proposes a third experiment. What if one of civilisation’s most fundamental technologies—language—has itself become an environment from which humanity must partially withdraw? This article develops Ayolov’s hypothesis from The Miscommunication Trilogy and The Planned Obsolescence of Language that language should no longer be regarded as the timeless essence of Homo sapiens, but as a comparatively recent cultural technology acquired by an evolutionarily much older organism. Language enormously expanded human capacities for abstraction, memory, coordination and collective organisation, yet it simultaneously generated new forms of misunderstanding, ideological conflict, symbolic domination and estrangement from immediate embodied experience. The post-linguistic society imagined here therefore does not require the literal abolition of speech. It proposes something more plausible and potentially more consequential: the deliberate reduction of linguistic domination over human consciousness and social organisation. Large Language Models make this thought experiment historically relevant. If sophisticated linguistic behaviour can be reproduced by artificial systems operating through statistical prediction without possessing a human biological language faculty, language can no longer function unproblematically as evidence of an autonomous inner linguistic self. Following Elan Barenholtz’s metaphor of LLMs as mirrors, artificial intelligence reflects linguistic performance back towards humanity and raises an uncomfortable possibility: humans may have attributed to language forms of interior depth that belong neither to words nor necessarily to the mechanisms producing them. Walden 3.1 takes the next step. If machines become increasingly capable of assuming civilisation’s repetitive linguistic labour—documentation, translation, administration, classification and unlimited textual production—humans may paradoxically acquire an opportunity to become less linguistically saturated. The scarcity of the future may no longer be information but silence, attention, direct perception and unmediated experience. Yet the post-linguistic society cannot simply resurrect the mythology of autonomous man rejected by Skinner, nor can it escape the political problem established in the preceding analysis of tyranny: large populations still require coordination, and every organised environment influences behaviour. Walden 3.1 is therefore conceived not as a society without control, but as an experimental society that attempts to make control visible, revisable and empirically accountable while progressively transferring human experience away from compulsory linguistic mediation. Its ultimate ambition is neither to return humanity to prehistory nor to destroy language, but to recover dimensions of the prelinguistic human that linguistic civilisation has increasingly obscured. After millennia in which civilisation taught humans to inhabit words, the planned obsolescence of language begins with a different possibility: language may remain one of humanity’s greatest technologies without remaining the place in which humanity is compelled to live. Keywords Henry David Thoreau; B. F. Skinner; Walden; Walden Two; Walden 3.1; Noam Chomsky; Universal Grammar; language; planned obsolescence of language; Miscommunication Trilogy; prelinguistic cognition; Large Language Models; Elan Barenholtz; Peter Ayolov; autonomous man; behavioural design; artificial intelligence; linguistic minimalism; post-linguistic society; embodied experience; cultural technology; silence; human nature Introduction: Three Waldens and the Escape from Civilisation “Men build society and society builds men.” B. F. Skinner, Walden Two (1948) The history of Walden can be understood as a history of escape, but each Walden attempts to escape from something deeper than the one before it. Henry David Thoreau withdrew to Walden Pond in 1845 in order to discover what remained of human existence when the accumulated necessities of civilisation were stripped away. His experiment was directed against excess: too many possessions, too much labour, too many obligations, too many artificial wants separating the individual from immediate experience. A century later B. F. Skinner transformed this solitary experiment into a social one. Walden Two does not ask how an individual might escape an imperfect civilisation, but why an imperfect civilisation should be accepted in the first place. If human behaviour is shaped by environments, then environments themselves can become objects of experimental design. Education, work, family relations, incentives and social practices need not be inherited as sacred traditions; they can be evaluated by their consequences and altered when they fail. Walden 3.1 begins with this Skinnerian reversal but directs it towards something neither Thoreau nor Skinner could have encountered in its contemporary form: the saturation of human existence by language. The deepest artificial environment may not be the city, factory, market or state, but the symbolic environment humans carry continuously within themselves. Language does not merely describe civilisation; it makes large parts of civilisation possible. Laws, religions, ideologies, bureaucracies, nations, money, social identities and historical narratives depend upon collective systems of signs capable of coordinating strangers across enormous distances and periods of time. Yet Ayolov’s Miscommunication Trilogy reverses the traditional celebration of this achievement. The expansion of linguistic communication does not necessarily produce an equivalent expansion of understanding. It can generate ever more abstraction, interpretation, ideological incompatibility and competing verbal realities. The Planned Obsolescence of Language therefore asks a question that neither Thoreau’s retreat nor Skinner’s behavioural community could ask: what if simplification must eventually include language itself? Walden 3.1 does not imagine that language can simply be prohibited, nor does it romanticise an impossible return to an untouched prehistoric consciousness. Its premise is different. The human organism is evolutionarily older than the linguistic systems through which contemporary humans interpret themselves. Bodies perceived, moved, feared, desired, remembered and formed social attachments before there were written laws, philosophical concepts or theories of grammar. If the argument developed throughout this series is accepted, language should therefore be treated not as the metaphysical essence of Homo sapiens but as an extraordinarily powerful cultural technology that entered an already existing biological organism and gradually reorganised its consciousness and social world. Large Language Models make this proposition newly disturbing. Barenholtz’s metaphor of the LLM as a mirror suggests that artificial systems reflect language back towards humanity stripped of many qualities humans traditionally assumed must exist behind fluent speech. Prediction can produce linguistic coherence; linguistic coherence can produce the appearance of personality, intention and interiority. The historical irony may therefore be that machines become increasingly linguistic precisely when humans acquire the possibility of becoming less so. AI could assume expanding quantities of bureaucratic, classificatory, translational and repetitive symbolic labour while human beings deliberately reclaim domains of existence in which experience need not continuously be converted into words. But the political problem established in Evolution of Tyranny remains. Large societies cannot survive through silence alone. Coordination does not disappear when autonomous man is rejected, and behavioural influence does not disappear merely because a society calls itself free. Walden 3.1 must therefore combine Thoreau’s simplification with Skinner’s experimentalism while refusing the fantasy of a society without control. Its objective is not to eliminate control but to expose it; not to eliminate language but to limit its sovereignty; not to manufacture a final perfect society but to create institutions capable of recognising their own provisional character. Walden 3.1 begins where the previous three articles end: after Universal Grammar loses its privileged position, after the linguistic self loses its metaphysical sovereignty, and after the tyrant is discovered to be only the visible face of collective systems of control. The fourth Walden asks what civilisation might become once humanity ceases to treat the technologies that govern it as eternal properties of human nature. “Walden 3.1 does not promise to free man from his environment. It promises something more difficult: to make the environment visible, to make its design reversible, and to give the human animal places where neither the machine nor another human voice is required to speak.” Walden 3.1 Part I: Before the Word: Rediscovering the Prelinguistic Human “We do not choose a way of life in which we are free; we choose one in which we feel free.” B. F. Skinner, Walden Two (1948) Walden 3.1 begins with a simple chronological reversal: before there was a linguistic human being, there was a human animal, and before that animal there stretched an immense evolutionary history of organisms capable of surviving without words. Language arrived extraordinarily late when measured against the history of life, yet linguistic civilisation has gradually learned to interpret almost everything that preceded it through linguistic categories. Hunger becomes a concept, fear receives a name, relationships acquire definitions, experience becomes narrative, memory becomes autobiography and eventually the individual learns to describe an imagined interior entity called “myself”. The result is a peculiar inversion in which the younger technology comes to appear older than the organism using it. Ayolov’s hypothesis reverses this hierarchy. The prelinguistic human should not be imagined as an unfinished linguistic subject waiting impatiently for grammar, but as an embodied organism already possessing immensely sophisticated systems of perception, affect, movement, attachment, recognition and non-verbal memory. Language subsequently entered this older architecture through other humans. Nobody chooses the first language that colonises consciousness; its sounds, distinctions, categories and inherited associations arrive through the surrounding verbal community until what originally came from outside begins to be experienced as an internal voice. Walden 3.1 therefore does not seek the absurd goal of deleting language from the adult brain or reconstructing an inaccessible prehistoric consciousness. It proposes something more modest and more radical: reducing the compulsory linguistic mediation of experience. Silence need not indicate ignorance. An emotion need not immediately receive a diagnostic or psychological label. A landscape need not become a caption. A memory need not become a story. Education need not transform every discovery into verbal examination. The body does not require a commentary in order to exist. The first principle of the post-linguistic society is therefore subtraction: wherever words add no necessary function to an experience, their absence should again become possible. This is where artificial intelligence creates a historical opportunity that earlier projects of simplification could not possess. Industrial technology was expected to liberate humans from physical labour, yet frequently created new forms of labour; information technology promised liberation from ignorance while producing unprecedented informational saturation. LLMs introduce a different possibility because they automate precisely the activity civilisation has treated as uniquely and almost sacredly human: linguistic production. The machine can summarise, translate, classify, explain, rewrite, document and generate endless sequences of plausible words. Barenholtz’s mirror therefore presents humanity with an uncomfortable reflection. If a predictive system can perform increasing quantities of linguistic labour without possessing a human autobiographical self behind every sentence, then linguistic performance can no longer serve automatically as evidence of the mysterious interior being that produced it. The Planned Obsolescence of Language takes this observation in an unexpected direction. Why compete with machines to become still more linguistic? Why educate humans for decades to perform symbolic routines that machines can increasingly execute faster? Walden 3.1 would instead transfer appropriate repetitive linguistic functions towards machines while protecting specifically embodied human time from the resulting flood of machine-generated language. This qualification is essential: AI should not become an infinitely speaking oracle filling every silent moment with generated text. That would represent the final victory of language rather than its obsolescence. The post-linguistic society would use AI partly to remove words from human life—to compress bureaucracy, reduce repetitive documentation, translate when translation is necessary and make information accessible without demanding continuous linguistic labour from everyone. Human education could consequently restore greater importance to observation, craft, movement, music, visual reasoning, experimentation, spatial intelligence and direct encounters with material reality. Language would remain indispensable, but it would lose its monopoly over intelligence. Walden 3.1 therefore proposes no return to the cave. It proposes the deliberate rediscovery, within a technologically advanced civilisation, of capacities that were never created by language in the first place. The end of language means neither muteness nor ignorance. It means the end of the linguistic occupation of every available territory of human experience. “The first mistake of the old civilisation was to believe that whatever could not be said could not be known. In Walden 3.1, children learn the opposite: some things must be experienced before words arrive, and some things are damaged when words arrive at all.” Walden 3.1 Part II: Walden 3.1: Designing the Environment After Autonomous Man “We can achieve a sort of control under which the controlled, though they are following a code much more scrupulously than was ever the case under the old system, nevertheless feel free.” B. F. Skinner, Walden Two (1948) The most disturbing proposition inherited from Skinner is not that human beings can be controlled, but that they are already being controlled. Every society constructs environments that make certain behaviours easier, more rewarding, more prestigious or more dangerous than others. Architecture directs movement; schools establish acceptable forms of knowledge; markets organise incentives; laws attach consequences to behaviour; interfaces organise attention; media systems determine visibility; algorithms decide which fragments of reality appear first. The mythology of autonomous man conceals these processes by placing the apparent origin of behaviour inside an independent self: individuals choose, believe, desire and decide, while the environmental architecture surrounding those choices disappears from view. Walden 3.1 begins by rejecting this comforting fiction without accepting the opposite fantasy of total social engineering. If environments inevitably influence behaviour, the political question becomes who designs them, according to which objectives, with what evidence and under what conditions they can be changed. This is where the argument of Evolution of Tyranny becomes indispensable. Dictatorship was revealed not merely as the will of one dictator but as a distributed accomplishment reproduced through institutions, administrators, experts, technologies and ordinary compliance. Walden 3.1 therefore cannot defeat tyranny merely by abolishing the tyrant. Nor can it pretend that democracy eliminates behavioural control simply because authority possesses an electoral justification. The alternative is to make the architecture of influence visible. Institutions must state what behaviours they are attempting to encourage, what consequences they expect, what evidence would count as failure and when an experiment must be abandoned. Social arrangements cease to be sacred inheritances and become provisional designs. Skinner’s experimentalism is preserved, but its most dangerous temptation—the possibility that enlightened planners might permanently define the good life for everybody else—must be rejected. No expert class can be permitted to transform provisional knowledge into unquestionable authority. For this reason the foundational political principle of Walden 3.1 would be reversibility. A society genuinely organised according to empirical principles must possess mechanisms for admitting that its experiments have failed. Policies should therefore be treated less like commandments and more like hypotheses: limited where possible, measured by consequences, exposed to criticism and capable of being withdrawn. Artificial intelligence could make such evaluation more sophisticated, but it simultaneously creates the greatest danger. The algorithmic sovereign imagined in the previous article could observe populations, predict behaviour and continuously optimise environments until political control became almost invisible. Walden 3.1 must deliberately prevent experimental governance from becoming precisely this cybernetic tyranny. Prediction cannot confer moral authority. Efficiency cannot define the common good. A model capable of forecasting behaviour does not acquire the right to determine behaviour. Protected zones of opacity, privacy, dissent and non-participation therefore become essential precisely because a technologically advanced society could increasingly eliminate them. Even linguistic minimalism must remain voluntary rather than becoming another compulsory doctrine: a state that prohibits language in order to liberate people from language would merely create a new linguistic orthodoxy enforced through silence. The paradox at the centre of Walden 3.1 is consequently unavoidable. Humans cannot live outside environments, but environments need not become prisons; societies cannot exist without coordination, but coordination need not justify unlimited authority. The goal is not a perfect system engineered once and preserved forever. Perfection would itself be dangerous because a political order convinced of its perfection would lose the capacity for correction. Walden 3.1 is instead permanently unfinished—hence the deliberately provisional version number 3.1. There must always be the possibility of 3.2. Its deepest institutional principle is therefore not obedience to the designer but the planned obsolescence of the design itself. The society that understands behavioural control most clearly must also be the society most willing to dismantle its own mechanisms when their consequences become destructive. “The old tyrant said: obey because I command you. The new machine said: obey because the data prove this is best for you. Walden 3.1 distrusts them both. No prediction is a commandment, no designer is permanent, and every system built by human beings must contain the instructions for its own replacement.” Walden 3.1 Part III: After the Tower of Babble: The Planned Obsolescence of Language “The simple fact is that we must delegate control of the population as a whole to specialists—to police, priests, teachers, therapies, and so on, with their specialized reinforcers and their codified contingencies.” B. F. Skinner, Walden Two (1948) Civilisation has responded to the failures of communication with remarkable consistency: whenever communication fails, it produces more communication. Misunderstanding demands explanation; political disagreement generates speeches and counter-speeches; bureaucracy answers complexity with additional documents; education answers uncertainty with more terminology; media multiply interpretations; digital networks promise universal connection by placing billions of people inside a permanent exchange of signs. The result is the paradox developed throughout Ayolov’s Miscommunication Trilogy: the quantity of communication can increase while common understanding does not. The Tower of Babel should therefore be inverted. Its contemporary meaning is not that humanity once possessed a perfect language and catastrophically lost it, but that linguistic civilisation eventually becomes buried beneath the abundance of its own words. The digital world accelerates this process because almost every action produces linguistic or symbolic residue—messages, comments, descriptions, classifications, instructions, metadata, opinions, explanations and responses to responses. LLMs transform abundance into something approaching infinity. For the first time in human history, linguistic production no longer requires equivalent human time. A machine can generate in seconds quantities of text that once demanded days or months of human labour. This is the decisive historical reversal behind The Planned Obsolescence of Language. For most of civilisation, words were scarce and expensive. Scribes copied them, monasteries preserved them, printers multiplied them, schools taught populations to produce and interpret them. The twenty-first century approaches the opposite condition: words become practically inexhaustible while the human attention capable of receiving them remains biologically finite. The scarce resources of post-linguistic civilisation are therefore silence, attention, presence and direct experience. The problem is no longer merely how to give everybody access to information. It is increasingly how to protect the human organism from informational environments capable of producing more language than any human being could possibly consume. Walden 3.1 consequently proposes planned obsolescence rather than linguistic destruction. Obsolescence occurs when a technology remains possible but ceases to be necessary for functions that newer arrangements can perform more efficiently. Language would survive as literature, conversation, scientific notation, historical memory, humour, poetry and voluntary expression while gradually losing its compulsory occupation of activities for which it has become redundant. AI could translate without requiring everybody to share a dominant language, compress bureaucratic documents into actionable decisions, retrieve information without forcing individuals to memorise verbal inventories and mediate repetitive administrative exchanges that currently consume enormous quantities of human attention. Yet the greatest danger is obvious: instead of liberating humanity from linguistic saturation, generative AI could complete it. A machine capable of speaking endlessly can become the ultimate Tower of Babble, filling every classroom, workplace, street, screen and private moment with personalised synthetic language. Walden 3.1 must therefore impose its principle in the opposite direction: technological progress should be measured partly by how much unnecessary mediation it removes. The successful AI assistant would sometimes produce no words at all. The successful institution would require fewer forms, fewer explanations, fewer slogans and fewer compulsory declarations of identity. The successful school would know when explanation should end and observation begin. This is not an attack upon knowledge but upon the confusion of knowledge with verbal accumulation. It also changes the political problem inherited from the previous article. The most powerful future tyranny may not silence populations; it may overwhelm them with language, narratives and artificial choices until distinguishing experience from interpretation becomes increasingly difficult. A post-linguistic society must therefore protect silence with the seriousness with which earlier societies protected speech. The final liberation from the Tower of Babble is not the discovery of the perfect language. It is the recognition that humanity does not need to answer every word with another word. “The old civilisation feared silence because silence escaped its categories. It filled the world with books, laws, advertisements, explanations, screens and finally machines that could speak without end. Walden 3.1 made the opposite discovery: when words became infinite, silence became the last scarce human resource.” Walden 3.1 Conclusion: Walden 3.1 — The Society After the Word “We are only beginning to understand the power of the environment in shaping human behavior.” B. F. Skinner, Walden Two (1948) The Skinner–Chomsky debate began with an apparently specialised question about language acquisition, yet beneath it stood a far more consequential disagreement about what a human being is. Chomsky’s revolution relocated language inside the organism, making linguistic creativity evidence for an innate architecture that could not be explained adequately through environmental contingencies. Skinner travelled in the opposite direction, dissolving the autonomous originator of behaviour into a biological organism continuously shaped through interaction with its environment. More than half a century later, LLMs have unexpectedly reopened this old battlefield. They do not prove that human beings acquire language exactly as machines do, nor do they settle the biological question of innateness, but they destroy one comfortable philosophical inference: sophisticated linguistic performance cannot by itself demonstrate the existence of the deep autonomous linguistic being traditionally imagined behind it. Prediction can generate grammatical novelty. Statistical regularities can produce astonishing linguistic competence. A machine without childhood, human embodiment or a Chomskyan Universal Grammar can nevertheless enter the symbolic world and speak from within it. Barenholtz’s mirror therefore reflects something potentially disappointing but intellectually liberating back towards humanity. Perhaps too much metaphysical significance was attributed to language because humans encountered language only in other humans and consequently confused the properties of linguistic behaviour with the properties of the organism producing it. Ayolov’s Miscommunication Trilogy and The Planned Obsolescence of Language push this possibility towards its radical conclusion. If language is a historically accumulated cultural technology rather than the timeless essence of Homo sapiens, then humanity need not remain permanently organised around maximum linguistic production. The evolutionary sequence can be reversed conceptually: underneath the linguistic self remains the vastly older biological organism—the perceiving, moving, desiring, fearing, touching and remembering animal upon which words were subsequently imposed. Walden 3.1 therefore does not announce the literal death of language. It announces the end of language as sovereign. The post-linguistic human can continue speaking while refusing to believe that everything valuable within human experience must first become a sentence. This fourth Walden also completes the political argument of the series. Once autonomous man is abandoned, the fantasy of a society without behavioural influence becomes difficult to sustain. Every civilisation structures environments, establishes consequences, distributes resources, controls access and creates systems of coordination. The previous article proposed that dictatorship itself can be understood partly as an organisational phenomenon appearing at different scales: tyranny is a collective effort disguised by the spectacular visibility of the tyrant. Walden 3.1 therefore cannot promise liberation through the simple destruction of authority. Nor should it disguise concentrated power behind a new benevolent façade. Its alternative is experimental, visible and reversible coordination. Skinner’s crucial insight is retained—that environments can be designed—but combined with a principle that protects society against the designers themselves: no expert, institution, algorithm or political arrangement should become immune to revision. AI makes such experimental coordination technically imaginable while simultaneously making its authoritarian opposite terrifyingly possible. A system capable of observing populations, modelling behaviour and adjusting informational environments could become the most efficient machinery of control ever constructed. Walden 3.1 therefore places limits precisely where technological capability would encourage their removal. Privacy, opacity, dissent, voluntary silence and the right to remain behaviourally unpredictable become political values. The goal is not to create perfectly predictable human beings. A perfectly predictable society would be a dead society. The objective is to reduce avoidable catastrophe without extinguishing the variability from which adaptation, creativity and change emerge. This is why Walden 3.1 must remain 3.1 rather than Walden Final. There must always be a 3.2, a 4.0 or an abandonment of the experiment altogether. Thoreau asked what could be removed from individual life. Skinner asked what could be redesigned in collective life. Walden 3.1 asks whether civilisation can finally redesign its relationship with the technology through which it has spent millennia explaining itself to itself. The answer is neither compulsory silence nor another linguistic utopia. It is the restoration of language to the status of a tool: extraordinary, dangerous, useful and replaceable. Humanity does not have to stop speaking. It has to stop mistaking the voice inside its head for the whole of the human being. Even Charles Darwin famously argued that the term "species" is an arbitrary label used for our own convenience. In his 1859 book On the Origin of Species, he famously wrote: "I look at the term species as one arbitrarily given, for the sake of convenience, to a set of individuals closely resembling each other." By treating "species" as a convenient fiction rather than a cosmic truth, he solved a massive headache for 19th-century naturalists. Before On the Origin of Species, scientists spent careers arguing over whether a specific bird was its own distinct species or just a local variety. Darwin pointed out that these arguments were endless because the boundary lines simply do not exist in nature. This includes the specie of Homo sapiens, of course. In recent decades, modern genetic science and fossil discoveries have vindicated Darwin's view of Homo sapiens as a fuzzy, poorly defined box. “Walden 3.1 was never built to become the final society. A final society would be another prison with a perfect name. It was built so that every institution could be changed, every machine could be switched off, every word could be questioned, and every human being could sometimes exist without explaining what they were. The old world searched for freedom inside language. We discovered something older: the freedom, occasionally, to have no words at all.” Walden 3.1 References Ayolov, P. (2026) The Miscommunication Trilogy. Unpublished manuscript. Ayolov, P. (2026) The Planned Obsolescence of Language. Unpublished manuscript. Barenholtz, E. (2025) ‘Large language models as mirrors of human language and cognition’. Center for the Future Mind, Florida Atlantic University. Chomsky, N. (1959) ‘A review of B. F. Skinner’s Verbal Behavior’, Language, 35(1), pp. 26–58. Chomsky, N. (1965) Aspects of the Theory of Syntax. Cambridge, MA: MIT Press. Chomsky, N. (1971) ‘The case against B. F. Skinner’, The New York Review of Books, 30 December. Chomsky, N. (1995) The Minimalist Program. Cambridge, MA: MIT Press. Chomsky, N., Roberts, I. and Watumull, J. (2023) ‘The false promise of ChatGPT’, The New York Times, 8 March. Clark, A. (1997) Being There: Putting Brain, Body, and World Together Again. Cambridge, MA: MIT Press. Clark, A. (2008) Supersizing the Mind: Embodiment, Action, and Cognitive Extension. Oxford: Oxford University Press. Deacon, T.W. (1997) The Symbolic Species: The Co-evolution of Language and the Brain. New York: W.W. Norton. Guriev, S. and Treisman, D. (2022) Spin Dictators: The Changing Face of Tyranny in the 21st Century. Princeton, NJ: Princeton University Press. Hobbes, T. (1651) Leviathan, or The Matter, Forme, & Power of a Common-Wealth Ecclesiasticall and Civill. London: Andrew Crooke. Kuhn, T.S. (1962) The Structure of Scientific Revolutions. Chicago: University of Chicago Press. Márquez, X. (2025) ‘Ancient tyranny and modern dictatorship’, The Review of Politics, 87(1), pp. 67–92. Saffran, J.R., Aslin, R.N. and Newport, E.L. (1996) ‘Statistical learning by 8-month-old infants’, Science, 274(5294), pp. 1926–1928. Skinner, B.F. (1948) Walden Two. New York: Macmillan. Skinner, B.F. (1953) Science and Human Behavior. New York: Macmillan. Skinner, B.F. (1957) Verbal Behavior. New York: Appleton-Century-Crofts. Skinner, B.F. (1971) Beyond Freedom and Dignity. New York: Alfred A. Knopf. Skinner, B.F. (1974) About Behaviorism. New York: Alfred A. Knopf. Skinner, B.F. (1981) ‘Selection by consequences’, Science, 213(4507), pp. 501–504. Skinner, B.F. and Segal, E.F. (1988) Discussion of verbal behavior. Interview conducted at Harvard University, February 1988. Thoreau, H.D. (1854) Walden; or, Life in the Woods. Boston: Ticknor and Fields. Tomasello, M. (1999) The Cultural Origins of Human Cognition. Cambridge, MA: Harvard University Press. Tomasello, M. (2003) Constructing a Language: A Usage-Based Theory of Language Acquisition. Cambridge, MA: Harvard University Press. Varela, F.J., Thompson, E. and Rosch, E. (1991) The Embodied Mind: Cognitive Science and Human Experience. Cambridge, MA: MIT Press. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, Ł. and Polosukhin, I. (2017) ‘Attention is all you need’, Advances in Neural Information Processing Systems, 30, pp. 5998–6008. West, G.B. (2017) Scale: The Universal Laws of Life, Growth, and Death in Organisms, Cities, and Companies. New York: Penguin Press. Zuboff, S. (2019) The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. New York: PublicAffairs.