Lost Safety: Education in a World That No Longer Requires Knowledge
Since the neoliberal turn of the 1980s, a new settlement has gradually emerged in which extensional forms of educational organisation are not an accidental byproduct but are functionally convenient for a stratified society. Once we distinguish extensional, intensional, and hyperintensional forms of thought with some precision, it becomes possible to argue that what has been eroded with the shrinking and disciplining of the middle class is not only material security but a whole cognitive form of life. The old welfare state and postwar settlement sustained, however unevenly, a broad social zone in which people could expect enough stability, time, institutional trust, and public investment to cultivate forms of judgement that exceeded immediate performance. As that settlement has been hollowed out, the social basis of intensional and hyperintensional life has also been hollowed out. That is my hypothesis here.
To make this clear, the terms need to do real work. Extensional thinking is the simplest. It is concerned with outputs, visible results, correct answers, measurable equivalences. If two pupils receive the same mark, or two answers generate the same officially recognised outcome, an extensional system tends to treat them as educationally equivalent.
Intensional thinking is more demanding. It is concerned not merely with whether two statements or performances come out the same, but with how they are presented, how they are reached, what conceptual route is being taken, and what would happen if things were slightly different.
Hyperintensional thinking is more demanding again. It distinguishes between things that may be extensionally and even intensionally equivalent, yet differ in their grounds, their internal structure, and the kinds of reasons that support them. Two students may reach the same conclusion about a poem, or a theorem, or a historical event, but one may have done so through a rich web of conceptual relations while the other has merely assembled the right surface cues. An extensional system sees sameness of result. A hyperintensional one sees a difference in the very being of the judgement.
That sounds abstract until one notices that the broad architecture of performative schooling is almost perfectly designed to privilege extensional equivalence. Modern accountability systems need outputs that can be compared across schools, regions, cohorts, and populations. They need indicators, profiles, dashboards, interventions, and performance data. Current English school accountability reforms explicitly integrate school performance data with point in time report cards and school profiles, which shows how central data visibility and comparability remain to system design. OECD work on test based accountability similarly defines accountability in terms of using student results to assess schools’ and teachers’ performance, which is already a clue that what matters institutionally is not primarily the structure of understanding but the measurable output that can enter a governance loop.
Now once a society builds schooling around those extensional demands, something deeper happens. The school ceases to be primarily a place where one is inducted into thick structures of judgement, interpretation, and reason giving, and becomes instead a site for producing commensurable performances. The task is no longer, or no longer mainly, to cultivate a mind that can move among different presentations of the same object, distinguish equivalent from non equivalent formulations, test grounds, handle ambiguity, and sustain inference through uncertainty. The task is to generate acceptable outputs under conditions of audit. This is why so much schooling now feels less like an education and more like a training regime in recognisable response forms. The mark scheme, the success criterion, the exam technique, the model answer, the scaffold, the feedback code, the retrieval quiz, the intervention cycle, the data drop, these are not incidental features. They are the institutional grammar of extensionality.
The disturbing point is that such a system progressively removes the need for epistemic safety. Safety, in epistemology, means that one’s belief could not easily have been false. Put in educational language, it means that a pupil’s way of arriving at an answer remains true across nearby variations. A pupil who understands a mathematical principle in a robust way will survive small alterations in form, context, and framing. A pupil who has merely learned the pattern will fail when the surface cues change. Knowledge, on this account, requires modal resilience. It requires the ability to remain right when things are slightly different. But performative schooling increasingly narrows the range of relevant variation. It teaches children to succeed in a highly curated environment where what counts as success is known in advance, where acceptable forms are modelled, where ambiguity is trimmed back because it threatens reliability, and where the surrounding apparatus is designed to reduce surprise. The result is not exactly the fake barn county in its original form, but an inversion of it. In fake barn county the environment is saturated with misleading alternatives, so true belief becomes unsafe because one could easily have been wrong. In performative schooling the environment is purified of many alternatives, so unsafe cognition can pass for knowledge because the situations that would expose it never arise. The pupil is kept in a world where extensional correctness is enough.
This matters politically because the old middle class, especially in the postwar welfare state imagination, was not just an income band. It was also a social formation with a certain relation to time, language, institutions, and culture. However mythologised that formation may have been, it depended on public systems that at least aspired to produce people capable of judgement, deliberation, interpretation, and self direction. It depended on libraries, universities, civic institutions, publicly funded arts, relatively secure professions, a thicker press culture, and a broader legitimacy for forms of knowledge not reducible to immediate market value. It was, in that sense, an intensional and often hyperintensional class. Not because all its members were philosophers, but because its social reproduction required sensitivity to nuance, mediation, delayed judgement, professional discretion, and conceptual handling of complexity. The teacher, civil servant, editor, solicitor, academic, architect, public broadcaster, local government officer, social worker, librarian, and union organiser inhabited institutional worlds in which the grounds of a claim mattered, not merely the visible output. One had to interpret, justify, discriminate, argue, and often revise.
What has happened since the 1980s in many high income countries is that the economic and political foundations of that world have been eroded. OECD analysis has documented a long squeeze on the middle class, stressing both its role in sustaining social protection and the pressures placed on it by slower income growth, higher housing and education costs, and wealth concentration, with the top ten percent holding half of total wealth across OECD countries. World Bank analysis likewise notes marked rises in inequality in countries such as the United Kingdom and United States from the 1980s and early 1990s onward, a temporal pattern that fits the wider neoliberal shift.
I am not saying those reports use my philosophical vocabulary. They do not. But they describe the material reorganisation within which such a vocabulary becomes illuminating. As material security erodes, the social space for intensional life erodes with it. People with less time, less institutional confidence, and more exposure to economic precarity are pushed towards short horizon optimisation. Schools under more pressure become more performative. Universities become more managerial. Professional judgement is replaced by targets, indicators, competencies, frameworks, and evidencing. The social world itself becomes more extensional. That is the stronger thesis. Neoliberalism is not just a set of market reforms. It is an extensional regime of social organisation. It does not abolish knowledge. It reorganises value so that what can be counted, displayed, benchmarked, and traded gains priority over what requires thick judgement.
That is why credentialisation expands even as genuine epistemic security often declines. The credential is an extensional object par excellence. It is a visible token that stands in for a socially recognised capability. It allows institutions to sort, compare, and allocate without entering the dense internal structure of what someone actually knows. The more unequal and competitive the society becomes, the more such tokens matter, because institutions need quick selection devices. At the same time, their inflation is almost inevitable, because as more people seek security through credentials, more credentials are required to distinguish them. So society becomes both more credential obsessed and less genuinely knowledge centred. The contradiction is not an accident. It is built into an extensional order.
At this point the class fissure becomes sharper. The elite do not need to choose between extensional and intensional education in a simple either or form. They can have both. Their children can be trained to succeed in formal systems while also inhabiting environments saturated with intensional and hyperintensional cultivation. Small seminars, tutorial conversation, travel, museums, music lessons, serious reading, dinner table argument, internships mediated by social capital, adult attention, confidence in questioning, foreign languages, unhurried holidays, and schools or informal spaces where ambiguity is not immediately punished, these form a parallel curriculum of deep structure. Even where elite schools outwardly participate in the same assessment regime, they often surround it with richer ecologies of discourse. The non elite, by contrast, are more likely to receive only the extensional shell. They are taught to perform school, to satisfy rubrics, to accumulate credits, to navigate bureaucratic checkpoints. What is withheld is not information alone, but access to forms of life in which reasons can be explored beyond immediate performative demand.
Holidays are not trivial. Holiday here should not be understood merely as leisure but as suspension of extensional compulsion. It is time and space in which thought can move without immediate audit. For some children, holidays are full of books, conversation, travel, boredom that turns into inquiry, visits, theatre, historical sites, family friends in professions, and the tacit absorption of a world in which concepts connect. That is a hyperintensional factory in all but name. Idling is another name for this. For others, holiday means economic stress, limited space, entertainment as distraction, and very little access to the institutions or adult attention that would thicken conceptual life. So the class divide is reproduced not only in school funding or exam outcomes but in the distribution of possible worlds, in who gets to inhabit environments where thought is allowed to ramify.
Once this is seen, authoritarianism itself can be interpreted through the same lens. Authoritarian politics is not merely intolerance or state force. It is often a hostility to intensional and hyperintensional life, because such life slows command, complicates slogans, and multiplies grounds. Authoritarian cultures prefer extensional obedience, visible loyalty, measurable conformity, binary classification, and simplified narratives. They dislike institutions whose work depends on ambiguity, contestation, and professional judgement independent of central command. That is why attacks on universities, arts, critical media, teacher autonomy, civil service discretion, and welfare professions so often accompany authoritarian turns. These institutions are habitats of intensionality. They preserve spaces where one can ask not only whether a claim works politically, but what exactly it means, what follows from it, whether two equivalent sounding positions differ in structure, whether the grounds are sound, whether the rhetoric hides an invalid move.
In this sense the decline of the middle class is connected to the decline of an intensional and hyperintensional class fraction, those social layers whose existence once depended on mediating institutions rather than raw capital or raw command. Schooling then becomes a crucial battleground. If one wanted, not necessarily by secret design but by structural preference, a population more governable under conditions of inequality, one would favour an education system in which extensional success can be widely distributed while intensional depth remains scarce. One would favour constant testing, visible accountability, competence statements, narrow notions of “what works,” behavioural compliance, and increasingly technological assisted production of acceptable outputs. One would not abolish aspiration. On the contrary, one would intensify it through creditisation. Everyone must chase qualifications, badges, certificates, portfolio items, employability markers, micro credentials. That keeps populations invested in the game while leaving untouched the deeper inequality in access to knowledge rich environments. The social order can then say, with perfect ideological sincerity, that opportunity has been expanded, because more people possess more credentials than ever before.
Yet what has expanded is extensional certification rather than access to forms of judgement that allow genuine autonomy. There is evidence, even in mainstream policy literature, that standardised assessment captures only a fraction of what matters in life and learning. OECD materials explicitly note that performance in standardised assessments explains only a fraction of later life outcomes and that wider personal and social attributes matter, though they are harder to measure. But I think the very difficulty of measuring such things becomes a reason systems abandon them or simulate them thinly. Once the measurable becomes sovereign, the immeasurable either disappears from the official curriculum or survives only for those insulated enough to pursue it privately.
AI makes the whole tendency more dangerous. AI is an extensional accelerant. It is superb at generating outputs that satisfy formal criteria. It can produce the essay, the summary, the revision notes, the answer, the tone, the format. For systems already disposed to judge by output, AI is a gift. It allows the extensional regime to intensify while appearing innovative. But from a safety perspective, and from an intensional or hyperintensional perspective, this is catastrophic unless resisted. The student may possess the output without the route, the proposition without the grounds, the answer without the modal resilience that would make it knowledge. If elite settings respond by doubling down on oral defence, tutorials, open ended projects, and humanly dense intellectual life, while mass systems rely ever more on managed outputs and surveillance, then AI will widen not merely a skills gap but an ontological gap in the forms of mind being produced.
So my thesis becomes this. Since the neoliberal turn of the 1980s, the erosion of welfare state institutions and the disciplining of the middle class have not only redistributed wealth and risk. They have also reorganised the cognitive ecology of society. Public schooling has increasingly been remade as an extensional apparatus of performative output, precisely because extensional systems are easier to govern, audit, compare, and align with unequal labour markets. The old middle strata were partly carriers of intensional and hyperintensional forms of life, because their institutional roles required mediation, interpretation, and thick professional judgement. As those strata are hollowed out, and as authoritarian forms rise, the social demand shifts towards compliance, visible performance, and simplified cognition. The elite tolerate or even welcome this, because their own children still have access to protected spaces where richer structures of thought are cultivated. The result is a society in which extensional education is massified while intensional and hyperintensional education is increasingly privatised.
That is a structural class theory of educational form. And if it is even partly right, then the fight over schooling is much deeper than a debate about standards or curriculum content. It is a fight over whether knowledge, in the demanding sense of safe, grounded, conceptually differentiated understanding, remains a public good at all, or becomes a privately hoarded class resource.
But the thesis doesn't end there. Though elites thought they may somehow have secured a permanent exemption from the extensional world they have helped normalise the media and political ecology now developing is corrosive enough to begin dissolving even the elite protected zones that once reproduced intensional and hyperintensional forms of life.
The original development was that broad publics could be managed through increasingly performative institutions, while elite families, elite schools, elite universities, and elite cultural circuits could still reproduce slow judgement, conceptual nuance, historical memory, and the ability to move among different levels of abstraction. But social media, platform capitalism, and now generative AI threaten to degrade even those enclaves, because they do not merely distribute content. They reorganise time, attention, prestige, and the conditions under which thought is socially rewarded.
That is why the issue may now be a civilisational mutation that still advantages elites materially while simultaneously eating away at the intellectual conditions that once stabilised elite rule. To see why, it helps to return to the distinction between extensional, intensional, and hyperintensional forms of thought.
Extensional thought is satisfied when outputs line up. It asks whether the answer is right, whether the token has been produced, whether the mark has been obtained, whether the credential has been earned. Intensional thought asks a thicker question. It asks how the same thing is being grasped, under what description, by what conceptual route, and with what sensitivity to nearby alternatives. Hyperintensional thought pushes further and asks about grounds, internal structure, and the fine grained differences between claims or arguments that may come out extensionally equivalent. A society with strong intensional and hyperintensional institutions does not simply value correct answers. It values different ways of arriving at them, the reasons that justify them, and the ability to tell when two superficially similar claims are in fact structured differently. That is the world presupposed by much serious law, scholarship, diplomacy, science, and humane education. It is also the world presupposed by the technical apparatus of a Kripke, Stalnaker, Williamson, Fine, and Hawthorne, because all of them in different ways insist that truth is not exhausted by immediate outcome and that the structure of possibility, presentation, and ground matters.
What becomes visible in the present media ecology is that extensionality has escaped the school and become ambient. It is not just that schools measure outputs. The entire social environment increasingly trains people to treat visible performance as reality. Platforms privilege what can be instantly registered, amplified, ranked, reacted to, clipped, and circulated. Ofcom’s latest adult tracker found that fewer UK adults now feel free to be themselves online than a year earlier, and only a minority say they can share opinions more easily online than offline, which suggests that online life is not simply expressive but highly performative and monitored. The same Ofcom research also shows the scale of passive social media use and the growing integration of AI into ordinary media habits, while children’s reports show social media and messaging use reaching younger and younger age groups.
That matters philosophically because intensional and hyperintensional thinking require conditions that platforms systematically weaken. To think intensionally you often need to hold several descriptions of the same object in mind, delay closure, compare perspectives, and test whether equivalence at one level conceals difference at another. To think hyperintensionally you need even more patience, because you must attend to grounds, inferential articulation, hidden assumptions, and the difference between conclusion and route. None of this is impossible online, but it runs against the reward structure of platforms optimised for speed, salience, novelty, and engagement. OECD data now show that more than half of students across OECD countries report being easily distracted, especially those spending more than an hour per weekday on social networks or online leisure browsing, and that attention in class is diverted by peers’ devices on a large scale. UNESCO has similarly stressed that smartphones and excessive or poorly integrated technology use can have negative effects on learning, and the rapid spread of school phone bans worldwide is itself evidence that governments increasingly read the surrounding digital ecology as educationally corrosive.
Now the stronger and more unsettling claim is that this ecology may be undermining even elite reproduction. Historically, elite assurance rested partly on insulation. Wealth could buy quiet, books, tutoring, travel, conversation, cultural confidence, and institutions where ambiguity was still survivable. Those buffers still matter. They remain real advantages. But insulation is weaker when the general prestige economy itself has shifted. If status now flows through visibility, immediate reaction, brand coherence, ideological signalling, and perpetual audience management, then elite actors too are pulled toward extensional life. The child of privilege may still attend the tutorial, but that tutorial now exists inside a broader world that rewards posting over reading, certainty over qualification, and circulation over depth. The elite advantage does not disappear, but it can change in kind. It becomes less an assurance of deep conceptual formation and more an assurance of superior navigation of the same extensional regime, better branding, better coaching, better curation, better reputational management. That is a more brittle ruling culture than the old one.
This is one reason why the rise of figures such as Trump matters as more than a political event. It indicates not simply the capture of institutions by a demagogue, but the arrival of extensional cognition at the summit of rule. I do not mean that Trump lacks all cunning or that his supporters are incapable of nuance. I mean something more specific. The surrounding regime increasingly privileges visible loyalty over grounded judgement, public performance over internal competence, and immediate rhetorical effect over institutional memory. Reuters reported in February and March 2026 that the administration moved to strip job protections from up to 50,000 federal employees through a new civil service category, making it easier to remove career officials deemed to influence policy, and that unions challenged the change as a direct threat to the independence of the civil service.
Reuters had already reported in 2025 on leak investigations and loyalty testing inside the administration. Taken together, these moves are not merely conservative governance. They structurally subordinate institutions that traditionally depended on professional, reason giving, quasi intensional judgement to a logic of visible allegiance and presidential control. That is exactly the kind of shift one would expect if intensional and hyperintensional practices were under threat. A mature civil service, an independent legal bureaucracy, a serious diplomatic corps, a strong university sector, and a genuinely professional teaching class all depend on the idea that not every equivalence at the level of outcome is a real equivalence. They require people who can say, this policy sounds like that one but is grounded differently, this legal claim reaches the same conclusion by an illegitimate route, this apparent loyalty masks institutional damage, this statistic is extensionally impressive but inferentially empty. In other words they require forms of mind trained to detect differences that extensional politics wants to erase. When such institutions are weakened, either through direct political attack or through the broader platform culture that hollows patience and memory, the result is not merely more propaganda. It is a shrinking of the social habitats in which fine grained distinctions remain thinkable.
At this point Williamson, Fine, Hawthorne, and Stalnaker stop looking esoteric. Their technical apparatus begins to look like a defence of endangered capacities. Stalnaker’s work on possible worlds and informational alternatives concerns what it is to situate a claim in a space of possibilities rather than merely assert it. Williamson’s safety condition asks not simply whether a belief is true but whether it could easily have been false, which is another way of insisting that knowledge requires robustness across nearby alternatives. Hawthorne’s mapping of Williamson shows how many careful decisions are needed before one can say that a belief is genuinely safe. Fine’s hyperintensionalism insists that necessity and truth value do not settle all questions, because grounding, essence, and internal articulation matter too. None of this is merely scholastic. It encodes habits of mind that are socially fragile. A culture drifting toward extensionality does not just stop reading technical philosophy (it hardly even started!). It loses the felt need for the distinctions such philosophy articulates.
That is why contemporary schooling matters so much. If schools become increasingly performative, if credentials continue to multiply while tasks are narrowed to what can be audited, if AI lets students and institutions substitute acceptable outputs for actual understanding, then schooling ceases to function as a mass initiation into intensional and hyperintensional life. OECD’s recent work on digital technologies and students’ learning warns that distracting interactive features can cause cognitive overload and that social media integration creates distraction and privacy concerns, while reporting around the OECD’s 2026 digital education outlook has highlighted worries about chatbots creating a false sense of mastery and masking failures to learn. Those warnings fit the deeper philosophical concern almost perfectly. An extensional system cannot easily distinguish between having the answer and owning the route to the answer.
The result may be a strange double movement. On the one hand, elites still retain advantages, because they can pay for human density, tutoring, calm, selective schools, and cultural capital. On the other hand, the ruling strata themselves become shallower in the long run because the whole environment in which elites are formed is increasingly platformed, accelerated, and reputationally volatile. That helps explain why contemporary leadership in several democracies often appears both highly schooled in performance and remarkably thin in judgement. It is not that intelligence has vanished. It is that the ecology rewarding intensional and hyperintensional discipline has weakened, while the ecology rewarding reaction, visibility, and loyalty has strengthened.
The old class compromise, unequal but still partly dependent on a relatively serious public sphere and a relatively serious professional class, may be giving way to something more unstable. If the non elite are mass educated into extensional performativity while elites are no longer fully able to reproduce their own deeper cognitive traditions, then we do not get a stable aristocracy of thought ruling a managed majority. We get a more general degradation in which wealth still buys advantage, but cannot fully buy wisdom, and in which leadership itself becomes increasingly extensional. Trumpism is then not simply the political success of an anti intellectual faction. It is evidence that a media saturated, metrics saturated, loyalty saturated order can elevate precisely those figures most native to extensional life, those who treat visibility as truth, transaction as reason, branding as substance, and immediate effect as sufficient proof of reality.
There is one qualification worth keeping in play, because otherwise the picture becomes too neat. Social media does not only erode knowledge. Under some conditions it can widen access to information and even improve knowledge and belief accuracy. A 2025 Nature Human Behaviour field experiment found that following news on Instagram could increase political knowledge and belief accuracy. That matters because it means the platform environment is not ontologically doomed to superficiality. The issue is not the existence of digital media, but the dominance of engagement driven forms that reward brevity, outrage, perpetual novelty, and behavioural capture. The danger lies in the prevailing platform logic, not in communication technology as such.
So I'm saying that the recent ecosphere does not just threaten one class more than another. It threatens the very social reproduction of intensional and hyperintensional life, though not evenly and not symmetrically. Elites still possess buffers, but those buffers are weaker than before because the prestige, media, and political environment now rewards extensional habits even at the top. Schooling, social media, managerial governance, and authoritarian politics are converging around a common form, one that prizes visible outputs, measurable compliance, and rapid circulation over grounded, counterfactual, structurally articulated understanding. In that setting, the technical work of Williamson, Fine, Hawthorne, and Stalnaker begins to look less like niche analytic refinement and more like a record of capacities that a healthy intellectual culture once needed, and may now be in danger of forgetting how to demand. If that is right, then the defence of serious education is no longer only a defence of fairness or standards. It is a defence of the social conditions under which non extensional thought can still exist.
If we now bring in a more explicit account of ideology, the picture becomes sharper and, in a certain way, more paradoxical. What looks like a drift toward extensional schooling, extensional media, and extensional politics can be redescribed, in the spirit of Brian Leiter and Jamie Edwards style approaches to ideology, as a transformation in how social interests are stabilised and obscured. But the crucial twist is that the very mechanisms that once helped to stabilise elite power, including complex justificatory narratives, professional mediation, and institutionalised forms of reason giving, are now weakening. The result is not simply domination, but a more exposed, and potentially more brittle, form of domination.
Leiter’s broadly Marxian understanding of ideology is helpful here. For Leiter, ideology is not primarily a set of explicit lies. It is a system of beliefs, norms, and justificatory practices that serve the interests of dominant groups while appearing natural, inevitable, or rational to those who inhabit them. Crucially, ideology often operates through what seem like reasonable, even sophisticated, forms of thought. It does not require crude falsehoods. It can operate through legal argument, economic theory, educational doctrine, and moral language. In that sense, ideology has historically depended on intensional and sometimes hyperintensional structures. It has needed the ability to draw distinctions, to reframe issues, to present interests as principles, and to embed power within apparently neutral forms of reasoning. Edwards emphasises how ideology is reproduced through institutions and practices rather than merely through explicit doctrine. It is enacted in how systems are designed, how incentives are structured, how language is used, and how certain forms of reasoning are privileged over others. Ideology, on this view, is not just what people say. It is what they are enabled and disabled from thinking. If we now return to the extensional shift I have been tracing, a striking possibility emerges. The neoliberal and post neoliberal transformation of institutions may still be guided by elite interests, in the sense that the resulting structures continue to advantage those with wealth, capital, and social power. But the mode of ideological operation has changed. Instead of relying heavily on complex justificatory frameworks that require intensional and hyperintensional literacy to sustain, the system increasingly operates through direct performativity, through metrics, outputs, and visible compliance.
In older formations, the legitimacy of elite dominance often depended on elaborate narratives. Legal systems articulated doctrines of fairness, rights, and procedure. Economic systems were justified through theories of efficiency, growth, and welfare. Educational systems were framed in terms of merit, cultivation, and public good. These narratives were not always sincere, but they required a certain level of conceptual engagement. They depended on a population, or at least a significant stratum within it, capable of following, contesting, and internalising complex arguments.
That is where intensional and hyperintensional thinking played a stabilising role. They allowed ideology to operate at a level where interests could be refracted through principles, where contradictions could be managed, where competing claims could be held in tension. The system did not merely impose itself. It justified itself, and in doing so, it relied on forms of thought that could sustain that justification. What happens, then, if those forms of thought are eroded? If schooling becomes increasingly extensional, if media environments reward immediate reaction over sustained reasoning, if institutional life is reorganised around metrics rather than judgement, then the cognitive conditions for that kind of ideology weaken. People become less accustomed to navigating complex justificatory structures. They become more attuned to visible outputs, to who benefits, to who wins and who loses.
This is where the paradox appears. One might think that a more extensional population would be easier to govern, and in some respects that is true. It may be more responsive to simple signals, more compliant with clearly defined expectations, more focused on immediate outcomes. But it may also be less susceptible to certain kinds of ideological masking. If people are no longer trained to follow or respect complex justificatory arguments, then attempts to cloak self interest in such arguments may fail. In that sense, the erosion of intensional and hyperintensional thinking can strip away a layer of ideological cover. What remains is a more direct perception of interests. Policies that favour certain groups may appear less as the outcome of intricate justifiable reasoning and more as straightforward exercises of power. Economic arrangements that concentrate wealth may be seen less as the product of neutral market forces and more as the result of decisions that benefit some at the expense of others. This helps to explain the peculiar character of contemporary political discourse.
Figures such as Donald Trump often do not attempt to provide elaborate justifications for their actions in the traditional sense. Instead, they operate in a register that is highly extensional. Success is measured in deals, wins, ratings, visibility, dominance. Language is used performatively rather than argumentatively. Claims are asserted rather than carefully grounded. From the perspective of an older, more intensional political culture, this appears crude, even shocking. From within an extensional ecology, it can appear direct, authentic, or simply effective. Yet this very directness exposes something. Without the mediating layer of intensional justification, actions can appear as what they are in a more naked form. When a policy benefits a narrow group, there may be less effort to embed it within a complex narrative of general good. When a leader pursues personal or factional advantage, the absence of sustained justificatory discourse can make that pursuit more visible.
The same processes that may have contributed to the erosion of a knowledge rich, intensional middle class, processes that advantage elites by making systems more manageable and more oriented toward measurable outputs, can also undermine the subtle forms of ideology that once stabilised elite rule. If people are less able or less willing to engage with complex reasoning, they may also be less willing to accept it as a cover for power.
But there is another layer to this. Intensional and hyperintensional thinking are not only tools for sustaining ideology. They are also tools for resisting it. They allow individuals to detect hidden assumptions, to distinguish between superficially similar claims, to trace the consequences of policies, and to challenge justificatory narratives. Their erosion therefore weakens critical capacity as well as justificatory capacity. This creates a volatile situation. On the one hand, people may see more directly that certain arrangements are unfair or self serving. On the other hand, they may lack the conceptual resources to articulate coherent alternatives or to sustain complex critiques. The result can be a form of oscillation. Periods of apparent clarity, in which power is seen as nakedly self interested, are followed by periods of confusion, in which simple narratives, conspiracy theories, or populist slogans fill the gap left by more structured forms of reasoning. The decline of intensional and hyperintensional thinking does not produce a stable, clear eyed public. It produces a public that is both more exposed to the realities of power and less equipped to process them in a sustained way.
From the perspective of elites, this is a dangerous equilibrium. It may be easier to manage systems in the short term through extensional mechanisms, metrics, compliance, visibility. But the long term stability of those systems may depend on forms of legitimacy that require more than immediate performance. They require trust, justification, and the sense that decisions are grounded in reasons that can be understood and, at least in principle, contested. If those forms of legitimacy erode, then power must rely more heavily on other mechanisms, on control of information flows, on institutional dominance, on the management of perception. This can reinforce authoritarian tendencies, because without a shared space of reason giving, disagreement becomes harder to manage through argument and more likely to be managed through force or exclusion.
So the deeper conclusion is not that elites have simply engineered a system that benefits them at the expense of others. It is that the transformation of the cognitive ecology of society, the shift toward extensional forms of schooling, media, and politics, may be both a product of and a threat to existing power structures. It advantages elites materially while simultaneously eroding the intellectual and institutional conditions that have historically made their dominance sustainable.
In that sense, Trump et al are not just beneficiaries of this shift. They are also symptoms of it. They operate effectively within an extensional environment, but their mode of operation also reveals the limits of that environment. It shows what happens when the capacity for sustained, grounded, intensional reasoning is no longer central to public life. The exposure of crude self interest is not necessarily a sign of deeper understanding. It may be a sign that the structures that once mediated between interest and justification have weakened, leaving power more visible but also more unstable.
Now I want to segue to Thomas Piketty, because he gives a long historical account of the distribution of capital, the transformation of inequality, and the changing structure of elites that maybe allows the cognitive distinctions I have been developing to be read as an underlying layer that helps explain, or at least illuminate, the phases he identifies.
To begin, it is worth recalling the broad arc in Piketty’s work, especially in Capital in the Twenty-First Century and Capital and Ideology. He argues that the period roughly from the 1930s to the 1970s in many Western countries saw an unprecedented compression of inequality. This was not an organic development of capitalism but the result of shocks, wars, political mobilisation, taxation regimes, and institutional redesign. Wealth concentration fell, top incomes were heavily taxed, public services expanded, and what he calls a “social state” or “egalitarian regime” emerged. This period also saw the expansion of education, the growth of professional classes, and a broader distribution of cultural and institutional participation.
After the late 1970s, this settlement began to reverse. Inequality increased, wealth reconcentrated, taxation became less progressive, and new elite formations emerged, including what Piketty calls the “Brahmin Left” and the “Merchant Right”, a configuration in which highly educated elites and economic elites align in complex ways. Education expanded further in formal terms, but its role in reproducing inequality became more pronounced. Credentials multiplied, but access to the most powerful forms of capital, economic, social, and cultural, remained uneven.
Now, if we place alongside this economic and sociological account the distinctions between extensional, intensional, and hyperintensional forms of thought, a suggestive alignment appears. The period from the 1930s to the 1970s can be read, cautiously, as a period in which the social conditions for intensional and even hyperintensional thinking were relatively widespread, at least compared to what came before and after. This is not to romanticise the period. Access was still stratified, and many groups were excluded. But the institutional structure of society created a broader zone in which people were required, and enabled, to engage in forms of reasoning that went beyond immediate outputs.
Consider what characterised that period. There was significant investment in public education, including universities that were less overtly marketised. There were strong professional classes, teachers, civil servants, engineers, doctors, whose work required judgement, interpretation, and the application of general principles to particular cases. There were relatively stable career trajectories, which allowed for the development of expertise over time. There were public institutions, broadcasting, print media, cultural organisations, that supported extended forms of discourse. There was also, crucially, a political culture in which competing ideologies, socialism, liberalism, conservatism, had to be articulated, defended, and contested in relatively structured ways. All of this can be redescribed in our terms as an ecology that required and sustained intensional thinking. People had to navigate different descriptions of the same phenomena, to understand how policies could be framed in multiple ways, to follow arguments that connected principles to outcomes. In some domains, especially in higher education and professional life, hyperintensional distinctions also mattered. The difference between two legal arguments that reached the same conclusion, the structure of a scientific explanation, the interpretation of a literary text, these were not reducible to simple outputs.
Importantly, this ecology was not confined to a tiny elite. It was supported by the broader social state. Welfare provision, labour protections, and public investment created a degree of security that made it possible for larger numbers of people to participate in such forms of life. The middle class that expanded in this period was not only an income category. It was a carrier of these cognitive practices.
Now consider the period before this, in the late nineteenth and early twentieth centuries, which Piketty describes as an era of high inequality and patrimonial capitalism. Wealth was concentrated, and large segments of the population had limited access to education and institutional participation. In such a context, the social space for intensional and hyperintensional thinking was narrower. It existed, often in highly developed forms, but it was largely confined to elite institutions. For most people, the demands of life did not require, and did not support, engagement with such forms of reasoning. In our terms, the cognitive ecology was more sharply stratified.
The equally interesting question is what happens after the 1970s. Piketty documents the rise of inequality, the transformation of elites, and the expansion of education alongside increasing stratification. On the surface, one might think that the expansion of higher education would lead to a further spread of intensional and hyperintensional thinking. More people attend university, more credentials are obtained, more knowledge is produced. But if we look more closely, the structure of education and work has changed in ways that may push in the opposite direction. Education becomes more credentialised, more tied to labour market outcomes, more subject to measurement and accountability. Universities become more managerial, more oriented toward rankings, outputs, and performance indicators. Schools become more standardised, more focused on test results and measurable attainment. Professional life becomes more audited, more governed by targets and metrics. Media environments become faster, more fragmented, more oriented toward immediate engagement. In terms of our distinctions, this can be seen as a shift toward extensionality. What matters increasingly is the production of outputs that can be measured, compared, and traded. Intensional and hyperintensional aspects of knowledge, how it is structured, how it is grounded, how it behaves under variation, are harder to capture in such systems and may therefore be de-emphasised.
This does not mean that intensional and hyperintensional thinking disappear. They persist, often at high levels, in certain domains. But their social distribution and their institutional support becomes more uneven. They are concentrated in elite institutions, specialised fields, and certain professional contexts, while broader educational systems focus more on extensional performance. Piketty’s analysis of the “Brahmin Left” is particularly interesting in this regard. He describes a class of highly educated individuals who dominate cultural and educational institutions but who are increasingly disconnected from the economic concerns of less educated groups. One way of reading this, through our lens, is that this group retains access to intensional and hyperintensional forms of thought, but within a social structure that is becoming more extensional for others. The result is a divergence not only in income and wealth but in cognitive ecology.
However, as we have already begun to suggest, this divergence may not be stable. The broader shift toward extensional systems, in schooling, media, and governance, can feed back into elite contexts. Universities that are subject to rankings and metrics may find it harder to sustain forms of teaching and research that require time, ambiguity, and deep engagement. Professional fields that are increasingly audited may prioritise measurable outcomes over nuanced judgement. Media environments that reward speed and visibility may shape even elite discourse.
So my mapping onto Piketty is not a simple one to one correspondence. It is not that there was a pure intensional age from the 1930s to the 1970s, followed by a pure extensional age thereafter. Rather, the suggestion is that the period Piketty identifies as relatively egalitarian also coincided with, and helped sustain, a broader social distribution of intensional and hyperintensional practices. As that egalitarian settlement has been dismantled, the institutional and social conditions that supported those practices have also been weakened. What makes this connection interesting is that it allows us to reinterpret some of Piketty’s findings in cognitive terms. The rise of inequality is not only about income and wealth. It also involves a reorganisation of how people are trained to think, what kinds of reasoning are valued, and how knowledge is structured and assessed. The decline of the middle class, in Piketty’s sense, can then be seen as linked to the decline of a social stratum that sustained certain forms of thought.
At the same time, the argument I have been developing adds a further layer of tension. If the current trajectory continues, the shift toward extensional systems may not only disadvantage the non elite. It may also undermine the conditions that allow elites to reproduce their own forms of knowledge. The very processes that increase inequality may also erode the cognitive practices that have historically underpinned stable forms of elite rule. So the alignment with Piketty does not replace his economic analysis. But it provides a way of thinking about how changes in the distribution of capital and power might be connected to changes in the distribution and structure of knowledge, and how the historical period he identifies as exceptional may also have been exceptional in cognitive as well as economic terms.
I now want to do a Daniel Bell and go back and think about the cultral contradictions of capitalism, which he developed out his thinking about Marx way back in the 1960's. As we all recall Karl Marx locates the instability of capitalism in its economic structure, in the tension between capital and labour, in the tendency of the rate of profit to fall, in crises of overproduction, and in the social relations that arise from private ownership of the means of production. On that picture, capitalism undermines itself materially. It generates conditions, inequality, immiseration, concentration, that eventually destabilise the system. What I have been circling, however, is the possibility that there is another layer of contradiction, one that Marx could not fully articulate because it depends on later developments in epistemology, logic, and media ecology.
The contradiction may not lie only, or even primarily, in the economic mechanics of capitalism, but in the cognitive structures that capitalism increasingly requires and simultaneously erodes. To see this, we need to bring together three strands.
First, capitalism, especially in its more advanced forms, depends on highly developed knowledge systems. It requires science, engineering, law, finance, logistics, and increasingly complex forms of coordination. These domains are not extensional in any simple sense. They depend on intensional and hyperintensional thinking. A financial derivative is not just a number. It is a structure of conditional claims. A legal argument is not just an outcome. It is a chain of reasoning grounded in precedent and interpretation. A scientific theory is not just a set of predictions. It is a network of explanations, models, and inferential commitments. In all these cases, what matters is not only that something works here and now, but that it would continue to work under variation, that it is grounded in reasons that can be extended, revised, and defended.
Second, as capitalism evolves, especially under neoliberal and platform conditions, it increasingly reorganises institutions around extensional metrics. Profit, growth, performance indicators, rankings, outputs, these become the dominant currencies. This is not simply an ideological choice. It is a functional response to the need to manage complex systems at scale. But it has consequences. It privileges what can be measured and compared, and it tends to marginalise what cannot.
Third, the media and educational ecologies that accompany this shift, performative schooling and universities, social media, AI assisted production, further reinforce extensional habits of thought. They reward speed, visibility, and output over depth, structure, and grounding. They reduce the space of variation in which safety, in Williamson’s sense, would be tested. They weaken the everyday cultivation of intensional and hyperintensional capacities.
Now, if we put these strands together, a contradiction begins to emerge. Capitalism requires, at its most advanced levels, agents and institutions capable of intensional and hyperintensional reasoning. It needs people who can design systems, interpret complex information, navigate uncertainty, and sustain chains of reasoning across changing conditions. But the broader organisation of society increasingly produces agents who are trained, and incentivised, to operate in extensional modes, to produce outputs, to meet targets, to respond to signals. In other words, the system depends on a form of cognition that it is simultaneously undermining.
This is not entirely new. Marx himself, in a different vocabulary, recognised tensions between the development of productive forces and the social relations of production. But what is new here is the focus on cognition as the site of contradiction. The issue is not only that capital concentrates or that labour is exploited. It is that the forms of thought required to sustain complex systems are not easily reproduced within a culture that privileges extensional performance. We can make this more concrete. Consider finance. Modern financial systems rely on highly sophisticated modelling, risk assessment, and legal structuring. These are paradigmatic cases of hyperintensional reasoning. Small differences in assumptions, in the structure of a contract, in the interpretation of a clause, can have large consequences. Yet the same system also produces incentives for short term performance, quarterly results, visible gains, and the packaging of complex instruments into simplified products.
The global financial crisis can be read, in part, as a failure of safety. Instruments that appeared stable under certain conditions proved fragile under slight variation. The models did not generalise. The beliefs were not safe. Or consider technology. The development of AI systems requires deep theoretical and technical expertise. But their deployment often occurs in environments that prioritise scale, speed, and user engagement. The result can be systems that perform impressively in narrow domains but fail in unexpected ways when conditions change. Again, the issue is not simply economic. It is cognitive. The system rewards outputs that appear correct without ensuring that they are grounded in methods that are robust across variation.
Education, as we have discussed, sits at the centre of this. If schooling becomes increasingly extensional, producing students who can generate acceptable outputs without developing safe beliefs, then the supply of individuals capable of intensional and hyperintensional reasoning may become more restricted. It may be concentrated in certain institutions or networks, but not widely distributed. At this point, the contradiction takes on a social form.
A system that depends on a narrow stratum of highly trained individuals, while producing a broader population oriented toward extensional performance, becomes vulnerable in multiple ways. It may struggle to reproduce the expertise it needs. It may become more dependent on a small elite whose own formation is increasingly fragile. It may face a population that can see the outputs of the system but lacks the means to understand or contest its underlying structure.
This is where the earlier discussion of ideology returns. In a more intensional society, ideology operates through complex justificatory frameworks. People can be persuaded, or at least stabilised, by arguments that connect principles to outcomes. In a more extensional society, those frameworks weaken. People see more directly that certain arrangements benefit certain groups, but lack the conceptual resources to articulate alternatives. The result can be a form of instability that is not easily resolved.
Now, if we bring Marx back into the picture, we can say that the traditional account of capitalism’s contradictions may have misidentified the primary site of instability. Capitalism may not collapse because the rate of profit falls or because production outstrips consumption. It may destabilise because the cognitive ecology it produces cannot sustain the forms of reasoning it requires. It generates agents who are highly effective within extensional frameworks but less capable of navigating the intensional and hyperintensional structures that underpin complex systems.
This is a speculative claim, but it has some resonance with contemporary developments. We see systems that are highly efficient in producing outputs but prone to unexpected failures. We see political environments in which visibility and performance dominate over deliberation and justification. We see educational systems that expand access while narrowing the forms of knowledge that are cultivated. We see elites who retain advantages but are shaped by the same extensional pressures.
I don't see a happy ending. It may lead to a prolonged period of instability, in which systems continue to function but in increasingly brittle ways. Failures will become more frequent, trust will erode, and the gap between visible performance and underlying structure may widen. We'll probably blow ourselves up.
The crisis of contemporary capitalism, and our geopolitical mess is as much a crisis of cognition as of economics. Educational is right at the heart of it.