"Speaking Truth to Power" Is Bad Epistemology

If intellectuals want to hold power accountable, they should focus less on power and more on truth.

One of the most influential ideas among “intellectual elites” in the broadest sense—academics, public intellectuals, writers, pundits, journalists, artists, and so on—is that their job is to “speak truth to power”. Even when they don’t use this specific phrase, it captures how many intellectuals understand their ethos and social responsibility: to unmask and confront entrenched interests, official narratives, and dominant institutions.

Historically, this ethos has been associated with the left, especially in traditions such as critical theory, “counter-cultural” art, and activist scholarship. Today, it increasingly captivates the right as well, where an “anti-elite” politics paints right-wing intellectuals as brave dissidents exposing where real power lies in society: in the “liberal establishment”, the “regime”, “the Cathedral”, or sinister networks of the “deep state”.

It’s easy to understand the ethos’s appeal. The world contains oppression, exploitation, and extreme power inequalities, many of which are unjust and harmful. Societies can’t rely on the powerful to check themselves. They will spread, fund, and amplify self-serving propaganda. So, intellectuals surely have a responsibility to push back against such falsehoods—to expose deception, unmask mystifying ideologies, and reveal what is really going on.

It’s easy to think of cases that fit this model of courageous intellectual activity: Mary Wollstonecraft challenging the subordination of women, Ida B. Wells documenting lynching in the American South, Solzhenitsyn exposing Soviet oppression, the reporters who broke Watergate or decades of Catholic Church abuse, the scientists who exposed tobacco industry propaganda, and so on.

Nevertheless, I will argue that in practice, the widespread embrace of this ethos among large segments of the Western intelligentsia is often harmful and counterproductive. It encourages intellectual laziness and self-deception in ways that undermine its stated aim.

To hold power accountable, societies need access to trusted truths about what is happening. At least in modern liberal-democratic societies, these truths are often highly complex and counterintuitive. Determining what they are is challenging. The ethos of speaking truth to power encourages intellectuals to think that such truths have been established before inquiry has even begun. It replaces a difficult epistemic task—finding out what is true, where power lies, and how truth and power interact in specific cases—with the simpler, more self-flattering goal of summoning intellectual courage.

Of course, intellectual courage is extremely important. But without first carefully establishing what’s true, it’s either pointless or harmful. So intellectuals’ primary responsibility is simply to seek and speak the truth, full stop. This job description is less heroic, but it’s more intellectually and socially useful, and it’s less vulnerable to self-deception precisely because it’s less heroic.

At least, that’s what I will argue here. Specifically, I will argue that an ethos of speaking truth to power has three problems: it prejudges what inquiry is supposed to discover; it licenses motivated reasoning under a heroic self-image; and it exempts intellectual elites from the suspicion they direct at others. For these reasons, it also threatens public trust in the institutions that democratic societies depend on to make sense of society and hold power accountable.

I will illustrate these arguments with several examples: the strange dogmatism of “critical” theory, the left’s failure to grapple seriously with AI, and the modern right’s collapse into conspiracism.

First, though, it will be helpful to start with a more general framework for understanding what power even is, and how it relates to the domain of ideas.

The Causes and Complexities of Social Power

Historically, many intellectuals assumed that different kinds of power could be reduced to one fundamental kind. The canonical example is Marx’s view that economic power relations are upstream of all others.

As the sociologist Michael Mann argues in his series The Sources of Social Power, this reductionist perspective seriously misrepresents human societies. Power, Mann argues, has at least four distinct sources: political, economic, military, and ideological. Although they often reinforce one another, the connections are complex and “promiscuous”, and no power source can be reduced to the others.

Nevertheless, if we focus on the distribution and character of power in recorded human history prior to the emergence of modern liberal democracies, two broad patterns emerge.

First, most power relations were illegitimate and extractive. Although power took different forms and shifted among different elites, the elites and institutions they fought over were highly exploitative. Second, ideological power was typically harnessed to support and legitimise such extractive social orders. Hence Marx and Engels’ famous observation that “the ideas of the ruling class are in every epoch the ruling ideas.”

This “dominant ideology thesis” is an exaggeration. Even in the most extractive regimes, ideological power is typically at least partly independent of other sources of power and shows a surprising tendency to be unleashed and to disrupt existing social orders (Christianity, Islam, Protestantism, liberalism, fascism, communism, and so on). Moreover, the masses often see through elite ideologies, even if they pay lip service to them to avoid trouble.

However, as with most of Marx’s ideas, the thesis contains a grain of truth. Throughout most of recorded history, extractive elites conscripted and coerced ideological elites (priests, theologians, intellectuals, artists, and so on) to produce and distribute legitimising myths. These myths then dominated society’s formal communication channels.

In that kind of world, an ethos of speaking truth to power can serve as an understandable guide for brave, public-minded intellectuals. Because illegitimate power structures are relatively easy to identify and ruling ideologies are systematically false and self-serving, intellectual courage and accuracy often overlap. When a regime depends on enforced lies, refusing to go along with those lies is a contribution to the truth. When exploitation is naked, the main challenges public-minded intellectuals confront are practical and moral—being killed, imprisoned, or silenced—not epistemic.

Both today and throughout history, the courageous writers, reporters, intellectuals, and artists who have spoken truth to power in such regimes often deserve deep admiration and respect.

The problem is that the Western intelligentsia no longer lives in that world.

The Sources of Social Power in Open Societies

Today, Western intellectual elites live in liberal democracies characterised by universal rights, formal equality, the rule of law, pluralism, constitutional limits on state power, and significant political and economic freedoms. We also live in an era of unprecedented material prosperity, in large part a consequence of how economic freedom and competition are channelled into productivity and innovation through free markets, and political freedom and competition are channelled into social welfare, insurance, and public goods through democracy.

In these societies, social power takes on a very different form.

First, although nobody should think that power is always distributed in fair or functional ways in liberal societies, it’s even more misguided to treat all power asymmetries as illegitimate or harmful. Consumer choice and voting mean that political and economic power is more evenly distributed, and disparities in wealth and political influence often reflect not extraction but fair, socially beneficial processes. For example, wealth is often downstream not of theft or rent-seeking but of free markets that reward innovation and efficiency, benefiting everyone. Similarly, political power is typically allocated not by brute force but through complex processes of democratic participation or the broadly meritocratic selection of civil servants.

Again, the point is not that there is no corruption, exploitation, rent-seeking, or domination in modern liberal societies. There is a lot. The point is simply that allegations of such things must be carefully adjudicated on a case-by-case basis. They cannot simply be assumed in advance of inquiry.

Second, ideological power operates very differently in liberal-democratic societies. As figures from Karl Popper to Jonathan Rauch have argued, open societies are distinguished as much by their free, competitive, and pluralistic epistemic characteristics as by their political or economic systems.

One aspect of this concerns the emergence of imperfect but real epistemic freedom for the first time in human history, including freedom of belief, free speech, a free press, and academic freedom. Open societies lack extensive top-down regulation of the information environment, enabling a vibrant, pluralistic public sphere and marketplace of ideas. Another, related dimension concerns the deliberate creation and funding of epistemic institutions, including news organisations, universities, fact-finding agencies, and more. Whether these institutions receive state or private funding, they are typically defined by their independence from other power structures in society and derive public legitimacy and support precisely from such neutrality.

Again, the point is not that in liberal-democratic societies ideological power floats completely free of other power sources. As I have argued elsewhere, the liberal “marketplace of ideas” often functions as a de facto marketplace of rationalisations, and political and economic elites often have greater “purchasing power” in such markets than ordinary people.

Nevertheless, a simple picture in which ideological elites are mere hand-puppets of other elites is a gross distortion. In liberal societies, ideological elites wield substantial independent power and are often far more influenced by their own norms, fashions, and status games than by top-down political authority or economic influences.

In fact, in a clear trend that has accelerated since at least the 1960s and the birth of a powerful “counter-culture” prestige economy, much of the intelligentsia and art world defines itself in explicit opposition to established political and economic power centres. Hence the widespread appeal and embrace of an ethos of speaking truth to power. And today, of course, the media environment in liberal societies is characterised by strong demand on both the left and right for populist denunciations and condemnations of establishment institutions. Critiquing the powerful can function as a lucrative source of cultural esteem and financial rewards.

Three Problems with “Speaking Truth to Power”

Under these conditions, a governing ethos of speaking truth to power no longer makes practical sense for intellectuals.

1. The Replacement Problem

First, if the ethos is to function as a guide to action, the intellectual must already know three things: what the truth is, where power lies, and how truth and power conflict in specific cases. These are precisely the things that intellectual work is supposed to establish, however. So, in practice, the ethos typically puts the cart before the horse, replacing challenging epistemic labour—figuring out what is actually happening in societies characterised by profound ambiguity, uncertainty, and complexity—with a much simpler moral posture.

Consider the following questions. Are experts defending genuine knowledge or protecting their institutional status? Is growing populist distrust of elites a rational response to establishment failure or a symptom of pervasive misinformation? Does social media democratise public debate, degrade it, or merely reveal pre-existing conflicts? What is the empirical track record of “neoliberal” policies? Did globalisation fuel anti-immigrant sentiment across Western countries? Is AI just another “normal” technology or is it on track to trigger an “intelligence explosion” that will upend our economy, politics, and culture?

Answering these and countless other questions requires patient, careful inquiry and data collection, rigorous social-scientific methods, intellectual humility, and deliberate attempts to seek out disconfirming evidence and dissenting viewpoints. A simple heuristic that power and truth systematically collide doesn’t get you anywhere. Powerful people often speak the truth; the powerless often speak nonsense; and very often nobody—neither the powerful nor the powerless—has a clue what is going on, which is why we need rigorous, trustworthy, truth-seeking epistemic institutions in the first place.

The Left’s Embrace of Sophisticated Conspiracy Theorising

In many ways, the clearest illustration of the replacement problem lies in “critical theory” and much of academic “critical studies” in modern universities. Although this work comes in many different forms, its unifying commitment is to some version of Marx’s dominant ideology thesis: that the responsibility of intellectuals is to unmask and oppose ruling (“official”, “hegemonic”) ideologies (“discourses”, “regimes of truth”, etc.), thereby emancipating the downtrodden and oppressed from the illusions that sustain their subordination.

In other words, it is to speak truth to power.

Although there is lots of interesting and insightful work in the broad tradition of critical theory, Joseph Heath is right when he observes that the most striking thing about critical studies today is how thoroughly uncritical—how dogmatic—it tends to be. Much of its intellectual energy and activity proceeds on the basis that the fundamental truths about society have already been established. It is simply taken for granted that debunking, suspicion, and unmasking of designated villains (capitalism, racial capitalism, neoliberalism, neocolonialism, and so on) is valuable. “Research” proceeds primarily by applying and extending this worldview, typically through the hermeneutic parsing of sacred texts (Marx, Adorno, Gramsci, Foucault, Fanon, Butler, and so on), rather than through rigorous social-scientific analysis that leaves open the question of whether the worldview is actually accurate or applicable in specific cases.

In fairness, the intellectual problems with this approach have gradually dawned on some within this tradition, especially once it became clear that the political right could simply appropriate the same toolkit to demonise the left intelligentsia and discredit the causes it champions. (More on this below.) Unsurprisingly, for example, the fashionable idea that all knowledge is socially constructed and entangled with power lost some of its appeal once climate activism became a defining left-wing cause. The analysis is not exactly helpful for encouraging people to “Trust the Science”.

As Bruno Latour belatedly acknowledged, a totalising hermeneutics of suspicion directed towards “power” and “dominant institutions” is no better than high-IQ conspiracy theorising: it simplifies complex realities, flattens important distinctions, and protects itself against refutation by dismissing all criticisms as predictable responses from those in power.

Similar lessons generalise to activist scholarship more broadly. When scholarship is guided by activism—in universities today, that mostly means “social justice” activism, but the point generalises—scholars assume that they have a clear and accurate sense of what justice requires. Their job is merely to “advocate” for justice through their research. Although this work can be valuable, it too often encourages a kind of cosplay of intellectual activity, in which the goal is not to figure out what’s true but to recruit any intellectual ammunition (evidence, facts, arguments, testimony, interpretations) available to support and extend a predetermined political vision. The question of whether the vision is actually correct or applicable in a given case is settled before inquiry has even begun.

Of course, these problems might not matter if they were confined to esoteric corners of humanities departments. But the core ethos driving these problems now shapes large bodies of work produced among the left-wing intelligentsia and pundit class.

The “AI Con” Con

Consider artificial intelligence. In the space of a few years, AI systems have gone from barely producing coherent text to achieving expert-level performance across a vast range of benchmarks, attaining superhuman coding abilities in many domains, and increasingly behaving like general-purpose, tool-using, agentic assistants. Across countless distinct evaluations and metrics, their capabilities have improved at a staggering exponential rate. For these reasons and more, AI progress is now backed by one of the largest capital expenditures in human history, and Anthropic, a frontier AI company, may be the fastest-growing company in the history of capitalism.

These developments throw up many difficult questions. What should we make of modern AI’s capabilities? As these systems improve across many aspects of AI research and development, are we on the cusp of a process of “recursive self-improvement” in which AI systems automate the task of producing new AI systems, fuelling explosive, rapid technological growth? How will we control this technology as it becomes more powerful? How will it impact and disrupt the labour market, our political institutions, culture, and humanity’s sense of meaning and purpose?

These are challenging questions. There are profound expert disagreements and ambiguous, contested evidence across many fields. And yet, in the face of these challenges, large segments of the left-wing intelligentsia and commentariat have coalesced around a simple viewpoint: that AI is basically a nothingburger, a “con”; that state-of-the-art large language models are little better than “fancy autocomplete” or “stochastic parrots”; and that anyone who claims otherwise, including experts who have long worried that AI is extremely dangerous and will pose existential risks, is simply part of an elaborate propaganda machine designed to over-hype a product being sold by deceptive AI companies.

This attitude is certainly not universal among the left, especially as its collision with reality is becoming increasingly clear. There are thoughtful left-wing analyses and growing attention to the topic among influential leftist politicians, including Bernie Sanders. Moreover, there is reasonable disagreement about how powerful AI systems actually are and are likely to become, how informative benchmark-based exponential trends are, and whether AI should be thought of as a sui generis, abnormal technology.

Still, it’s difficult to overstate the scale of the intellectual failure that has occurred among many left-wing intellectuals here—and, by extension, the many departments in universities where they dominate—which has gone far beyond responsible scepticism into the domain of outright denial. This failure is especially galling because a thoughtful, reality-oriented, left-wing intellectual culture would have much to contribute to many of the issues raised by highly capable and rapidly advancing AI, including oligarchy, concentration of power, economic inequality, and labour automation.

As Will MacAskill observes, it is striking that in the midst of a sustained attempt by some of the most powerful capitalists in human history to build a technology openly designed to replace human workers—something that Marx himself literally predicted would happen—the Bluesky intelligentsia has pre-decided it is all hype and propaganda.

What happened? Obviously there is no single answer to this question, but one part of the answer is how much of the modern left’s self-identity of “speaking truth to power” has replaced intellectual inquiry with the application of a predetermined template. In this case, the application of the template runs: Powerful “tech bros” say AI is powerful; The claims of capitalists can be dismissed as self-serving propaganda; So, AI is not really powerful.

There’s no need to check whether this template is actually applicable in this specific case, or to every detail of the case. The core issues have been decided in advance. Indeed, many on the left take pride in refusing to even use AI systems like ChatGPT, even when their literal job is to write about or research them, and even when they make confident claims about what these systems cannot do. The role of intellectual inquiry is to take the predetermined template and find ways of formulating, extending, and rationalising it.

The problem here is not just the bad epistemics. The deeper problem is that the bad epistemics are self-defeating. If your goal is to speak truth to power, it’s important to first carefully determine what is actually true. If you don’t do that, you undermine your goal by misrepresenting the challenges in ways that discredit critics among anyone with actual expertise on the topic. You make power less accountable, not more so.

2. The Motivated Reasoning Problem

There is a second, related problem that arises when intellectual elites embrace an ethos of speaking truth to power: it fuels corrosive forms of motivated reasoning. It doesn’t just let a pre-existing worldview shape inquiry; it tends to systematically distort that inquiry, giving favoured conclusions a much easier pass than unfavoured ones.

Of course, motivated reasoning is a universal tendency. But whereas people should generally be embarrassed by it, an identity of speaking truth to power tends to reframe it as a virtue.

First, it casts intellectuals as brave truth tellers, opposing illegitimate power on behalf of the powerless. Once one identifies with this self-image, isolated demands for rigour no longer seem like a failure of rigour. They look like bravery. Courageous, truth-seeking intellectuals doubt the powerful; only cowards and apologists for power doubt the powerless. Second, the ethos creates a status game where prestige flows to those who unmask and oppose power, and anyone who complicates or challenges those critiques is viewed with suspicion or outright contempt.

In intellectual environments that operate according to this ethos, the result is an informational ecosystem in which claims that demonise villainous elites and institutions are credulously accepted and amplified, whereas any claims that cast them in a positive light are scrutinised, ignored, or dismissed. Once again, this might not be such a problem in societies in which all power is nakedly illegitimate and extractive. But in contexts involving much greater complexity and opacity, it is intellectually poisonous.

This dynamic is on steroids in the AI debate. In Karen Hao’s highly influential, award-winning book Empire of AI, for example, possibilities that many of the world’s leading AI experts take extremely seriously—that we might soon build general-purpose, autonomous, super-intelligent AI—are simply dismissed as “a fantastical, all-purpose excuse for OpenAI to continue pushing for ever more wealth and power.” At the same time, she reports wildly inaccurate claims about AI water use. (Hao has since corrected these mistakes, which were identified by Andy Masley.)

This is a general pattern in much left-wing commentary on this topic. Claims that might suggest frontier AI systems are actually highly capable, useful to many people and businesses, or carry societal benefits (e.g., in science, healthcare, and education) are dismissed as industry hype, even when they come from independent experts. Even claims that such systems have extremely dangerous capabilities are dismissed as further hype because they imply the systems in question are genuinely capable in the first place.

This extreme scepticism and cynicism then suspiciously vanish when confronted with any claims that cast modern AI in a negative light, often no matter how misinformed or dubious they are. (See, for example, prominent but highly misleading claims about AI and the environment, AI and racial discrimination, AI and misinformation, or, the strange, continued insistence that AI systems can’t do things that they can demonstrably do.)

The motivated reasoning problem is that these highly asymmetric standards aren’t recognised as an intellectual failure. They are experienced as heroic activism.

The Right’s Embrace of Unsophisticated Critical Theorising

These problems are not unique to the left. In the last couple of decades, large parts of the right have adopted the identity of anti-establishment politics, embracing much of the aesthetics, rhetoric, and cognitive style of left-wing critical theory and counter-culture. Once progressives came to dominate establishment institutions, the right became the anti-institutional faction that views itself as speaking truth to power. As Joe Rogan once put it, “The rebels are Republicans now. You want to be punk rock? You want to like buck the system? You’re conservative now.” (In support of this assessment, The Sex Pistols’ John Lydon is now a MAGA Trump supporter.)

This anti-establishment, counter-cultural energy has fuelled the demand for a new right-wing intelligentsia—I use the term loosely—capable of articulating and rationalising the core worldview. So, gone are the days of serious intellectuals like Friedrich Hayek, Milton Friedman, Thomas Sowell, or Roger Scruton. For much of the modern right, the intellectual energy now resides with writers, pundits, and social media posters such as Curtis Yarvin, Christopher Rufo, Peter Thiel, Tucker Carlson, and Elon Musk.

Despite superficial differences in how these figures view the world, the core worldview is similar: that real power in the modern world lies not with ordinary political, economic, or military elites but with the coercive ideological power exercised by established institutions such as legacy media organisations, universities, government agencies, and so on. This is “the Cathedral”, the “regime”, the “deep state”, etc., which enforces a stifling and destructive progressive orthodoxy.

There is, of course, some irony here. When the ideal of a counterculture and critical theory began to gain popularity in Western universities in the 1960s, the “system”—the enemy—was capitalism, capitalist elites, and social conservatives. Now, pretty much the same worldview (and sometimes even the same intellectual gurus) has been systematically inverted by a powerful new political movement for whom—as J.D. Vance has put it in a title of a 2021 speech—“the universities are the enemy”.

Just as with the traditional left’s critique of capitalism, there are some important grains of truth in the anti-establishment, “dissident” right’s critique today. It’s simply true that progressive groupthink is too common in many establishment institutions, and that the cultural values and priorities of the professional-managerial class that staffs these institutions play a powerful role in modern Western societies. In Mann’s vocabulary, cultural progressives wield extraordinary ideological power in the modern West, and this power is worthy of scrutiny and critique.

Nevertheless, the right intelligentsia’s new self-image as speaking truth to power exhibits even worse intellectual pathologies than those that have long characterised left-wing critical theory. In place of careful, evidence-based investigations into universities or the administrative state, these institutions are reduced to simplistic, stick-figure caricatures aimed at homogenising and demonising them in extreme ways. Any evidence that supports such caricatures, no matter how cherry-picked or misinformed, is amplified, and any evidence that complicates or challenges them is either dismissed or ignored.

Leftist Indoctrination Camps

Consider universities, for example. Universities in the US and other Western countries are highly complex institutions. They often house thousands of highly trained researchers working on many distinct scientific and intellectual projects and oversee a diverse curriculum across subjects such as quantum mechanics, macroeconomics, cancer research, and personality psychology. Unsurprisingly, given the sheer amount of intellectual ingenuity and research there, they have also been a major force for scientific and technological progress in the modern world.

As Steven Pinker observes of Harvard, although there are undeniable issues with progressive groupthink and free speech in the university, most faculty don’t identify as “very liberal” or on the “radical left”, and there is considerable intellectual diversity and disagreement in terms of research, the personal views of researchers, and what gets taught. Only a small minority of courses with relatively small enrolments are explicitly “woke”, whereas the most popular undergraduate courses tend to be in fields like economics and computer science.

In other words, despite the institution’s obvious imperfections, it holds immense value, and it is a highly intricate system that defies simplistic generalisations. And yet, if you read the “dissident” right-wing intelligentsia, facile slogan-based demonisation is pretty much all you get, with status flowing to those who denounce Harvard and other institutions in the most hyperbolic ways possible—as a “national disgrace”, a “woke” or “Maoist indoctrination camp,” and so on—all dressed up as “based” denunciations among those brave enough to be “red-pilled”.

“Gay Race Communism”

Consider Curtis Yarvin, for example, a tech entrepreneur-turned-intellectual who coined the terms “red pill” and “the Cathedral” as they are used today, and whom J.D. Vance has cited as an intellectual influence.

In recent reflections on why he supports the Trump administration and its assault on many establishment institutions, Yarvin declares that the governing “regime” has “invented a pandemic and killed 20 million people, for absolutely no sane reason at all”, and “has spent the last century managing public opinion with every carrot and stick it can find, up to and including asking professors to compose their own inventive and detailed loyalty oaths to gay race communism.”

As with the rest of Yarvin’s intellectual outputs, it’s challenging to overstate how low-quality these arguments are. Not only is there no monolithic governing “regime”, but the existing evidence for the lab leak hypothesis is highly mixed; almost all the actual evidence we have comes from the very scientists, intelligence agencies (the “deep state”), and legacy media reporters treated as part of this “regime”; it was Donald Trump’s first administration that ended an Obama-era pause on funding high-risk gain-of-function research in 2017; whereas establishment scientists and academics wrote critiques of such research, MAGA supporters and “dissident” right intellectuals didn’t; it was scientists who invented vaccines that saved millions of lives, along with countless other medical and technological advances over the past century; and the framing of DEI statements (in my view, a genuinely objectionable practice) as “loyalty oaths to gay race communism” is a preposterous exaggeration.

In a functional intellectual culture, this sort of analysis would be a source of extreme embarrassment and shame. But for Yarvin and other pundits and writers operating within the “contrarian” right-wing status economy, they are experienced as bravery—as part of a red-pilled escape from being a “libtard and a coward”.

This is what happens when the task of intellectual inquiry is replaced by a governing ethos of speaking truth to power. It allows charlatans and know-nothings to reframe intellectual malpractice as a virtue.

3. The Self-Exemption Problem

The example of the modern “dissident” right also highlights a final, obvious problem with this ethos: many of the movement’s central figures are millionaires or billionaires who wield immense power and considerable influence over the governing administration of the world’s most powerful country.

On the one hand, this illustrates why this intellectual movement’s dysfunctional epistemics are so costly. For example, the modern right’s lazy, conspiracist worldview has had countless terrible real-world consequences, not least in the millions of lives projected by some experts to be lost or put at risk by Elon Musk’s inept DOGE cuts on USAID.

However, it also illustrates a deeper flaw with how the ethos of speaking truth to power operates in practice. By its very nature, this self-image exempts intellectual elites from the suspicion they direct at others. It implicitly treats the intellectual as either powerless or a defender of the powerless.

Although the absurdity of this is especially salient in the case of the modern right-wing intelligentsia and pundit class, it also applies to large swathes of the left-wing Western intelligentsia, whose disproportionate influence in universities, legacy media outlets, and the art world grants them considerable ideological power.

As Mann’s analysis of social power reveals, such ideological power—the power to shape ideas, cultural fashions, and norms—is real. This is why modern culture wars—conflicts over norms, ideas, and symbols—are so heated, and why those who already wield considerable economic power (e.g., Musk, Peter Thiel, Marc Andreessen, and so on) are so obsessed with converting it into ideological power, often through incessant social media posting. If ideological power were merely derivative of economic power, these efforts wouldn’t be necessary.

The existence of ideological power raises awkward questions for those who wield it under the banner of speaking truth to power. If the ethos’s animating insight is that the powerful often embrace self-serving, self-aggrandising narratives, shouldn’t we also turn such suspicion towards intellectual elites themselves? Mightn’t the ethos itself function as a kind of legitimising myth, a way of dressing up activity often rooted in grubby motives—self-aggrandisement, status competition, demonising rivals, and so on—in suspiciously noble clothing?

Once again, this worry is not pressing for true dissidents fighting clear cases of oppression and exploitation. Those who risk life, limb, or reputation to speak truth to power in such cases do so from a position of subordination and personal risk. This is what makes the behaviour so admirable.

For much of the Western intelligentsia today, however, this is simply not the situation. For Ivy League professors, New York Times journalists, the podcast class, or Hollywood actors, “speaking truth to power” is often met not with crushing opposition from a brutal regime but with applause, approval, and accolades.

If we should be highly suspicious of power and its self-serving propaganda, why do intellectuals expend so little effort turning this suspicion inwards on ourselves? (There are some excellent exceptions here.) If power corrupts cognition—a genuine insight—then so, presumably, does ideological power, along with the status games through which it is allocated in universities, journalism, media, and the arts.

Of course, one might respond that this kind of view is too cynical, and that the mere fact that intellectual elites might have impure motives doesn’t show that their claims are wrong. And I agree. We shouldn’t pre-judge that the powerful’s claims are self-serving propaganda before inquiry has even begun—before we’ve undertaken the hard work of figuring out what is actually true.

In other words, an ethos of “speaking truth to power” is bad epistemology.