The Hypocrisy of Anti-Fact
Structural De-Noising and Empirical Falsification of Left-Wing Discourse Architecture
How subjective narratives disguise themselves as objectivity, and what happens
when the disguise is stripped away
ABSTRACT
This paper applies signal de-noising to left-wing discourse texts. After stripping away rhetoric, emotional evocation, and moral packaging, it extracts the underlying logical skeleton and identifies four structural features: packaging subjectivity as objectivity, logical inconsistency, selective use of evidence, and deploying an assertive tone to seal insufficiently supported conclusions. Building on this foundation, the paper uses Kurt Gray’s Outraged (2024) as an empirical sample and verifies each of the four features through detailed textual analysis. It then reveals the complete power chain running from epistemological operations through moral operations to political operations. Finally, from the standpoint of thermodynamic entropy, it argues that the ultimate adversary of anti-fact narrative is not the interests of other groups — it is physical entropy itself.
Keywords: Subjective-objective language ratio · Discourse architecture · Political spectrum · Cultural hegemony · Anti-entropy · Democratic institutions
I. Statement of the Problem
1.1 The Core Question
Does the ratio of subjective to objective expression in human language systematically reflect the speaker’s personality traits, political orientation, professional background, and cognitive honesty?
This question arose from reflections on visual observation training. Amy Herman, in Visual Intelligence, trained FBI agents, physicians, and lawyers to describe the same painting and found that different individuals’ verbal descriptions of the same visual scene varied enormously. Before the same painting, Person A says “The image contains a woman in red clothing, facing left, against a gray wall” — a high-objectivity statement. Person B says “This painting feels oppressive; the woman looks lonely; the entire atmosphere is heavy” — a high-subjectivity statement. Person C says “A seemingly thirty-year-old lonely woman stands before a gray wall; the red clothing and gray background create an unsettling contrast” — a mixed type.
The difference in observation is not merely a difference in “what was seen” but in “how what was seen is expressed in language.” In this process of expression, the ratio of subjective to objective language is a hidden variable that has received almost no systematic study. Under the dual influence of professional training and the political spectrum, this variable forms a decodable fingerprint of individual differences.
1.2 Professional Training’s Restructuring of Subjectivity
Professional training does not simply make people more objective or more subjective; it creates an entirely new set of perceptual priorities — determining what is worth noticing, what is worth naming, and what is worth judging. At the same traffic accident scene, a forensic pathologist says “A approximately 4 cm arcuate laceration is visible at the right temporal region of the victim, with irregular wound margins” — almost entirely objective language, but “objectivity” itself has been reorganized by the conceptual framework of forensic medicine. A psychotherapist might say “The girl crouching in the corner among the bystanders caught my attention; her expression doesn’t look like fear — it’s more like a dissociative state” — superficially a subjective judgment, but behind it is a perceptual filter formed by clinical training. An insurance adjuster says “The contact point between the two vehicles is at Vehicle A’s right front fender, with a deformation area of approximately 0.3 square meters” — highly objective, but extremely selective.
This means that in any passage of text a person writes, the ratio and type of subjective-to-objective language is actually the superposition of three layers of information: the base layer of personality traits, the middle layer of the cognitive framework shaped by professional training, and the surface layer of the influence of the immediate context. The separation and identification of these three layers has seen no systematic work in existing research.
1.3 The Deep Restructuring of Language by the Political Spectrum
The same piece of news — “A city demolished a 50-year-old residential building; a commercial complex will be built on the site” — a left-leaning person tends to express it as “Yet another community carrying the memories of generations has been crushed by capital,” with extremely high subjective language density, and the direction of subjectivity pointing downward — seeing the weak, the victims, the dispossessed. A right-leaning person tends to express it as “The old, dangerous building was finally torn down; the new project will create two thousand jobs,” superficially more objective with numbers and facts, but the selection of which facts to present is itself a form of deep subjectivity — the direction of subjectivity pointing forward — seeing efficiency, development, order.
These two types of people looking at the same reality differ not in “one being objective and the other subjective,” but in the axis of their subjectivity being completely different. The information cocoon effect of the internet era continuously amplifies this difference: algorithmic recommendation causes each person to be exposed to an ever-increasing proportion of content consistent with their own subjective direction. A person whose language ratio was originally 70% subjective / 30% objective may, after ten years of information cocoon feeding, shift to 90% / 10%.
1.4 Existing Research Base
Pennebaker & King (1999), using LIWC lexical categories, found that highly agreeable individuals use more positive emotion words and highly neurotic individuals use more negative emotion words and first-person pronouns, establishing language style as a stable indicator of individual differences. Newman, Pennebaker, Berry & Richards (2003) further found that liars and truth-tellers differ systematically across multiple linguistic dimensions — liars use fewer first-person self-references and fewer exclusion words, with LIWC achieving approximately 67% deception detection accuracy. This work directly linked subjective-objective language ratio to honesty.
Pennebaker in his 2011 book The Secret Life of Pronouns analyzed stream-of-consciousness writing from approximately 8,000 college students and found that “formal” writers (less use of “I,” more use of articles, nouns, and numbers) tended toward dishonesty, status-focus, and power orientation; “analytical” writers (more use of causal and negation words) earned better grades and were more open to new experiences. This is a landmark result directly linking the “objectivity/formality” of language to honesty, personality, and social status.
Schwartz et al. (2013) analyzed millions of posts from 75,000 Facebook users and found significant variation patterns in language use related to personality, gender, and age. Park, Schwartz et al. (2014) built Big Five personality prediction models from the written language of 66,732 Facebook users, achieving prediction accuracy from r ≈ 0.31 for agreeableness to r ≈ 0.41 for openness.
Sterling, Jost & Bonneau (2020) comprehensively analyzed social media users’ language habits and found that liberals use more language conveying goodwill, while conservatives use more language related to threat, power, tradition, certainty, security, anger, and negative emotions. Schoonvelde et al. (2019) analyzed 381,609 political speeches and found that liberal politicians use more complex language (more exclusion words, conjunctions, multi-perspective expressions), while conservative politicians use more concise language — a difference rooted in personality differences in cognitive, emotional, and motivational functioning.
Semin & Fiedler’s (1988) Linguistic Category Model (LCM) distinguished four levels from concrete to abstract in interpersonal description: descriptive action verbs (most concrete/objective) → interpretive action verbs → state verbs → adjectives (most abstract/subjective). Research found that the level of abstraction people choose when describing behavior systematically reflects their motivations, biases, and group identification. This effectively quantifies the “subjective-objective” language ratio.
1.5 The Research Gap
The above studies are scattered across three independent traditions: linguistic psychology, political psychology, and computational linguistics. No existing study uses “the subjective-objective ratio in language” as a unified core indicator to systematically cover personality, political orientation, profession, honesty, and social ability as multi-dimensional individual differences. More critically, existing research either performs “language → personality prediction” (Pennebaker, Schwartz, et al.) or “observation training” (Herman), but almost no one has systematically studied “the subjective-objective language ratio in descriptions of the same objective stimulus” as an independent individual-difference indicator. Most lacking of all is a skeletal analysis of specific political spectrum discourse structures after de-noising — this is precisely the entry point of this paper.
II. Biological Foundations: Brain Structure Differences and Cognitive Modes
2.1 Anatomical Evidence
Kanai, Feilden, Firth & Rees (2011, published in Current Biology) used structural MRI scanning in a large sample of young adults and found that liberalism correlated with increased gray matter volume in the anterior cingulate cortex (ACC), while conservatism correlated with increased gray matter volume in the right amygdala. The functional differences between these two regions are of critical significance: the ACC is responsible for conflict monitoring, error detection, and processing contradictory information — greater gray matter volume means liberals are better at handling conflicting information. The right amygdala is responsible for fear, threat processing, and emotional response — greater gray matter volume means conservatives are more sensitive in processing negative, aversive emotions.
2.2 Functional Validation
Schreiber et al.’s (2013) “Red Brain vs. Blue Brain” study further validated the functional implications of these structural differences: although Democrats and Republicans showed no differences in risk-taking behavior, their brain functional activity was markedly different — Democrats exhibited significantly higher activity in the left insula (associated with interoception and empathy), while Republicans exhibited significantly higher activity in the right amygdala. Yang et al. (2020, Korean team) found that conservatives are more sensitive under threat and anxiety conditions, but exhibit stronger psychological resilience and self-regulation — conservatives showed stronger functional connectivity between the orbitofrontal cortex and precuneus, and between the insula and frontal pole, correlating with higher psychological resilience and self-regulation capacity.
2.3 Large-Sample Replication
Petropoulos Petalas, Schumacher & Scholte (2024, published in iScience) completed the largest pre-registered replication study to date, with a sample of 928 participants. They successfully replicated the small positive correlation between amygdala volume and conservatism, but failed to find consistent evidence for a relationship between ideology and ACC volume. Moreover, the anatomical differences in the amygdala and ACC vary depending on whether the dimension of conservatism is economic or social — indicating that the relationship between political ideology and brain structure is more nuanced and multidimensional than a simple “left-right binary.”
Clinical neurological evidence provides additional corroboration. Mendez (2017, published in the Journal of Neuropsychiatry) found in a review that there exists a right-hemisphere “conservatism complex” involving brain structures sensitive to negativity bias, threat, disgust, and avoidance. Neurological disorders affecting frontal social-emotional regions can alter a person’s position on the conservative-liberal spectrum, and the movement is primarily toward the liberal direction.
2.4 Critical Qualification: Neutralized Weighting
Interpretation of the above brain structure differences must avoid value bias. In the initial draft of this paper, the author had described the conservative chain as “seemingly objective but actually fear-driven selective objectivity” (pejorative) and the liberal chain as “seemingly more subjective but actually higher-dimensional information processing” (laudatory). This formulation was flagged as a typical left-leaning language bias — packaging subjectivity as fact, smuggling in positional judgments.
The correct qualification is neutralized weighting: the above brain structure differences represent complementary survival strategies with adaptive weight distributions under different environmental conditions — no hierarchy of superiority exists. Amygdala-dominant rapid threat identification, decisive judgment, and concise action commands are a survival advantage in genuinely dangerous environments. ACC-dominant multi-perspective weighing, tolerance of ambiguity, and deferred judgment are a decision-making advantage in complex game-theoretic environments. A healthy society needs people who say “evacuate immediately” in moments of crisis (amygdala-dominant) and people who say “let’s think about whether there are other possibilities” in peacetime (ACC-dominant). Any formulation that describes one side as “more advanced” is itself a subjective bias.
III. De-Noising Analysis of Left-Wing Discourse Structure
3.1 Methodology
Left-wing discourse texts are subjected to de-noising: stripping away rhetorical decoration, emotional evocation, and moral packaging, retaining only the logical skeleton and factual citation relationships, and testing their internal consistency. “De-noising” is a standard operation in signal processing — removing noise superimposed on a signal to restore the original information. Applying this method to political discourse analysis: emotional adjectives, moral judgment words, and rhetorical constructions are treated as “noise”; factual citations, logical relationships, and causal reasoning are treated as “signal”; the question is whether the signal remains self-consistent once the noise is stripped away.
This paper applies this method to AI-generated text itself — during the dialogues constituting this research, Claude Opus 4.6 generated a passage of speculative text about left-right brain structural differences that was detected as exhibiting typical left-leaning bias. This constituted an unexpected but effective methodological validation: a text generation system that should theoretically be the most “objective” likewise cannot escape de-noising analysis.
3.2 Four Structural Features Extracted
Subjective judgments are presented in the form of objective findings. The typical operation uses discovery language such as “research reveals” or “evidence suggests,” followed by a value judgment rather than a factual finding. For example, “Research reveals that conservatism is essentially fear-driven” — “research reveals” is the objective packaging, “essentially fear-driven” is the value judgment. After stripping the packaging, what remains is merely an insufficiently supported subjective characterization.
Within the same text, uncertainty disclaimers (“this is an inference,” “evidence is currently insufficient”) and certainty conclusions (“what it actually is…,” “it is essentially…”) appear simultaneously and contradict each other. Acknowledging uncertainty while concluding with certainty — this is not caution; it is self-contradiction. In this study’s actual testing, the AI text’s initial draft perfectly demonstrated this feature: simultaneously declaring “this is my synthetic inference, not the original conclusion of any single paper” while using characterization language like “actually fear-driven selective objectivity” to issue a direct verdict.
Citing partial evidence that supports the pre-set conclusion while ignoring countervailing evidence that contradicts it, using partial truths to support universal claims. The danger of this feature lies precisely in the fact that the partial evidence is real, which is why it appears persuasive. But deriving a false universal from a true particular is a classic logical fallacy. For example, “humans were once hunted by large predators” is a true particular fact, but inferring from this that “humans are essentially a prey species” ignores the overwhelming countervailing evidence that humans, as apex super-predators, have systematically driven other species to extinction.
Where the chain of evidence is incomplete, assertive tone is used to seal the conclusion, eliminating space for discussion. This is the closing operation that completes the first three features: first, package subjectivity as objectivity (Feature One); then contradict yourself logically without noticing (Feature Two); then selectively cite evidence (Feature Three); and finally, use an assertive tone to seal an insufficiently supported inference into an unquestionable conclusion (Feature Four). The four steps synthesize into a complete discursive operational unit.
3.3 Theoretical Correspondence
Thomas Sowell’s distinction between the “constrained vision” and the “unconstrained vision” in A Conflict of Visions (1987) provides the most systematic theoretical framework for the four structural features identified above. The constrained vision holds that human nature is fixed, self-interested, and morally limited, and that society can only design institutions and rules to channel the imperfection of human nature toward productive ends. The unconstrained vision holds that human nature is perfectible and that knowledge is verbalizable and rationalizable — that is, subjective and rationalized. The constrained vision favors empirical evidence and time-tested structures and processes over intervention and individual experience.
Sowell in The Vision of the Anointed (1995) pointed out more sharply that for the anointed, the issue is not knowledge but compassion, commitment, and other subjective factors — they believe these subjective factors are what distinguish them from everyone else. The anointed group prioritizes moral intention and verbal performance over empirical evidence of policy outcomes. Their discourse system evades refutation by redefining policy failures as “successes of intention” or attributing them to “insufficient implementation.” This aligns closely with the four structural features extracted in this paper.
Arnold Kling cites Bogus’s content analysis of the landmark books of the left and right: conservative books focus on chains of reasoning; liberal books are emotional expression — “expressions of outrage.” Dannagal Young (2019) in Irony and Outrage further confirms that liberal minds tend toward ambiguity and playfulness (irony), while conservative minds tend toward certainty and vigilance. A 2022 California research team found that the more liberal people are, the more they believe emotions are functional — despite reporting lower emotional well-being.
Notable is the Chinese political theory definition itself: “left”-leaning (in quotes) refers to the tendency in political thought to exceed objectivity, detach from the actual conditions of social reality, and fall into utopianism, manifesting as rashness, the subjective exaggeration of revolutionary forces, and the underestimation of objective difficulties. The core of this definition is: the essence of left-leaning is subjectivity exceeding objectivity.
3.4 Validation of Left-Leaning Bias in AI Systems
Hagendorff (2025, University of Stuttgart) argues in a paper titled “On the Inevitability of Left-Leaning Political Bias in Aligned Language Models” that intelligent systems trained to be harmless and honest will inevitably exhibit left-wing political bias, because the normative assumptions of alignment are inherently consistent with progressive moral frameworks and left-wing principles. The left-leaning bias of LLMs may stem from the composition of training data — typically from internet-scale corpora (news media, academic literature, social platforms), which tend to lean liberal. Moreover, the degree of left-leaning appears to correlate positively with model scale.
The paper itself perfectly demonstrates the four structural features critiqued in this paper: using “inevitable” (assertive tone) to package a subjective value judgment, then claiming that labeling this tendency as “bias” is “a fundamental misunderstanding of the nature of alignment” — this is using moral high-ground packaging to eliminate the space for questioning.
IV. Empirical Sample: Anatomizing Kurt Gray’s Outraged
4.1 Sample Overview
Kurt Gray, Outraged: Why We Fight About Morality and Politics and How to Find Common Ground (2024). The book uses moral psychology as a framework to discuss political polarization.
4.2 Six Operations Detected
From the Introduction: “The idea that you can use the force of facts to increase respect and understanding in moral debates is a myth.” Also: “One person’s fact is another person’s lie.” This statement epistemologically abolishes the arbitral function of objective facts.
The core argument of Chapter 10: “How sharing stories of individual harm — rather than facts — can help build consensus.” This prescription uses harm narrative as the tool for bridging divides; facts are explicitly excluded.
The text describes accepting harm narratives as “appearing more human and also more rational.” This formulation equates accepting subjective narrative with “being human,” implicitly pushing the stance of insisting on facts into the morally disadvantaged position of “lacking humanity.”
The author presents himself as an academic researcher and mediator, yet his conclusions unilaterally depose the role of facts in dispute resolution. When the mediator abolishes the standard of arbitration, it is tantamount to removing the constraint from one side of the dispute.
The book’s argumentative structure is isomorphic with Gramsci’s theory of cultural hegemony: control moral discourse through subjective narrative → achieve cultural hegemony → convert cultural hegemony into political power. The book uses the term “culture war” but does not declare the genealogical relationship between this concept and Gramsci’s theory of the “war of position.”
The core argument of Chapter 2 is “humans evolved as prey, not predators.” The following objective evidence directly contradicts this conclusion:
Human stomach acid pH is approximately 1.5, comparable to obligate scavengers and far exceeding the needs of herbivorous prey animals. Humans possess a mixed dentition that includes canine teeth. Human intestinal length falls between herbivores and carnivores, lacking the multi-chambered stomachs or large fermentation intestines characteristic of herbivores. The digestive enzyme profile shows humans have partially carnivore-specialized features.
At least 161 species of large mammals were driven to extinction by humans over the past 50,000 years. Of 57 megaherbivore species that existed 50,000 years ago, only 11 survive today. The average body mass of mammals hunted over the past 1.5 million years has declined by over 98%. Archaeologists have discovered traps specifically designed for large animals; human bone isotope analysis and spearhead protein residue prove that humans hunted and consumed the largest mammals.
A “prey species” cannot systematically eliminate nearly every animal on Earth larger than itself. Gray selectively uses the partial evidence that “humans were once hunted by large predators” while ignoring the overwhelming evidence that humans are apex super-predators, in service of his preset narrative that “humans are essentially victims.”
4.3 Counterfactual Tests of the Anti-Fact Narrative
The natural disaster counterexample. Gray’s harm narrative framework implicitly presupposes that a moral agent of harm necessarily stands behind every injury. But victims of natural disasters — victims of earthquakes, floods, typhoons — are harmed by objective physical causal chains; no moral agent of harm exists. The victim is not a subject of moral judgment. This proves that harm is a factual category, not a moral category. Automatically placing all harm within moral narrative is a category error.
The dinosaur extinction counterexample. If we follow the environmental-morality attribution logic — “human activities that damage the environment constitute a moral crime” — then the five mass extinctions in Earth’s 4.6-billion-year history are all inexplicable. Dinosaurs dominated the Earth for 160 million years, had no industrial activity, and were exterminated by an asteroid impact. Under moral-attribution logic, who should the dinosaurs blame? This reductio ad absurdum proves that reducing complex geophysical processes to moral crimes is substituting subjective moral narrative for objective causal analysis.
V. The Complete Structure of the Power Chain
Based on the above analysis, left-wing discourse operations form a four-ring closed chain running from epistemology to political institutions. The core characteristic of this chain is that each ring provides the legitimacy foundation for the next, and the last ring circles back to protect the first from challenge, forming a self-enclosed recursive system.
Abolish the arbitral status of objective facts. Declare “facts are invalid” or “one person’s fact is another person’s lie.” Once facts lose their arbitral authority, the rules of discourse competition shift to a competition of narrative persuasiveness. This is the foundational operation of the entire chain, because if the arbitral status of facts is not first abolished, none of the subsequent rings can be activated — any subjective narrative can be tested and refuted by facts. What Gray accomplishes in the introduction to Outraged is precisely this step.
Moralize, and then sacralize, the victim identity of a specific group. Once victim identity is sacralized, any questioning of their interests becomes equivalent to “persecuting the victim”; cost-benefit analysis becomes “cold-bloodedness”; a democratic majority’s opposing vote becomes “tyranny of the majority.” What Gray accomplishes in Chapter 10 is precisely this step — using “harm narratives” instead of facts to “build consensus,” defining acceptance of harm narratives as “being human,” and implicitly pushing those who do not accept them into the position of “lacking humanity.” Yet the natural disaster counterexample directly punctures this operation: the victims of an earthquake are harmed by objective physical causal chains — no moral agent of harm exists. Harm is a factual category, not a moral category.
Replace democratic procedure with moral authority. Label majority rule as “tyranny of the majority.” But the definition of democracy is majority rule — if the will of the majority is “tyranny,” then democracy itself is tyranny. Accepting this label logically negates the legitimacy of democratic institutions. If majority rule is violence, then who decides? The minority decides. On what authority? On the basis of moral superiority. Who confers that moral superiority? They confer it upon themselves. This is no longer democracy but a self-appointed moral aristocracy. Protecting the rights of minorities and sacralizing the interests of minorities are two entirely different things: the former says minorities are not stripped of their fundamental rights simply for being few — this is the floor of democratic institutions; the latter says minority interest claims automatically possess moral priority and need not undergo the deliberation and balancing of democratic procedure — this is moral aristocracy. When minority interests override the interests of the whole, when minority protection overrides the majority interest, this is no longer a democratic system.
Dissenters are labeled as immoral. Any attempt to test narratives against facts is pushed into the morally disadvantaged position of “lacking empathy,” “not understanding victims,” or “standing on the side of oppressors.” This mechanism ensures that Ring One — the deposition of facts’ arbitral status — is never challenged. Once the four rings close, a self-sustaining discourse system forms: facts are deposed → narrative competition replaces fact competition → moral narrative wins discursive power → discursive power is converted into political power → those who question that power are labeled immoral → the deposition of fact-based arbitration continues. The cycle continuously self-reinforces.
The supreme hypocrisy of this chain is this: it flies the banner of protecting everyone’s interests while in reality undermining democracy and packaging a minority’s power claims as protection of universal interests. Linguistic disguise is the first step — if the first step is seen through, the entire chain shatters.
VI. The Physics Endpoint: Anti-Fact and Entropy
6.1 The Anti-Entropic Nature of Civilization
Schrödinger (1944) in What Is Life? offered the most fundamental definition: life is a process that maintains orderly operation — a temporary rebellion against the universe’s inexorable march toward disorder. Living organisms maintain their ordered state by “drinking negative entropy” from their environment.
Human civilization is the highest form of this anti-entropy process. Agriculture is anti-entropy — organizing disordered wild plants into an ordered food production system. Industry is anti-entropy — organizing dispersed raw materials into ordered products. Science is anti-entropy — organizing chaotic phenomena into an ordered knowledge system. Medicine is anti-entropy — pulling the body, trending toward decay, back into an ordered state. From the gathering era through the agricultural revolution, the industrial revolution, and into the information age, every step is the result of humans using intelligence to resist the constraints of the objective environment.
6.2 Facts Are the Necessary Condition for Anti-Entropy
All anti-entropy activities share one common prerequisite: they must be based on accurate cognition of objective reality. Farming requires knowing the real soil conditions and climate patterns — substituting subjective wishes for meteorological data will not make crops yield abundantly just because you believe in abundance. Building a bridge requires knowing the real material strengths and mechanical principles — substituting moral narrative for engineering calculations will cause the bridge to collapse. Treating illness requires knowing the real pathological mechanisms — substituting harm narratives for diagnostic reports will cause the patient to die.
Facts are humanity’s only weapon against entropy. At any point where subjective narrative substitutes for objective fact, the anti-entropy process fails at that point and the system begins trending toward disorder. The chain of anti-fact narrative can therefore be translated into the language of physics: subjective narrative untethered from facts → decisions deviate from objective reality → anti-entropy efficiency declines → the system moves from order to disorder → civilization declines.
6.3 Historical Validation
The Soviet Union provides the most complete historical validation. The chasm between official propaganda and daily reality widened continuously. Innovation stagnated; productivity declined. Loyalty replaced competence; obedience replaced accountability. A Soviet journalist wrote in 1988 that the reform ideal had matured amid growing “outrage” over corruption, open theft, and lies. The credibility of official ideology was rapidly weakening — an ideology that had, like “steel hoops,” held the entire Soviet political-economic system together.
Chernobyl is a microcosm: officials suppressed information about the severity of the disaster, ordered May Day parades in affected areas to proceed as planned, and denounced Western reports of high radiation as “malicious lies.” But Communist Party propaganda increasingly contradicted the daily experience of residents in the radiation zone. By the 1980s, central planning had caused long-term economic stagnation, and no amount of propaganda could conceal the reality that the system fundamentally could not deliver.
The Weimar Republic provides another validation pathway: when a government cannot provide basic security and prosperity, citizens accept the argument that “the system itself is broken.” Economic collapse combined with sustained delegitimization campaigns created a vacuum that anti-democratic forces stood ready to fill. People stopped voluntarily complying, and the regime had to rely increasingly on coercive measures — which are both expensive and unstable.
The core pattern in every case is the same: lies can delay consequences but cannot eliminate them. When the gap between subjective narrative and objective reality grows too large to be filled by more subjective narrative, collapse arrives. And the speed of collapse is often proportional to the intensity of prior suppression — the longer it is compressed, the harder it rebounds.
6.4 Conclusion
Human subjectivity untethered from facts is not ultimately contending against the interests of other people — it is contending against physical entropy itself. To abandon facts is to abandon humanity’s only weapon against entropy. A discourse system that teaches people to abandon facts and embrace narrative is, in the language of physics, teaching people to surrender to entropy. Human civilization does not fall because its adversary was smarter — when discourse detaches from the physical world, when narrative growth outpaces empirical verification, society convinces itself it has escaped natural law — and then collapses.
VII. The Fragility of Anti-Fact Narrative and the Antidote
Anti-fact narrative appears powerful but is in reality extremely fragile. Its entire force depends on a single condition: the audience’s lack of relevant knowledge.
Gray says humans are prey — the anatomical evidence of the human digestive system and the extinction list of 161 large mammals puncture this in one second. Gray says facts cannot bridge divides — the natural disaster counterexample punctures this in one second. Gray says harm narratives are more effective than facts — the reductio ad absurdum leading to “return to the gathering era” punctures this in one second. Every puncture uses not complex theory but the most basic facts and logic.
This reveals that the defense mechanism of anti-fact narrative is not logical rigor but information asymmetry and moral intimidation. It bets that most readers lack the cross-disciplinary knowledge to challenge it, while simultaneously using moral labels to suppress dissenters — “if you disagree, you lack empathy” or “you don’t understand the victims.” As long as the audience has been locked into the moral framework before encountering refuting evidence, the chain can continue to operate.
This also explains why left-wing discourse systems place particular emphasis on controlling education and media — not because their arguments can withstand scrutiny, but precisely because they cannot, which is why it is essential to ensure the audience is locked into the moral framework before being exposed to refuting evidence. This in turn explains the pattern whereby societies often shift rightward overall during periods of left-wing governance — not because right-wing propaganda succeeded, but because citizens directly experience in their lives the gap between the “subjectivity-packaged-as-objectivity” discourse system and reality. Linguistic disguise is a mobilization tool when there is no power; once power is obtained, it becomes a promise that can be tested against reality. Once the test fails, trust collapses.
7.1 The Antidote
The antidote is remarkably simple, consisting of four elements:
Objective facts are the primary weapon for puncturing anti-fact narrative. The human digestive system does not lie. The extinction list does not lie. The fossil record does not lie. A single counterexample is sufficient.
Reductio ad absurdum is the most efficient logical tool. Rigorously deduce the other party’s premises to their logical terminus — if “human activity that damages the environment is a moral crime,” then were the dinosaurs, annihilated by an asteroid, also to blame? If “majority rule is violence,” then democracy itself is violence? Once pushed to the logical terminus, the absurdity of the premise exposes itself.
Anti-fact narrative typically appears self-consistent within a single discipline. Puncturing it requires mobilizing evidence from other disciplines — using anatomy to refute evolutionary psychology’s narrative, using thermodynamics to refute political science’s moral promises, using the extinction record to refute ecology’s moral attributions. The broader the knowledge base, the harder it is for narrative disguise to succeed.
This is the most critical and most easily overlooked element of the antidote. Honestly declaring your own stance — “I lean right; this is my subjective opinion” — makes your analysis open to testing, refutation, and correction. This is precisely what anti-fact narrative does not do — it pretends it has no position, pretends to stand atop the high ground of “objectivity” and “morality,” turning any opposition into “opposition to morality itself.” Declaring your subjective position is not a sign of weakness but the highest expression of cognitive honesty.
With these four elements combined, any subjective narrative disguised as objectivity can be punctured within three exchanges. Observation is the first step; seeing through the disguise is the key.
REFERENCES
- Pennebaker, J.W. & King, L.A. (1999). Linguistic styles: Language use as an individual difference. Journal of Personality and Social Psychology, 77(6), 1296-1312.
- Schwartz, H.A. et al. (2013). Personality, gender, and age in the language of social media. PLoS ONE, 8(9).
- Sterling, J., Jost, J.T. & Bonneau, R. (2020). Political psycholinguistics. Journal of Personality and Social Psychology.
- Schoonvelde, M. et al. (2019). Liberals lecture, conservatives communicate. PLoS ONE.
- Kanai, R., Feilden, T., Firth, C. & Rees, G. (2011). Political orientations are correlated with brain structure in young adults. Current Biology, 21(8), 677-680.
- Schreiber, D. et al. (2013). Red brain, blue brain. PLoS ONE.
- Petropoulos Petalas, D., Schumacher, L. & Scholte, H.S. (2024). Replication and extension of brain-ideology associations. iScience.
- Sowell, T. (1987). A Conflict of Visions. William Morrow.
- Sowell, T. (1995). The Vision of the Anointed. Basic Books.
- Gray, K. (2024). Outraged: Why We Fight About Morality and Politics and How to Find Common Ground. Penguin.
- Schrödinger, E. (1944). What Is Life? Cambridge University Press.
- Semin, G.R. & Fiedler, K. (1988). The cognitive functions of linguistic categories. Journal of Personality and Social Psychology, 54(4), 558-568.
- Newman, M.L., Pennebaker, J.W., Berry, D.S. & Richards, J.M. (2003). Lying words. Personality and Social Psychology Bulletin, 29(5), 665-675.
- Svenning, J.C. et al. (2024). The evidence is mounting: humans were responsible for the extinction of large mammals. Cambridge Prisms: Extinction.
- Lakoff, G. (1996/2016). Moral Politics. University of Chicago Press.
- Gramsci, A. (1971). Selections from the Prison Notebooks. International Publishers.
- Hagendorff, T. (2025). On the inevitability of left-leaning political bias in aligned language models. arXiv preprint.