On the Abuse of Words
I have a confession. I’ve worked at the Cato Institute for 7 months and have barely read any of Cato’s Letters—the eighteenth-century essays on liberty by John Trenchard and Thomas Gordon that gave the organization its name. The pen name honored Cato the Younger, the ancient Roman senator who chose death over submission to Caesar’s tyranny.
I’m a psychologist, not a political philosopher, and the founding texts of classical liberalism were not assigned reading in my Harvard neuroscience seminars. But a smart colleague recently pointed me toward Letter 117, titled “Of the Abuse of Words,” and I haven’t been able to stop thinking about the first two sentences:
“I have often thought, that most of the mischiefs under which mankind suffers, and almost all their polemick disputes are owing to the abuse of words. If men would define what they mean by the sounds which they make use of to express their thoughts, and then keep to those definitions, that is, annex always the same ideas to the same sounds, most of the disputes in the world would be at an end.”
In plain terms, the world would be so much better if people were clear, honest, and straightforward. Most of the evil in the world is some downstream result of dishonesty or miscommunication. The result of the abuse of words.
What makes the abuse of words so grave is not just that clarity and truth is pragmatically useful. Philosophers across traditions have treated truth as sacred.
Truth is a sacred thing because it is, in some ultimate sense, unattainable. We do not perceive reality directly but only through the veil of our perception, our evolved biological constraints and cognitive biases. Socrates intuited this in declaring that he knew nothing, and considered that admission the starting point in all wisdom.
In theory, everything is knowable. Physicists dream of a “theory of everything”—a set of equations which could describe and predict all things that ever have happened or will happen. But metaphysical questions of what came before time and Being—if such questions even makes sense—or where our laws of physics came from in the first place, may be fundamentally unanswerable. Gödel proved that even mathematics cannot fully ground itself, resting on axioms it can never verify from within. If the most precise language humans have ever invented cannot capture the whole truth, what hope does ordinary speech have?
Kierkegaard understood that honest reasoning, pursued to its limit, arrives at a threshold where only a leap of faith can carry you further. Religious traditions across history have tried, with varying claims of certainty, to address that void. The opening line of the Tao Te Ching warns that the Tao that can be spoken is not the true Tao—the God that we can conceive of is inherently limited by human subjectivity and preconceptions. Humility is the right approach. The Gospel of John begins at the other side of that leap: the Word or Logos, the ordering principle that precedes and exceeds human understanding, is deemed unknowable but through Christ.
Spinoza, one of my greatest intellectual inspirations, took a different approach. He recognized that the grand Truth we so crave is fundamentally unattainable to humans, but that the universe has an inherent order to it. An omnipotent and omniscient God is essentially a personification of the forces of Nature that describe and govern all things. For Spinoza, Deus sive Natura—God or Nature—were one in the same.
Spinoza believed that to live according to reason was to live according to one’s true nature, and that deception, including self-deception, was a form of bondage. Kant argued that lying was always wrong, even to a murderer at the door, because dishonesty corrodes the very faculty of reason that makes moral life possible. Both of them believed that the pursuit of truth was a form of ethical striving, what Spinoza called the “intellectual love of God”:
“In so far as we are intelligent beings, we cannot desire anything save that which is necessary, nor yield absolute acquiescence to anything, save to that which is true: wherefore, in so far as we have a right understanding of these things, the endeavor of the better part of ourselves is in harmony with the order of nature as a whole.”
Humans know that truth is valuable and that we are ignorant in some fundamental sense. But we also crave certainty, and have remarkable capacity for reason. With that remarkable capacity also comes a temptation to manipulate words and claim false certainty.
Evolution exists in an unsteady dance with truth. On one hand, there is clear survival value in ability to perceive the world as objectively and accurately as possible. We have multiple senses because convergent detection of the same stimulus through vision, smell, sound, and touch greatly enhances our belief that we are perceiving something real. On the other hand, there is clear adaptive utility in certain cognitive biases that orient away from objective perception. We evolved to be threat-sensitive, because the relative cost of mistaking a stick for a snake is trivial compared to the cost of mistaking a snake for a stick.
Dishonesty also pays for motivated actors, as we know all too well from politics. But this too is a deep evolutionary instinct that plays out whether we are conscious or not. As evolutionary biologist Lixing Sun describes in The Liars of Nature and the Nature of Liars, deception is among the oldest and most pervasive strategies in the biological world. Orchids mimic the pheromones of female wasps to lure pollinators. Fireflies imitate the mating signals of other species to attract and devour them. Cuttlefish display female coloration on one side of their body to deceive rival males while courting a female on the other. These organisms are not lying in any conscious sense, but they are manipulating information to gain an advantage, and natural selection has rewarded them handsomely for it.
Humans do the same thing with language. We can deceive with words, with tone, with omission, with framing. We can deceive others and ourselves simultaneously, and often do both at once without realizing it. In competitive environments, the ability to manipulate information is often more advantageous than the ability to share it honestly. Honesty is a cooperative strategy, but cooperation is fragile. It flourishes only when the systems surrounding it make deception costlier than truth-telling. When those systems fail—or when the incentives flip—the abuse of words is the default.
I was a graduate student at Harvard on October 7th, 2023, the day Hamas launched the deadliest attack on Jewish civilians since the Holocaust. Over 1,200 people were murdered, many of them brutally on camera. Within 48 hours, student organizations released statements that did not merely criticize Israeli policy—a position one can hold in good faith—but characterized the massacre as an act of “resistance.” The word “decolonization” was deployed to redescribe the slaughter of civilians as a political project with moral legitimacy. Posters of kidnapped hostages were torn down, and the people tearing them down did not think of themselves as cruel. They thought of themselves as righteous. They had been given words that made terrorism feel like justice.
When the news broke, people around me seemed uncertain how I would react—as though my ethnic background predetermined my moral position, as though having Palestinian heritage meant I was obligated to see the massacre through a particular ideological lens.
Instead, my undergraduate students seemed to feel surprisingly validated when I expressed my own dismay at the situation, in polite terms, that this was a tragic clusterfuck. The attacks on Israel were horrific. The attacks on Gaza were horrific. The generations of mutual grievance, displacement, and retaliatory violence in the unholy Holy Land made any solutions seem impossible.
And yet I kept thinking: as far as we are from real progress, it would be tremendous progress if we could simply come to unilateral agreement that murdering civilians in their homes is wrong. It is bad enough that humanity still engages in war. It is bad enough that violence, terrorism, and cruelty persist in a world with the material abundance, institutional knowledge, and historical memory to know better. But it is much worse that we gaslight about it. That we dress atrocities in the language of liberation. That we redefine terrorism as resistance, civilians as collateral, and massacre as a complex geopolitical event requiring “context.” Language had been so thoroughly abused that the most basic ethical clarity and nuance became a controversial position in seminar rooms and dining halls.
Wolves and dogs cycle through an elaborate sequence of warnings before a fight: stiffened posture, growl, snarl, snap, and only then a bite. Each signal is an honest communication—an escalating declaration of intent that gives the other party a chance to back down. The system evolved this way because violence is expensive. Even the winner often gets hurt. The entire architecture of threat signaling exists to make actual combat unnecessary, and it works precisely because the signals are reliable. When a dog growls, it means something.
Human diplomacy follows the same logic at a larger scale. Sanctions, troop movements, trade restrictions, and ultimatums are the geopolitical equivalents of the growl and the snarl—costly signals designed to resolve disputes before blood is shed. The system breaks down when the signals become dishonest. When genuine threats are dismissed as posturing, the signaling architecture collapses and the probability of violence increases. And the abuse of words—redescribing aggression as defense, conquest as liberation, terrorism as resistance—is perhaps the most dangerous coordination failure of all, because it doesn’t just misrepresent the present. It makes the next step impossible to negotiate honestly.
How much moral progress would humanity make if we did not become more moral in our actions, but simply achieved universal consensus on the immorality of certain actions?
Every act of cruelty is preceded by an act of redescription. The abuse of words comes first. The torturer must first dehumanize. The mob must first be told it is the victim. The suicide bomber must first be convinced that death is glory. a refusal to see the other person as fully human, a motivated redefinition of who counts as innocent, a linguistic framework that transforms murder into something else.
Imagine a world where every act of violence still occurs—every war, every terrorist attack, every atrocity—but no one pretends it is justified. No one calls it resistance. No one frames it as decolonization. No one wraps it in the language of historical grievance or cosmic justice. Everyone who commits violence knows, and admits, that they are doing something wrong to a fellow human, and everyone who witnesses it agrees.
I suspect, as Cato’s Letter 117 argues, that such clarity would make the steps towards true progress suddenly obvious.
“But this would not answer the purposes of those who derive power and wealth from imposing upon the ignorance and credulity of others. And therefore, till the world can agree to be honest, and to buy and sell by the same measure (which they do not seem in haste to do), I doubt this evil is likely to go on.”
The letter goes on to discuss religious bigotry and the way zealots weaponize piety to serve their own interests.
We see the same abuse of words, weaponizing piety and sympathy, in my primary research field of child development and mental health. Every time there is a moral panic—whether over television, comic books, video games, or the present-day fears about AI and social media—people look to galvanize their arguments by invoking the vulnerability of children. The pattern is remarkably consistent. A new technology emerges. Children use it. Some children who use it have problems. The technology is blamed, and the word “harm” is attached to it before the causal evidence exists.
Jonathan Haidt’s bestselling The Anxious Generation argues that smartphones have “rewired” childhood and caused an epidemic of mental illness—language so vivid and morally urgent that it has already catalyzed legislation in multiple countries. But the largest meta-analysis on the subject, reviewing 143 studies featuring data from over a million adolescents worldwide, found associations that were small, inconsistent, and heavily confounded by personality and pre-existing mental health. The word “rewired” does not appear in that meta-analysis. Neither does “epidemic.” Those are rhetorical choices, not scientific findings, and they have done more to shape policy than the underlying data ever could.
Once “harm” is established in the public vocabulary, the debate shifts from “is this actually dangerous?” to “how do we protect children from this danger?”—a question that presupposes the answer to the one we skipped.
The COVID-19 pandemic followed the same rhetorical playbook in the opposite demographic direction. “Protect the vulnerable elderly” became the phrase that foreclosed cost-benefit analysis of lockdown policies. To ask whether school closures were doing more damage to children than the virus itself was framed not as a legitimate empirical question but as an act of callousness toward grandparents. The word “protection” was annexed to a specific set of interventions—lockdowns, closures, mandates—rather than to the broader obligation to weigh all harms against all benefits. Anyone who questioned whether the cure was worse than the disease was accused of not caring about the vulnerable, when in reality the question was about which vulnerable population we were willing to sacrifice for another.
This is how the abuse of words operates in policy: by attaching a morally charged term to one side of a genuinely difficult trade-off, so that the other side becomes unspeakable. “Save the children” and “protect the elderly” are not arguments. They are conversation-enders disguised as moral imperatives. And they work precisely because the underlying sentiments are real. We should care about children. We should care about the elderly. The abuse lies not in the caring but in the weaponization of that caring to shut down the very inquiry that responsible caring demands.
The worst medical sophistry has come in the field of transgender medicine, the subject of my doctoral dissertation. The phrase “gender-affirming care” is perhaps the most sophisticated example of the abuse of words in contemporary policy. It takes a set of interventions—puberty blockers, cross-sex hormones, surgical procedures—whose long-term effects on developing bodies and brains remain poorly understood, and packages them in language that makes skepticism sound like cruelty. To question “affirming care” is to be, by definition, non-affirming. (I’ve quipped that a better phrase for them would be “dysphoria-enabling interventions”).
A clinician who recommends watchful waiting for a distressed adolescent, or who explores whether anxiety or social contagion might be contributing factors, is not failing to affirm the child. They are practicing medicine. But the vocabulary has been arranged so that caution registers as hostility, and the word “care” has been annexed to a specific set of interventions rather than to the broader obligation to do no harm. Meanwhile, exploratory psychotherapy—the approach that the majority of countries that have independently reviewed the evidence now recommend as a first-line response—has been rebranded as “conversion therapy,” a term that carries the moral weight of a practice historically used to torture gay people into straightness. The conflation is not accidental. It is designed to make the most cautious, evidence-based clinical response feel morally equivalent to abuse.
The same rhetorical machinery that rebranded experimental medicine as “care” and exploratory therapy as “conversion” extended far beyond the clinic. Once the premise was established that gender identity overrides biological sex, the policy implications of the abuse of words followed. Biological males were placed in women’s prisons, and women inmates who objected were told their discomfort was bigotry. Transgender athletes with the physiological advantages of male puberty competed in women’s sports for over a decade, and female athletes who raised concerns about fairness were accused of hatred. Feminists who pointed out that sex-based protections existed for a reason—that shelters, sports, and prisons had been segregated by sex to protect women from male violence—were called exclusionary and hateful, often by the same progressives who claimed to champion women’s rights. An entire generation of young feminists learned that defending sex-based categories was impermissible, even as the practical consequences of abolishing those categories fell disproportionately on women.
Pediatric gender reassignment will go down in medical history alongside lobotomies—invasive, irreversible procedures celebrated by the profession, even winning the Nobel Prize in Medicine in their heyday—supported by thin evidence, and abandoned only after the damage was done. Almost all of it is now being walked back, and the speed of the reversal is itself an indictment of how flimsy the original consensus was. But what makes the gender ideology debacle distinctly modern is the role that the abuse of words played in sustaining it long past the point where the evidence had turned, and long past the point where the collateral damage to women’s sports, women’s safety, and feminism itself had become undeniable. And the abuse of words continues even in the retreat. The NCAA did not say it was wrong for fifteen years. It said it was adopting “clear, consistent, and uniform eligibility standards.” The AMA did not acknowledge that it had urged governors to protect access to procedures it now calls insufficiently supported. The same organizations that abused words to build the edifice are now abusing words to dismantle it quietly, hoping no one notices the rubble. The reckoning that honest language would make possible keeps getting deferred by the same dishonesty that made it necessary.
The thread that runs through all of these examples—from Hamas apologists on college campuses to pandemic policy to transgender medicine—is the same one articulated in Cato’s Letters three centuries ago. The abuse of words is not a series of isolated rhetorical failures. It is a single mechanism operating across every domain where power, fear, and moral urgency converge. Diplomacy works when signals are reliable. Medicine works when words like “care” and “harm” refer to observable realities rather than ideological commitments. Societies work when the people in them can say what they mean and mean what they say.
The temptation to abuse words will never disappear. Evolution built it into us, right alongside our capacity for reason. We are creatures who crave certainty in a universe that offers none, who possess language powerful enough to reshape perception but too imprecise to capture the whole truth, and who discovered long ago that manipulating information is often easier than sharing it honestly.
But recognizing that the temptation is permanent is not the same as accepting that it is insurmountable.
The authors of Cato’s Letters knew that the powerful would never willingly surrender the advantage that dishonest language provides. They said as much. But naming the problem clearly was itself an act of defiance against the very thing they described. The abuse of words sustains itself partly by making the distortion invisible, by ensuring that the euphemism feels like common sense. Simply insisting on clarity—this word means this, and you are using it to mean something else—is the first crack in the edifice.
That is what honest inquiry demands. Not certainty, which we cannot have. Not moral perfection, which we will never achieve. Just the discipline of meaning what we say and saying what we mean, even when, and especially when, the incentives run the other way. Define your terms. Hold to your definitions. Refuse to let language be turned against the things it was meant to describe. It is the simplest demand imaginable, but if we’d manage it, the solutions to insurmountable moral dilemmas might suddenly become clear.