On Language, Violence, and the Limits of Legal Definitions
A Response to "Speech is Not Violence"
My longtime friend and respected colleague Andrew Horn's recent essay "Speech is Not Violence" tackles crucial questions about language, power, and democratic discourse. Andrew and I share deep concerns about protecting free speech and maintaining the conditions for meaningful dialogue in polarized times. His essay makes important points about avoiding the weaponization of speech restrictions and offers valuable methodological guidance for evidence-based discourse. His methodological approach (checking sources, grounding arguments in evidence, and maintaining compassion in debate) offers valuable guidance for navigating polarized times. However, while I appreciate Andrew's intentions and share many of his concerns, his central argument rests on an unfortunate conflation of legal definitions with conceptual reality, and misses crucial dynamics of how language operates within existing power structures. This response is offered in the spirit of advancing our shared conversation about these vital issues.
The Definitional Authority Problem
Andrew presents legal definitions of violence as if they settle the broader conceptual question: "Violence (strict sense): intentional physical force causing injury or damage." But this approach puts the cart before the horse. Language preceded law, not the reverse. Legal definitions serve specific institutional purposes: they need to be concrete, enforceable, and limited in scope. But they don't exhaust the meaning or reality of the concepts they attempt to govern.
Moreover, legal definitions are transient and contingent. If Congress tomorrow passed a law defining certain speech as violent, would Andrew's position suddenly change? Laws are not morality. They occasionally align with moral behavior, but this seems more like bureaucratic synchronicity than principled foundation. To ground our understanding of violence in whatever happens to be legally codified at the moment is to build on shifting sand.
The etymology of "violence" traces back to the Latin violentia, meaning vehemence or impetuosity, and violare, meaning to treat with violence or to violate. Historically, "violent" has described things that are disruptive, forceful, or harmful in ways that extend far beyond physical force. To reduce violence to its narrowest legal definition is to ignore both the historical development of the concept and its lived reality.
More fundamentally, Andrew's argument assumes definitional authority without acknowledging it. When he states "words, while they may be hurtful, offensive, or inappropriate, are not violence," he presents this as objective fact rather than a particular framing that serves specific institutional purposes. This itself is an exercise in linguistic power: using language to control how we think about language.
The Violation of Autonomy
There's something deeper at work here about the concept of violation, which sits at the etymological heart of violence. In nonviolent communication practices (which Andrew explicitly rejects in favor of "objective communication"), there's recognition that certain uses of language violate another person's autonomy and subjective experience. When I tell you what you think or feel, when I make claims about your internal state without your consent, I'm violating the boundary of your subjective experience.
Consider the difference between saying "Charlie Kirk is a racist" versus "When I hear Charlie Kirk speak about certain topics, I feel a sense of alarm and concern about the impact on communities I care about." The first statement claims objective knowledge about another person's internal state and motivations. The second acknowledges the speaker's subjective experience while respecting the fundamental unknowability of another's inner world.
This violation of autonomy represents a form of linguistic violence that goes beyond hurt feelings. It's an assertion of power over another person's self-definition and internal reality. In men's work, which Andrew leads, we use "I statements" precisely because of this recognition. The foundational practice of speaking from our own perspective rather than making claims about others' internal states creates the safety necessary for authentic dialogue. This isn't just politeness; it's acknowledgment that there's something inherently coercive about claiming authority over another's subjective experience.
Missing the Institutional Dimension
Andrew's framework fails to account for how speech operates within existing institutional structures to cause direct physical harm. Consider these scenarios:
The Psychiatric Weapon: A classic example from the 1950s involves men who, having been unfaithful to their wives, would manipulate psychiatric institutions to silence them. The husband brings his wife to a psychiatrist and, through careful framing of her legitimate anger and distress as symptoms, secures a diagnosis of hysteria. She is medicated against her will and institutionalized, sometimes for years. The man never laid a hand on her, yet his words directly caused her physical confinement, forced medication, and bodily harm. Under Andrew's definition, this wouldn't qualify as violence because no direct physical force was used. But this seems to miss something essential about what occurred.
Bearing False Witness: I testify in court that I saw you commit a crime, knowing this to be false. My words directly lead to your imprisonment or worse. The commandment against bearing false witness sits alongside "thou shall not kill" because the consequences can be equally devastating. This person might still be alive, still be free, if I had not spoken these words. The causal chain from my speech to their physical harm is clear and direct.
These examples reveal a blind spot in privileging direct physical force over the complex ways language participates in systems that cause bodily harm. When speech becomes a necessary component in a causal chain leading to violence, the distinction between "speech" and "violence" becomes less clear-cut.
The Direct Biological Impact of Language
Even if we accept Andrew's narrow focus on physical harm, there's compelling evidence that language can directly cause measurable biological damage. Negative self-talk and verbal abuse don't just hurt feelings—they trigger stress responses that have concrete physiological effects.
Research consistently shows that harsh, critical, or threatening language can:
Elevate cortisol levels chronically
Suppress immune function
Increase inflammation markers
Contribute to cardiovascular disease
Accelerate cellular aging
Impact neuroplasticity and brain development
In extreme cases, the cumulative physical toll of linguistic abuse can contribute to life-threatening conditions. When I use the sounds from my mouth with intention to create stress responses in you that damage your cardiovascular system over time, how is this fundamentally different from using my fist to break your bone? Both involve using my body to directly cause physical harm to yours (one is just slower and harder to trace).
This isn't metaphorical harm or hurt feelings. It's measurable, objective, biological damage caused by specific sound waves organized into language. If we're going to insist on defining violence in terms of physical impact, then we must grapple with the reality that words can and do cause physical impact.
Beyond the Charlie Kirk Example
While Andrew's essay emerges from recent events involving Charlie Kirk, the question of whether language can be violent doesn't depend on parsing Kirk's specific statements. Whether or not Kirk's language was violent, language generally can be violent. This isn't about any particular speaker but about the nature of language itself as a tool that can cause real harm through various mechanisms.
The focus on specific examples can distract from the deeper conceptual question: do we understand language as merely symbolic representation, or as a form of action that can directly impact the physical world? The evidence suggests the latter.
The Spectrum of Linguistic Participation in Violence
Rather than a binary distinction between speech and violence, consider a spectrum of how language participates in violent outcomes:
Direct Command Violence: I have a voice-activated gun that fires when I say "shoot." When I say "shoot" and it kills someone, this seems clearly to be violent speech (language that directly causes physical harm through technological mediation).
Distributed Decision Violence: The gun requires a majority of people present to say "shoot" before it fires. When I add my voice to the chorus, am I participating in violent speech? My individual contribution may be necessary but not sufficient for the violent outcome, yet I'm still causally implicated.
Influenced Decision Violence: A group of decision-makers control the gun, and they listen to advisors. When I, as an advisor, use language to influence them toward shooting, the causal relationship becomes more complex but doesn't disappear. My speech shapes the violent outcome, even if more indirectly.
This spectrum suggests that the question isn't simply "is speech violent or not?" but rather "how does speech participate in violent systems, and what level of participation are we willing to tolerate?"
Language as Weapon
If we can't agree that language is inherently violent, perhaps we can at least acknowledge that language, like any tool, can most certainly be used as a weapon. A hammer can build a house or break a skull. The tool itself may be neutral, but its deployment determines its moral character. Similarly, words can heal wounds or inflict them, depending on the intention and context of their use.
The Violence of Institutional Suppression
Having experienced psychiatric institutionalization, I've witnessed how language gets weaponized to suppress those who've already been victimized. When someone speaks uncomfortable truths—about family abuse, institutional failures, or systemic harm—there's often a violent use of language and institutional power to pathologize, discredit, or silence them.
This represents a particular form of linguistic violence: using diagnostic language, legal procedures, or institutional authorities to inflict physical confinement and bodily harm on people whose speech threatens existing power structures. The violence isn't in the initial speech but in the institutional response designed to silence it.
Law, Morality, and the Violence Monopoly
Andrew's reliance on legal definitions obscures a deeper irony: law itself operates through violence. The entire legal system rests on what Max Weber called the state's "monopoly on legitimate violence." Every law, every court order, every legal proceeding is ultimately backed by the threat of physical force. As Tolstoy observed, to engage with legal mechanisms is to invoke this violent apparatus of the state.
Andrew talks about using "legal protections" and making "specific actions and behaviors legal or illegal through the law," but as Tolstoy correctly observed, to engage with legal mechanisms is to invoke the violent apparatus of the state. This isn't necessarily wrong, but it reveals the inadequacy of simply saying "speech is not violence" while simultaneously relying on violent institutions to regulate speech.
Pragmatic Acceptance vs. Conceptual Denial
Perhaps the more honest framework is this: language can indeed be violent, but we choose to tolerate certain levels of linguistic violence for the sake of other values, like free expression and democratic discourse. This parallels our approach to guns, as Charlie Kirk himself has noted: if we're going to allow firearms, we accept that some innocent people will die. You might disagree with this trade-off, but it's the reality of the choice we've made as a society.
Similarly, if we're going to have robust free speech, we're going to have some casualties from the violence of that language. The question becomes whether the benefits outweigh the costs, not whether the costs exist at all.
Andrew briefly gestures toward this when discussing free speech as "not the best option, but the least worst." But rather than extending this logic to acknowledge that we tolerate some linguistic violence, he insists on definitional purity that language simply isn't violent.
Ancient Wisdom on Language's Power
The recognition that language can cause real harm isn't new. James 3:6 warns that "the tongue is a fire, a world of unrighteousness. The tongue is so set among our members that it defiles the whole body, and sets on fire the course of nature; and it is set on fire by hell." James continues in verse 10: "Out of the same mouth come blessing and cursing. My brothers, these things ought not to be so."
These biblical texts formed part of the foundation of modern Western law, yet they recognized something that Andrew's legal framework misses: language has always been understood as having genuine power to harm or heal. The question isn't whether this power exists, but how we choose to wield it.
A Different Path Forward
I largely agree with Andrew's practical concerns about protecting free speech and avoiding the weaponization of expanded violence definitions. But I think we can hold these positions while being more intellectually honest about the nature of language and power.
Instead of denying that speech can be violent, we might say: "Yes, language can participate in violent systems and cause real harm, but the costs of broadly restricting speech outweigh the benefits. We choose to tolerate some linguistic violence for the sake of democratic discourse, just as we tolerate some physical violence in sports or law enforcement."
This approach acknowledges the legitimate concerns of those who experience language as violent while maintaining strong free speech protections. It also opens space for more nuanced conversations about when and how we might want to limit the most harmful forms of linguistic violence—through counter-speech, social consequences, or in extreme cases, legal intervention.
The Meta-Question
Finally, there's something almost poetic about this entire debate. In arguing that "speech is not violence," Andrew is using speech to shape how we think about speech itself. If we accept that language can be a tool of power that shapes institutional reality, then Andrew's definitional claims become their own exercise in linguistic power.
This doesn't make his argument wrong, but it does suggest we should be suspicious of any claim to definitional authority that presents itself as mere description rather than prescription.
Conclusion
The question of whether language can be violent touches on fundamental issues of power, causation, and moral responsibility. Rather than settling for binary answers, we might embrace the complexity. Language exists on a spectrum from clearly non-violent expression to direct participation in violent systems. Where we draw lines will always involve pragmatic judgments about what kinds of harm we're willing to tolerate for what kinds of benefits.
What matters most is that we make these judgments explicitly and democratically, rather than hiding them behind claims that legal definitions exhaust conceptual reality. Andrew's essay makes valuable contributions to this conversation, particularly in its emphasis on evidence-based discourse and compassionate engagement. But it would be stronger if it acknowledged the genuine complexity of how language operates in systems of power, rather than resolving that complexity through definitional authority.
I'm not arguing that we should restrict speech more broadly, nor that all harmful language should be suppressed. Like Andrew, I believe strongly in protecting democratic discourse. But I think we can do so while being more honest about the nature of language itself. Just as we allow certain forms of physical violence (boxing, police force, military action) for various social purposes while still calling them violent, we might acknowledge that language can be violent while still protecting most speech for the sake of democratic values.
The goal isn't to win a semantic argument but to build frameworks for thinking about language, power, and responsibility that help us navigate an increasingly complex world. That requires engaging with the full spectrum of how speech functions, not just the narrow slice that fits comfortably within existing legal categories.
Andrew, I hope you'll engage with these ideas and help refine them. As I'm sure you'd agree, the highest use of debate isn't to dominate one's opponent but to arrive together at deeper truths through the respectful collision of different perspectives. I welcome your response and the opportunity to continue this essential conversation about how we understand the power and responsibility that comes with human speech.
Anthony David Adams is widely known as the Unicorn Whisperer for his high-stakes performance work with once-in-a-generation talent from the founders of category defining companies like Asana, Bombas and Tuft&Needle to Broadway legends to top brain surgeons to professional champion athletes.
His work has been organically featured in hundreds of media outlets from PBS to The New York Times to Forbes to CNN to TIME, and has impacted millions of lives.
To join the waitlist for his 1:1 coaching or Flight Crew mastermind, be in touch.