When the Dictionary Surrenders: The Ancient Practice of Redefining Words Until They Obey

In 427 BCE, the Greek historian Thucydides observed something that no political scientist has improved upon in the twenty-five centuries since. Writing about the civil war in Corcyra, he noted that during periods of political extremity, words themselves changed their ordinary meanings. Reckless audacity was called loyal courage. Prudent hesitation was called cowardice in disguise. The person who thought before acting was suspected of hidden disloyalty.

Thucydides was not describing propaganda in the modern sense. He was describing something more fundamental: the way that political stress warps the shared vocabulary that communities use to coordinate their understanding of the world. He was describing what happens when the people who control language also control power, and when maintaining that power requires that certain words mean something other than what they have always meant.

Nothing about this has changed. The technology for disseminating redefined words has improved enormously. The human impulse behind the redefinition has not moved at all.

The Mechanism

Political language degrades through a specific and repeatable process. It begins with a word that carries genuine moral weight — liberty, security, justice, democracy, the rule of law. The weight is not arbitrary. These words have accumulated meaning through centuries of use, sacrifice, and contestation. People have died defending what they understood these words to mean. That accumulated weight is precisely what makes them valuable to those who would redirect them.

The redefinition rarely announces itself as such. It proceeds instead through a series of incremental contextual shifts, each of which seems defensible in isolation. Security comes to encompass surveillance that would previously have been called intrusion. Liberty comes to mean the freedom of the powerful from accountability rather than the freedom of the individual from coercion. Justice comes to describe outcomes that benefit particular constituencies rather than procedures applied without regard to identity.

Each shift is argued on its own terms, often by intelligent people making locally coherent arguments. The cumulative effect is that the word, while retaining its original emotional resonance, has been evacuated of its original meaning. It still functions as a rallying cry. It no longer functions as a description.

Rome and the Elasticity of Res Publica

The Romans built their entire political identity around the concept of the res publica — the public thing, the common concern, the shared enterprise of self-governance. For centuries, the phrase carried genuine descriptive content. It meant something specific about the distribution of power, the role of the Senate, the limits on individual authority.

By the late Republic, res publica had become an almost perfectly elastic term. Julius Caesar claimed to be restoring it when he crossed the Rubicon with his army. His assassins claimed to be defending it when they killed him. Augustus, who systematically dismantled every institutional feature of the Republic while preserving its ceremonial vocabulary, used the phrase constantly. He called his autocratic consolidation of power the restoration of the Republic. He kept the Senate. He kept the consulship. He kept the elections. He simply ensured that none of these institutions retained any of the power the words implied.

The genius of the Augustan settlement was not military. It was semantic. By retaining the language of the Republic while gutting its substance, Augustus allowed Romans to avoid the psychological disruption of acknowledging what had actually happened. The word res publica continued to do the emotional work it had always done. It simply no longer did the descriptive work.

The Constitutional Version

American political discourse offers a particularly rich field for studying this phenomenon, because the United States has a written constitutional text that both constrains and invites semantic manipulation.

The constraint is obvious: the document exists, can be read, and says specific things. The invitation is less obvious but equally real: the document uses words whose meaning has never been permanently fixed, and whose interpretation has always been contested. Commerce, necessary and proper, general welfare, due process, equal protection — each of these phrases has supported interpretations across a range that the document's authors would in some cases have found unrecognizable.

This is not inherently pathological. Language evolves, circumstances change, and a constitution that could not accommodate any interpretive development would be a constitution that could not survive. The pathology emerges when the interpretive process is driven not by honest engagement with the text and its purposes but by the need to reach a predetermined conclusion.

The tell is usually consistency. An interpreter who applies a particular reading of executive power only when their preferred party holds the executive is not engaged in constitutional interpretation. They are engaged in the use of constitutional language as a post-hoc justification for a position they reached by other means. The word constitutional in their usage no longer means consistent with the Constitution. It means useful to us right now.

This pattern has appeared across the entire ideological spectrum, across the entire history of the Republic. It is not a partisan phenomenon. It is a human one.

The Acceleration Problem

What the digital information environment has changed is not the underlying impulse but the speed and scale at which semantic redefinition can propagate.

In previous eras, the redefinition of political language was constrained by the friction of communication. A new usage of an old word had to travel through newspapers, pamphlets, speeches, and word of mouth. The process took time, and that time created opportunities for resistance — for people who remembered the older meaning to articulate the difference, to contest the shift, to slow the evacuation.

Contemporary information infrastructure has largely eliminated that friction. A redefined usage can reach tens of millions of people within hours, repeated across enough channels and by enough voices that it begins to feel like established meaning before any organized resistance can form. The word fake was redefined in American political discourse with a speed that would have been technically impossible twenty years earlier. So was election integrity. So, in different directions, were equity, woke, and populism.

This is not to say that any of these words has only one legitimate meaning, or that contestation over their meaning is inherently corrupt. Contestation over the meaning of important words is a normal feature of political life. What is worth noting is the acceleration — the degree to which the pace of semantic change now outstrips the capacity of public discourse to process it deliberately.

The Diagnostic Question

Thucydides' observation implies a diagnostic question that is still useful twenty-five centuries later: when a political actor uses a word of significant moral weight, ask whether they would apply the same word consistently if the beneficiary changed.

Would the person invoking liberty to oppose government mandates apply the same word to oppose government restrictions on other kinds of individual choice? Would the person invoking the rule of law to demand prosecution of their opponents apply the same standard to allies? Would the person citing precedent as sacred when it supports their position treat it as equally sacred when it does not?

Inconsistency does not automatically indicate bad faith. Reasonable people hold inconsistent positions without recognizing the inconsistency. But systematic inconsistency — the pattern of applying a word's full moral weight selectively, based on whose interests it serves — is the clearest signal that the word has been converted from a description into a weapon.

The weapon is ancient. The words change. The technique does not. Five thousand years of political history is, among other things, a very long record of people discovering that the most durable power is not the power to compel action but the power to define what the action means.