In August 1998, testifying before a grand jury, President Bill Clinton offered a sentence that became a national punchline: “It depends upon what the meaning of the word ‘is’ is.” Critics mocked it as evasion. In rhetorical terms it was something more precise. It acknowledged that the question before him turned less on what had happened than on what to call it.

Clinton was doing what advocates have done since antiquity. Classical rhetoricians called this the stasis of definition: the point in a dispute where both parties accept the facts but disagree about how to name them. Hermagoras, Cicero, and Quintilian taught that this stage often decides a case before a single piece of evidence is weighed.

Most of us believe our arguments are about facts or values. Very often, we are actually fighting over a word, and whoever controls the definition frequently controls the outcome. Learning to recognize these moves, and to contest them openly rather than resentfully, is one of the most practical skills rhetoric has to offer. It begins with seeing that definitions are rarely neutral.

A stipulative definition announces how a term will be used: “For the purposes of this discussion, by ‘success’ I mean…” In honest inquiry this is a virtue. It replaces vagueness with precision, which is why statutes open with definitions sections and philosophers begin by fixing their terms. The trouble starts when a stipulation is presented as neutral housekeeping while it quietly settles the substantive question in advance.

Consider a debate over whether a policy “reduced poverty.” If one side measures poverty by pre-tax cash income, benefits like food assistance or tax credits may not register at all. If the other side counts them, the same policy looks transformative. Neither figure is fabricated. The verdict was largely decided at the moment someone chose what to count, and the audience may never notice that a choice was made.

These moves have a recognizable signature. The definition arrives early, often casually, and asks for agreement before its consequences are visible: “We can all agree that X simply means Y.” Classical manuals treated the brief, favourable definition as a figure of thought in its own right, a way of framing the matter so that conclusions seem to follow naturally. The technique has long outlived the manuals that described it.

The strategic response is neither to accept nor to reject the definition wholesale. Ask what it includes and what it excludes. Test it against cases it handles poorly. Then propose your alternative openly and argue about which definition better serves the purpose at hand. This reflects Wayne Booth’s ethical ideal of rhetoric: the goal is not to capture the word but to make the choice visible, so the audience can judge it for themselves.

Takeaway

When someone asks you to agree on a definition, you are being asked to agree on a conclusion in disguise. Pause and ask what the definition makes impossible to see.

Categories carry baggage: rules, penalties, protections, and attitudes. To place a case inside or outside a category is to decide which bundle applies. Cicero, in De Inventione, offers a classic example. A man steals private money from a temple. Is he a thief, or guilty of sacrilege? No one disputes what he did. Everything depends on the label, because the label determines the punishment.

The same logic reached the United States Supreme Court in 1893. The Tariff Act of 1883 taxed imported vegetables but not fruit, and the Nix family, importers of tomatoes, argued that tomatoes are botanically fruit. In Nix v. Hedden, the Court conceded the botany but ruled that in ordinary language, the tomato served with dinner rather than as dessert is a vegetable. The duty applied. Both sides held legitimate definitions; the real question was which context governed.

Borderline cases are where boundary arguments live. Is a gig worker an employee or a contractor? Is a disruptive demonstration a protest or a riot? Is a misleading statement a lie? Each label triggers different consequences: labour protections, police responses, moral condemnation. Advocates push borderline cases toward the category whose consequences they favour, usually by stressing resemblance to a paradigm case, the clearest example everyone already accepts.

To respond strategically, identify the paradigm case each side is invoking. Then ask what features of that paradigm justify the category’s rules, and whether the borderline case shares them. The more productive question is rarely “Is it X?” but “Do the reasons we treat X this way apply here?” This moves the debate from labels to purposes. Sometimes the honest answer is that neither category fits, and a new one is needed.

Takeaway

A category is a bundle of consequences wearing a name. Argue about whether the reasons behind the rules apply, not merely whether the label fits.

“That’s not real art.” “He’s not a true conservative.” “This isn’t genuine democracy.” Appeals to a word’s true or real meaning sound like claims about language, but they are frequently claims about value. In 1938 the philosopher Charles Stevenson named this pattern the persuasive definition: a move that keeps a word’s favourable emotional charge while quietly changing what it refers to. Redefine “true freedom” as what you prefer, and your preference inherits freedom’s prestige.

In 1917 Marcel Duchamp submitted a porcelain urinal, signed “R. Mutt,” to an exhibition in New York. It was not displayed, and the question of whether Fountain was art has been argued ever since. But the argument was never really lexical. It concerned what art should do, whether craft matters, and who gets to decide. No dictionary could settle it, because everyone involved wanted “art” to carry its honour.

Antony Flew gave the most famous symptom of this pattern a name: No True Scotsman. When a counterexample appears, the speaker redefines the category to exclude it. No true Scotsman would do such a thing; this man did it, so he is no true Scotsman. The term becomes immune to evidence because its boundaries stretch to protect it. A reliable warning sign is that no conceivable case could ever count against the essence being claimed.

The remedy is translation. Convert “Is X really Y?” into “What do you value about Y, and does X honour or threaten it?” This is close to what Booth called listening-rhetoric: seeking to understand what the other person actually cares about before trying to change their mind. Translated this way, many essence disputes reveal agreement on facts and disagreement on priorities, which is a harder conversation but a far more honest and tractable one.

Takeaway

When someone insists on what a word really means, ask what they are trying to protect. The definition is the surface; the value is the argument.

Definitions are not preliminaries to argument. Very often, they are the argument. Stipulations, category boundaries, and essence claims can each decide outcomes while appearing to be neutral matters of vocabulary.

The practical guidance follows directly. When a debate stalls, ask whether you are really disagreeing about a word. Make your own definitions explicit and defend them by the purpose they serve. When contesting someone else’s, test the edge cases and surface the values underneath rather than accusing them of bad faith.

The ethical persuader does not hide the definitional choice but places it on the table where the audience can judge it. That transparency may cost a little advantage. It earns something more durable: trust, the oldest and strongest form of ethos.