Consistency as a Minimum Bar

Scenario

Arjun tells his book club, on a Tuesday, that lying is wrong without exception, and that people who carve out excuses for it are only excusing themselves. It is a tidy, principled thing to say, and nobody in the room objects. The following week, over dinner, he mentions that his mother called during a bad week at work, and that he told her everything was fine because he did not want to add his stress to hers. His friend points out that this was a lie. Arjun's answer is that of course it was, but it was obviously the right call, she did not need that weight on top of her own.

Neither conversation sounds strange by itself. Said to a room, "lying is wrong without exception" reads as a firm ethical line. Said to a friend, "I lied to spare someone" reads as an act of care. The problem shows up only when the two are placed side by side. The claim from the book club rules out every case of lying, including the one Arjun describes at dinner. The claim from dinner treats that very case as justified. Both cannot be true. If lying really is wrong without exception, the lie to his mother was wrong regardless of his reasons for telling it. If sparing her was the right call, then "lying is wrong without exception" was never actually the rule he was living by, whatever he said on Tuesday.

Definition

Call a claim anything a person asserts that could turn out true or false. A set of claims is inconsistent when there is no way the world could be that makes every claim in the set true at the same time. Arjun's two claims form a set of exactly two: "no lie is ever the right thing to do" and "this lie was the right thing to do." Try to build a version of the world where both hold. There is not one, because the second claim is an instance of the very thing the first claim forbids. That is what inconsistency means: no possible arrangement of the facts satisfies both claims together.

This is a stronger condition than two claims merely sitting awkwardly next to each other. Apparent tension often dissolves under a closer look: one claim might quietly be about a different case, or carry an exception the speaker simply failed to say aloud. "I value honesty" and "I sometimes withhold information from people" can both be true, since withholding is not the same act as lying, and a person can hold both without either one denying the other. A genuine inconsistency survives that check. State both of Arjun's claims as precisely as he would defend them, and they still rule each other out.

This gives the shape of the principle of non-contradiction, one of the older results in logic. Aristotle stated it as a constraint on how the world is: nothing can both have a property and lack it, in the same respect, at the same time. The same constraint applies to a set of claims a person asserts: a claim and its exact denial cannot both be true. Arjun's book club claim, "no lie is ever the right thing to do," denies exactly the case his dinner story affirms, "this lie was the right thing to do." Once the case is the same and the respect is the same, one claim affirms what the other denies, and the principle says a person cannot hold both as true.

Consistency and truth are different tests, and the difference runs in only one direction. A set of claims can be entirely consistent and every member false: someone could hold "the coin will land heads," "heads-landing coins bring good luck," and "luck is a physical force," and those three would never contradict each other while still managing to be wrong about all three. Consistency alone does not certify correctness. The other direction is where the bar earns its name: an inconsistent set guarantees, without needing any further evidence, that at least one of its members is false, because if every member were true, the set would already be a consistent description of the world, and by hypothesis it is not. Consistency does not say which claim is the faulty one, only that the set as a whole cannot be entirely true. Passing the test proves nothing has been ruled out; failing it proves an error is sitting somewhere in the set.

Common mistakes

Treating consistency as proof of truth. Arjun could revise his rule to "lying is wrong except to spare someone unnecessary pain," and the two conversations would stop conflicting. That revision buys consistency, and it does nothing to establish that the revised rule is correct; the new version could still be indefensible for reasons that have no bearing on whether it contradicts itself.

Calling every tension a contradiction. Two claims can look like they clash before either one is checked against the other precisely, and treating that discomfort as a confirmed inconsistency skips the actual test: does accepting one claim require rejecting the other in every possible circumstance. If it does not, the pair is merely uncomfortable, not inconsistent, and forcing a resolution onto it solves a problem that was not there.

Resolving a real inconsistency by force rather than by revision. Confronted with the gap between his two claims, Arjun could repeat the book club line louder, or dispute his friend's memory of the dinner conversation. Neither move touches the inconsistency; both avoid the question of which claim has to give. Resolving a genuine inconsistency requires dropping or narrowing one of the claims that produces it. Repeating either claim with more conviction changes nothing about whether the pair can both hold.

Try it yourself

Pick one topic you hold confident opinions about, work, money, a relationship, whatever you argue for most readily, and write down four or five things you actually believe about it, stated as flatly as you would say them out loud. Then look for two of them that cannot both be true, where accepting one requires rejecting the other in every circumstance you can imagine, the way Arjun's rule about lying and his account of the dinner conversation could not both hold once placed side by side. A pair that merely feels uncomfortable together does not meet that bar; keep looking until you find one where affirming one claim genuinely denies the other.

Once you have a real pair, decide which of the two claims you are willing to give up or narrow, and write the narrowed version down. Success on this task means three things end up on the page: the two original claims in your own words, a one-line demonstration that no possible situation makes both of them true at once, and the revision you are willing to make to restore consistency. If four or five claims are not enough to surface a genuine pair on the first pass, add a few more from the same topic; a person's confident opinions rarely stay arranged so carefully that none of them collide.