There's a test I keep coming back to, and I'll warn you now that it's unpleasant to run on yourself. Take a principle you're sure you hold. Free speech, fairness, honesty, whichever one sits closest to your identity. Now go find the situation where that principle would have cost you or your side something, and check what you actually did.
Most of us fail, and not because we're liars. The mind is very good at reaching for whichever principle happens to justify what we already wanted, and then quietly putting it back on the shelf when it points the other way. From the inside this never feels like hypocrisy. It feels like nuance. That's what makes it so hard to catch.
Framing does most of the work
The cleanest demonstration I know is over forty years old. In 1981, Kahneman and Tversky asked people to choose between programs to fight a disease expected to kill six hundred people. Describe one option as saving two hundred lives and most people prefer it to a gamble. Describe the mathematically identical option as four hundred people dying, and preferences flip toward the gamble. Same numbers, same stakes, different wording. Nobody in those experiments felt manipulated. They just answered the question in front of them.
That result matters for ethics more than it first appears to, because if a change of wording can flip your answer to the same question, your answer wasn't really yours. Whoever picked the wording picked it for you.
We build an ethics assessment at Agonora, and one of the things it measures is exactly this. We put the same dilemma in front of people twice, reframed and spaced apart, and check whether the two answers are compatible. I went in expecting a certain amount of drift. Watching how often the frame, rather than the values, decides the answer has still been humbling, and I'm not excluding my own practice runs from that.
The Emerson objection
The obvious pushback is Emerson's line about 'a foolish consistency' being 'the hobgoblin of little minds', and he was right about the thing he was attacking. Refusing to update when the evidence turns isn't a virtue; it's a different vice. But I think there's a distinction that dissolves most of the objection: consistency of principle versus consistency of position. A scientist who abandons a theory when the data breaks it is being consistent, because the underlying commitment, follow the evidence, never moved. The person who defends the theory forever isn't consistent. They're just stuck.
So none of this is an argument against changing your mind. It's an argument for knowing which layer moved when you did.
Where it shows up
The classic case is the free speech advocate, passionate right up until the speech in question is something they find repellent, at which point a remarkably sophisticated set of exceptions appears. I've watched this run in both political directions, occasionally in the same person within a single year. The principle was never speech. It was protection for speech they liked, which isn't a principle at all, just a preference in a suit.
Or, smaller: the friend for whom your lateness is a character flaw and their lateness is traffic.
Most of us are that friend somewhere. I certainly haven't located the corner of my life where I'm not.
Why I think the stakes are rising
Here's the claim I'd normally hedge into oblivion, so let me state it plainly and flag the uncertainty afterwards. Machines can now generate a tailored argument for any position, adjusted in real time to whatever seems to move you. I don't know how fast this scales or what forms it ends up taking; predictions in this space have a short shelf life, and mine are no exception. But the direction seems hard to argue with. The volume and precision of persuasion aimed at each of us is going up, a lot, and the cost of producing it is going to nearly nothing.
If your principles bend to framing, you are steerable, and the systems doing the steering are getting very good at framing. The unsettling detail in the framing research was always that being moved doesn't feel like being moved. Each new position feels like your own reasoning. There's no inner alarm that rings.
I should concede the strongest counterpoint, because it's a real one: the drive toward consistency is itself a manipulation lever. Robert Cialdini documented this decades ago in Influence. Get someone to make a small commitment and their need to look consistent will walk them into larger ones they'd never have accepted cold; every salesperson trained since has used it. So 'but you said' can be wielded against you as effectively as any reframing. The distinction I'd draw, and I admit it's easier to state than to live, is between being consistent and needing to appear consistent. The first is loyalty to a principle. The second is loyalty to an image of yourself, and images are easy to take hostage.
What actually helps, in my experience, is embarrassingly low-tech: decide the principle before the situation arrives, ideally in writing, while you don't yet know which side of it you'll be standing on. In the moment you will be brilliant at finding the exception; in advance you're merely honest. Ulysses tying himself to the mast is the old version of this, and the reason it worked is that he did it before he could hear the singing. And yes, I'm aware that an essay praising consistency is an open invitation to go check mine. That's fair. It's rather the point.