CreedOS

Home / Guides

How to write a personal principle about empathy you can actually check

Empathy is not a behaviour, and that is exactly why "be more empathetic" fails as a principle before you have even finished writing it. It names a capacity — the ability to register that someone else's experience is real and different from your own prediction of it — and a capacity is not something you do at 7pm on a Tuesday. What you can govern is the moment that capacity is tested: the point where you are already forming a response, and you either let the other person's actual account revise it or you don't. That moment is specific, it happens on a schedule your week already sets for you, and it is the only part of this a written rule can actually reach.

Why the vague version cannot be kept or broken

Ask yourself whether "I want to be more empathetic" has a moment where it is tested. It doesn't — you can sincerely hold it as a value while still interrupting a friend's explanation with your own conclusion about what they must be feeling, because the sentence never named the interruption as the thing to watch for. A principle needs an instant where someone, including you, could point at what just happened and say whether you kept it or broke it. "Empathy" describes an outcome you would like to be true of you in general; it does not tell you what to do the moment a coworker starts explaining why they missed the deadline and you feel the urge to explain, correct, or defend before they finish.

Find where your own version of it actually fails

Empathy rarely goes missing across the board — it goes missing in specific, repeatable conditions, and those are what a principle should target. Common ones: when you are being criticized and the criticism has any truth to it, when you are rushed and someone's explanation feels like it is delaying the actual point, when the other person's account conflicts with a story you have already decided is true, or when you are tired enough that listening starts to feel like a cost you would rather not pay. Look at the last conversation that went badly — not a stranger's, one with someone you actually care about — and find the exact sentence where you stopped taking in what they were saying and started composing your reply instead. That gap, between their sentence ending and your attention actually being on it, is where the rule needs to live.

Write the rule about the pause, not the feeling

The useful target is not "understand people better," which cannot be observed, but the specific action that empathy either produces or fails to produce: whether you ask a question before you respond with your own view, and whether the other person's answer is allowed to change what you say next. "When someone tells me I hurt or upset them, I ask one question about their experience before I explain mine" is checkable — there is a sentence someone could review, a moment it either happened in or didn't, and a version of you that skipped straight to explaining that anyone in the room could point to. Run the same two tests that apply to any principle: can it be violated — is there a real, visible instance where you asked and one where you didn't — and does it cost something, meaning it occasionally makes you sit with an account of yourself you would rather not hear.

Attach standards someone else could actually notice

For a principle like I let the other person finish their account before I give mine, workable standards might be: when someone raises a problem with something I did, I restate what I heard before I respond to it, so they can correct me if I got it wrong; I do not use the words "but" or "actually" as the first word of my reply to someone describing how they feel; if I notice I am forming my rebuttal while the other person is still talking, I stop and finish listening before I say anything. None of these require you to feel differently first, and none of them ask you to agree with the other person — only to let their account land before your own takes over. All three can be confirmed or denied by the person on the other side of the conversation, which is the difference between a principle and a private opinion of yourself that nobody else can check.

Where this differs from agreeing, and from being talked out of your own view

Empathy governed this way is not the same as capitulation, and a rule that quietly turns into "give in to whoever is upset" is not empathy, it is conflict avoidance wearing a better name — worth checking against a principle about people-pleasing if that pattern sounds familiar. Restating what you heard and asking a question before you respond does not commit you to agreeing with it; it commits you to responding to what was actually said instead of to the version of it you had already assembled before they finished the sentence. You can hear someone out completely and still hold your original position — the rule only governs whether their actual account was allowed to reach you before you decided that.

Stress-test it against the last conversation that went badly

Take that conversation again — the one where you found the exact sentence you stopped listening at — and run the rule against it directly. Would asking one question first, or restating what you heard, have changed what you said next? If it would have, the rule has teeth and you keep it as written. If it would not have changed anything, the rule is aimed at the wrong moment, and the honest move is to rewrite it rather than drop it, because the pattern you found — interrupting, correcting, defending before the other person finishes — is still real even if the first sentence you wrote for it did not quite name where it starts.

Social is one of the six areas CreedOS ships templates for, alongside efficiency, emotion, money, health, and cognition, each with curated creeds and behavior standards already written that you can take as they are or edit until they name the exact interruption your own pattern keeps making. An AI coach can also help think through what the standard should be once you have named the moment. It is free, with no in-app purchases.