How to write a personal principle about trust (deciding who has earned it back)

Trust is an unusual value to write a principle about, because the moment that actually tests it is not a daily one — it is the recurring, specific decision of whether to let someone back in after they have already shown you the version of them who let you down. "I'm a trusting person" and "I don't trust easily" are both self-images, not rules, and neither one tells you what to actually do the next time someone who hurt you is standing in front of you with a good excuse and a request for the access they used to have. A personal principle about trust has to hold in exactly that moment, not in the calm months on either side of it.

Why "I trust people who earn it" cannot be violated

Ask someone whether they are a trusting person and most will answer instantly, correctly by their own definition, whichever answer they already believed about themselves before you asked. That is the tell. A rule you can affirm without it ever costing you anything is not a rule — it is an identity, and identities are cheap to hold because no single day can disprove them. You can hand someone back exactly the access they had a week after they broke something real, and still think of yourself as someone who "does not trust easily," because the label was never actually pinned to a moment you could be shown to have gotten wrong.

The same self-image problem, running the other direction, produces people who call themselves careful or guarded, and use that language to avoid a less flattering one: still hurt, still keeping score, unwilling to update no matter what the other person actually does next. A principle that never specifies what would move the needle either way is not protecting you from anything. It is just delaying a decision you are going to make anyway, badly, at 11pm during an argument that is not really about trust at all.

Mine your own record of what actually changed your mind

Do not start by asking "who do I trust." Ask instead:

• The last time someone let me down in a way that mattered, what did they actually do afterward — and did it match what I said I needed to see, or did I just get tired of being upset with them?
• Is there someone I have quietly given back the same access they had before, without anything actually being different, because keeping the distance felt like more work than closing it?
• Is there someone I still keep at the old distance long after they did the specific thing I said I needed — and what am I actually protecting by not noticing that?
• When I decide to trust someone again, is it usually because of something they did, or because of how uncomfortable the waiting period had become?

Write the specific instance, not the tidy self-assessment. "I gave him the spare key back because the silence between us was awkward at family things, not because anything had changed" tells you exactly where a trust principle needs to hold. "I trust people who earn it" is a sentence anyone can sign without it costing anything, which is exactly why it will not help you the next time someone is asking for the old access back before they have actually shown you anything new.

Write the rule about the decision, not a verdict on the person

"I trust people who earn it" cannot be checked against any single event — it names a category of person, not an action you take. "When someone has broken something specific, I wait until I have watched them do the corrected version, unprompted, more than once, before I hand back what I took away — not because enough time has passed, and not because staying upset is exhausting" can be checked. There is a moment, a request on the table, and a next day on which someone could say honestly whether you gave the access back because of evidence or because of fatigue.

Run the same two checks that apply to any principle. Can it be violated — is there a specific, visible instance where you can say whether you kept it or broke it? And does it cost something — a rule that only ever agrees with what you were already inclined to do, whichever direction that runs, is not doing any work.

Attach behavior standards someone else could actually notice

Keep the principle general and put the specifics underneath it, where they can be revised as a particular relationship changes. For a principle like I restore access based on what I have watched, not on how long it has been, useful standards might be: I name the specific thing I need to see before I say "we're fine," so it is a fact I can check rather than a feeling that quietly wore off; I do not hand back a key, a confidence, or a responsibility until I have actually seen the corrected behavior at least once, not just heard it promised; if I catch myself closing the distance because the tension is exhausting rather than because something changed, I say that to myself before I act on it. Each of these is small enough to actually do and specific enough that you, or the other person, could say by the end of the conversation whether it happened.

Where earning it back is not the same as never trusting again

A principle about trust is easy to misread as license to keep someone on permanent probation, and that is a different failure running in the opposite direction. Requiring evidence is not the same as requiring it forever — a standard with no finish line is not caution, it is a longer punishment with better manners, and the person on the other end of it eventually stops trying, correctly, because nothing they do will ever be enough. The standards above need an actual endpoint: a number of instances, or a length of time attached to a specific behavior, not an open-ended wait for a feeling of safety that may never arrive on its own. Once that bar is met, the rule also obligates you to close it out in fact, not just in the words you use — a "we're past that" that comes with the old distance quietly left in place is not trust restored, it is trust withheld with a kinder sentence attached.

Stress-test it against the case you are actually deciding right now

Take the real person and the real request currently in front of you — not a hypothetical one — and run the principle against it directly. If the rule tells you something different from what you were about to do, either give in early because the tension is uncomfortable, or refuse indefinitely because you have not defined what would ever be enough, it has teeth; keep it as written. If it agrees with whatever you were already leaning toward, the principle is not adding anything, and it is worth rewriting until it can actually disagree with you. This is also where you find out whether the rule is genuinely yours or borrowed from someone whose history with trust looks nothing like yours — a rule you cannot picture holding at the exact moment someone is asking for the old access back will not survive being in that moment.

This is close to the shape a creed inside CreedOS is built to hold: a stated principle with two or three behavior standards attached underneath it, checked in a daily check-in that simply records what happened rather than grading it. 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 person you are actually deciding about. It is free, with no in-app purchases.