The complete script of this episode, about 1,891 words — roughly 13 minutes to read. Play the episode above and the transcript follows along; tap any paragraph to jump the audio there.
0:00What gaslighting actually takes from you
This is an episode about the part of recovery almost nobody prepares you for: what happens after you have named the gaslighting, after you have left or gone no contact, after the story is finally straight in your head — and you still cannot trust your own judgment about anything. You second-guess which restaurant to pick. You draft a two-line text four times. Someone asks what you want and your mind goes completely blank. That is not indecision, and it is not weakness. That is what happens when your instrument for reading reality has been deliberately knocked out of calibration, and it is the thing we are going to rebuild.
Here is what gaslighting actually takes. It does not take your memory. Your memory works fine. What it takes is your permission to rely on your memory. Every time you said, that is not what happened, and were told you were confused, tired, dramatic, remembering it wrong, you learned a small lesson: my read is unsafe to act on. Repeat that a few hundred times and you build a reflex — check with them first, wait for the correction, do not commit to your own version. The reflex outlives the relationship, because the reflex is what kept you safe inside it.
So the goal is not to become certain again. Certainty was never the thing. Plenty of people with excellent judgment are uncertain constantly. The goal is to be able to act on your best read without needing someone else to sign off on it. That is what self-trust is: not knowing you are right, but being willing to move on your own assessment and correct later if you were wrong. We are rebuilding a working relationship with your own perception, not installing a new authority.
One more thing before we start, because it matters. If you are recently out, some of what feels like broken judgment is simply exhaustion. Chronic vigilance burns the same fuel decisions run on. Sleep, food and time will hand back a surprising amount of clarity on their own. What follows is for the part that does not come back with rest.
2:25The self-doubt loop, and how it runs
Let me show you the loop, because you cannot interrupt something you cannot see. It runs in four moves and it runs fast, usually in under ten seconds. Move one: you have a perception. Something felt off in that conversation. Move two: the old rule fires — my perception is probably distorted. Move three: you go looking for outside confirmation. You reread the messages, you ask a friend, you rehearse the whole thing to see if it still holds. Move four: whatever you find, the loop restarts, because the search itself proved to you that your read needed a chaperone.
Notice what the loop never includes: acting. That is the tell. Healthy uncertainty ends in a decision made with incomplete information. The doubt loop ends in more searching. If you are gathering evidence with no threshold at which you would stop gathering, you are not investigating — you are stalling, and the stalling is the injury talking.
The second thing to notice is who the loop hands authority to. After gaslighting, it is almost never handed back to you. It goes to a friend, a group chat, a therapist, a comment section, or to the very person whose contradictions caused the problem. That is why hoovering works so well on people in this state. When you cannot verify your own reality, someone offering a confident version of it is offering relief, and relief is very hard to refuse.
So the loop has a shape, and the shape has a weak point: move three. You cannot delete a perception, and you cannot argue away a reflex. But you can change what you do in the ten seconds after the reflex fires. Everything practical in this episode lives in that ten-second window.
4:23Building the evidence base
The first practice is boring and it works better than anything else I know: keep a record, and then read it back. Not a journal about your feelings — a log of your reads. One line, written before you know the outcome. Something like: I think this meeting will get moved again. I think she is annoyed and will not say so. I think I will regret saying yes to this. Date it. Do not edit it later. Then, once a week, go back and mark which reads were right, which were wrong, and which are still open.
What this does is replace an argument with data. Right now, when you ask, can I trust my judgment, you are asking a question you answer with a feeling — and the feeling was installed by someone with a motive. After six weeks of logged reads, you are answering with a hit rate. Almost everyone who does this discovers the same uncomfortable, wonderful thing: their reads were mostly accurate the whole time, including inside the relationship. That is often the moment the reflex actually starts to loosen, because you cannot argue with your own handwriting.
Keep the log small enough to survive. Three lines a week beats a beautiful system you abandon on day nine. And include the ordinary ones, not just the emotionally loaded ones, because ordinary predictions are where your accuracy is easiest to see and hardest to explain away.
The second practice is separating observation from interpretation, out loud. Observation: he said he would call at seven and called at nine without mentioning it. Interpretation: he does not respect my time. Both may be true. But gaslighting attacks the observation layer — you did not say seven, you are making that up — and once the observation is in question, everything built on it collapses. Get in the habit of stating the observation first and holding it separately. Observations are checkable. Interpretations can stay provisional without taking the ground out from under you.
6:38Small decisions, made on purpose
Now the muscle work. Self-trust is not restored by insight; it is restored by repetition of a specific act — deciding, and then living with the decision. So we practice on stakes low enough that being wrong costs nothing.
Here is the drill. Pick decisions that are genuinely reversible and genuinely yours: what to eat, which route to take, which of two shirts, whether to answer the phone right now. Give yourself a hard limit — thirty seconds — and then commit. No polling, no rereading, no checking. Then, and this is the part people skip, notice what happened afterward. Almost always: nothing. The world did not correct you. Nobody arrived to explain that you had got it wrong. That absence of consequence is the medicine. You are teaching a nervous system that learned decisions are dangerous that most decisions are simply decisions.
When you do get one wrong — you will, everyone does — practice the repair without the collapse. The healthy version has three parts: name it plainly, fix what can be fixed, and keep your authority. That last part sounds like, that was the wrong call, I will do it differently, rather than, see, this is why I cannot trust myself. One wrong read is information about that read. It is not a verdict on your instrument. People who gaslight teach the opposite lesson, that any error proves total unreliability, and that lesson has to be actively unlearned each time you slip up.
Scale slowly. After a few weeks of thirty-second decisions, move to decisions with a small social cost — saying no to an invitation you do not want, ordering first at the table, telling someone your actual preference when they ask. Those are the ones that were punished, so those are the ones that rebuild the most.
8:41Rebuilding with other people
Self-trust does not get rebuilt in isolation, because the injury happened in relationship, and relationships are where the reflex fires hardest. But who you practice with matters enormously.
Look for people who can hold two things at once: they can disagree with your conclusion while still treating your perception as real. That sounds like, I see it differently, and I believe that is how it landed for you. Compare that with the version you are used to: that is not what happened. The first leaves your instrument intact. The second removes it. You are not looking for people who always agree with you — agreement is not the same as safety, and a friend who only ever validates you is training the same dependency from the other direction.
Be careful with over-consulting. Asking one trusted person for a second opinion is good practice. Asking six people the same question is the doubt loop wearing a social costume, and it usually ends with you outsourcing the decision to whoever sounded most confident. A useful rule: consult before you decide, never as a substitute for deciding, and cap it at one or two people.
Watch for the contact test, too. If the person who did the gaslighting is still reachable, every wave of self-doubt will present reaching out as a reasonable way to settle the question. It never settles it. The information you would get is generated by the same source that broke the calibration. If you need a script for this, the boundary scripts on the site are written for exactly that moment, and using a script is not a failure of nerve — it is what keeps a decision you already made from being renegotiated in real time.
10:38What recovered self-trust actually feels like
Let me set the expectation properly, because the wrong expectation makes people quit at the six-week mark. Recovered self-trust does not feel like confidence. It does not feel like certainty. It feels, mostly, like speed — the gap between having a read and acting on it gets shorter. It feels like fewer drafts. It feels like noticing something is off and simply adjusting, without the internal trial about whether you are allowed to have noticed.
The doubt does not vanish either. What changes is its status. Right now the doubt is a verdict; later it is just weather. You will still get the flicker of maybe I am imagining this, and you will register it and keep going. That is the finished state. Not silence — permission.
You will also notice something quieter: your standards come back. When you cannot trust your read, you cannot enforce a boundary, because every boundary rests on a judgment that something is not okay. As the calibration returns, so does the ability to say, no, that does not work for me, without building a legal case first. A lot of people find their boundaries return before their confidence does, and that order is fine. Acting on your judgment is what rebuilds the feeling, not the other way around.
So, the practice, in short. Log your reads and check the hit rate. Separate observation from interpretation. Decide small things in thirty seconds and watch nothing bad happen. Repair errors without surrendering authority. Practice with people who can disagree without deleting your perception. Cap your consulting. And measure progress by how fast you act, not by how sure you feel. If you want to keep going after this one, the gaslighting field guide, the boundaries guide and the recovery pages are all linked in the show notes — and you can register for the monthly digest to get each new episode and article as it lands.
