5.3 — Finding Your Triggers, and Defusing Them

Most people, asked to list what sets them off, produce three or four items and get at least two of them wrong.

Not through dishonesty. Through 2.4 and 4.5: you are reconstructing from the most memorable episodes, which are the biggest ones, which are not the most frequent ones. The trigger that fires twice a week at low intensity and quietly costs you your patience with your family is invisible to memory and obvious in a log.

So this page has two halves. Finding them, which takes three weeks of five minutes a day. And weakening them, which takes months and actually works.

Part one: the trigger log

Three weeks. One line per episode. Six fields. Keep it in a note on your phone so it can be filled in within a minute of the event, because filling it in that evening produces a reconstruction rather than a record.

Date and time. The time is not decoration — patterns by hour are among the most common findings.

What happened. Only what a camera would have recorded. "He said the numbers looked off." Not "he had a go at me."

Intensity, one to ten. Rate it in the moment. This is the field that reveals the low-level frequent ones.

What I felt. One specific word (4.3).

What I did. Also camera-level. "Said nothing, was short with everyone for an hour."

State. Sleep last night, hours since eating, how loaded the day already was.

Two rules. Log the small ones — a three is data, and the threes are where the pattern lives. And do not analyse while logging; the analysis happens at the end of the three weeks, because analysing as you go turns the log into an argument with yourself.

Reading the log

After three weeks, sit with it for half an hour and answer six questions in writing.

Which situations repeat? Group them. Most people find three or four clusters covering the large majority of entries.

What is the common threat underneath each cluster? From 5.2: not competent, not respected, not safe, not wanted, being controlled, treated as less than. Two clusters that look unrelated on the surface frequently turn out to share one.

Which people appear most? Not because they are the problem, though sometimes they are. Because it tells you where the work is.

What time of day? Almost everyone has one. Late afternoon and the first hour at home are the two commonest.

What was the state on the highest-intensity entries? Look specifically at sleep and accumulated load. The correlation is usually visible without any statistics.

And what did I actually do, in each case? Your default move under threat — withdraw, attack, appease, over-explain, take control, go silent, make a joke — will be extremely consistent. Naming it is one of the most useful things this exercise produces, because from then on you can catch the move starting.

What people typically find, and it is worth knowing in advance so you do not mistake it for failure: two or three real triggers rather than ten, one dominant underlying threat, a strong time-of-day effect, and at least one trigger they would not have listed and one they would have listed that barely appeared.

Part two: the four ways a trigger gets weaker

Now the useful half. These are ordered by how much they actually do, not by how they feel.

1. Remove the fuel

The cheapest and most neglected. The threshold moves with your state (5.2), so a large fraction of trigger episodes can be eliminated without touching the trigger at all.

Sleep in the right range. Eat before the meeting rather than after. Do not schedule the difficult conversation for 6 pm. Put fifteen minutes between work and home rather than walking straight from one into the other. Handle the noise, the heat, the pain.

This is unglamorous and it is where the largest single reduction usually comes from. People want the psychological work and skip the physical floor, then wonder why the psychological work is so hard.

2. Build the gap

You cannot prevent the surge. You can extend the time between the surge and the action, and that is where every other option lives.

The physical interrupt. One thing, always the same, practised until automatic. Two breaths with a long exhale (4.6). Feet flat on the floor. Both hands around a cup. Practise it when you are not triggered, several times a day for a fortnight, because under load you will only reach for something already grooved.

The naming sentence, silently. This is the competence one. Recognition converts an engulfing state into a familiar event.

The delay rule, set in advance. Above seven out of ten: no reply today. Written as a rule beforehand, because in the moment you will always believe this case is the exception.

3. New experience, repeated

This is the one that actually reduces the surge rather than managing it, and it is the same mechanism as 2.6: a prediction only updates when it is contradicted by evidence.

The implicit system predicts danger. Every time the situation occurs and nothing bad follows, that prediction takes a small hit. Every time you avoid it, the prediction survives intact — which is why avoidance feels like protection and functions as maintenance.

In practice this means deliberately staying in the situation without doing the defensive thing.

If your trigger is being questioned in public, the new experience is being questioned in public and responding neutrally, and surviving it. Not once — twenty times, over months.

The rules that make this work rather than becoming re-traumatising: start well below your limit and work up; stay until the arousal actually comes down rather than leaving at the peak, because leaving at the peak teaches escape; repeat frequently; and vary the setting, since the learning is context-bound (2.4).

This is exactly the exposure procedure that treats phobias and OCD (14.3), applied to an everyday trigger. It is slow, it works, and there is no faster route.

4. Change the meaning layer

Reappraisal (4.6), aimed specifically at the interpretation in layer three.

Find another reading that fits the same facts. "He asked me to double-check" also fits: he double-checks everything, he was covering himself, he had been burned by an error last month.

Attack the "should". Anger requires a rule (5.1). Say yours out loud and ask whether you would defend it in writing. People should not question me in meetings does not survive being said aloud. People should not mock me in meetings does — and knowing which one you are holding tells you whether to work on yourself or on the situation.

Separate the fact from the threat. He questioned the number is a fact. He thinks I am careless is a hypothesis. I am careless is a third thing entirely, and the slide from the first to the third takes under a second.

And use the distance moves — third person, the observer position, would-I-advise-a-friend (2.9, 1.7).

The special case: a person, not a situation

Sometimes the trigger is one individual, and everything they do lands wrong.

First, check whether it is a real pattern. Log their behaviour specifically, without interpretation, for two weeks. Sometimes the evidence shows a genuine pattern of undermining, in which case this is not a trigger to be worked on but a situation to be handled (13.2, 13.12).

Often the log shows something else — a small number of real incidents, a large number of neutral acts read as hostile, and an interpretation running ahead of the evidence (2.6's self-fulfilling model). If so, the work is on your model of them (10.11), and the most effective single intervention is a genuine conversation with them about something unrelated, because a specific human being is much harder to keep as a category.

And sometimes the person reminds you of someone. This is common, it feels absurd when you notice it, and noticing it is most of the cure. The reason a mild acquaintance can produce such a strong reaction is that they are pattern-matching to a face, a phrasing, or a manner from years ago.

Expectations, so you do not quit

Weeks one to three: you notice triggers after the fact. This feels like no progress and it is the necessary first stage.

Weeks four to eight: you notice during, and sometimes get the interrupt in. You still act badly, but less often and less far.

Months three to six: you notice at onset. The gap is real. The surge is the same size, and what you do with it is different.

Beyond six months: with repeated new experience, the surge itself begins to shrink for the specific triggers you have worked on.

One trigger at a time. Trying to work on four produces work on none. Take the one with the highest cost — usually the one that damages a relationship you care about — and give it three months.

What to do with this

Start the log tonight. Three weeks, six fields, one minute per entry, small ones included. Everything else on this page depends on having it, and no amount of reflection substitutes.

While you log, build the interrupt. Two breaths with a long exhale, practised twenty times a day for a fortnight, so that it is available before you need it.

Then pick one trigger and work all four levers on it — fuel, gap, new experience, meaning — for three months.

And tell one person what you are doing. "I'm working on how I react when I feel dismissed. If I get short with you, you can say so." This does three things: it makes the work real, it gives you an outside observer (4.5), and it repairs some of the damage the trigger has already done, because being told that someone is working on the thing that hurts you is itself a repair.

Next: 5.4 turns to the part of suffering that is not caused by the event at all — the second arrow, which is the oldest and most practical idea in this Part.