Appearance
9.8 — How a Person Actually Changes
Two claims are made about this and both are wrong.
"People never really change." Contradicted by the maturity data (9.1): the same people, measured across decades, become substantially more conscientious, more agreeable and less emotionally reactive. That is not nothing; it is a different person to live with.
"You can be anything you decide to be." Also false. Traits are 40 to 60 per cent heritable, change is slow, and most of it is in one direction.
The accurate version is narrower and more useful: specific things change, at specific rates, by specific mechanisms. This page is the map of which.
What changes readily
Skills. Anything trainable — communication, a technique, a language, the ability to handle a difficult conversation. Weeks to months with deliberate practice (3.5).
Knowledge and beliefs. Fast, if you are looking at evidence and slow if the belief is attached to identity (1.5).
Habits. Months (8.2).
Emotional reactivity to specific things. Individual triggers weaken with repeated new experience (5.3). Months per trigger.
How you interpret events. Reappraisal is trainable and this is what most therapy does (15.3). Weeks to months, with practice.
And what you do, which is the point — because behaviour is the thing everything else is measured through.
What changes slowly
Trait levels. Years, in the direction of the maturity trend, faster with deliberate structured effort or with therapy.
Attachment patterns. Years, mostly through a different kind of relationship or through therapy (9.5).
Baseline emotional reactivity. Years, and it does move.
Self-concept. Years, and only when the behavioural evidence has already accumulated (8.7).
What barely changes
Temperament. The starting point (9.4). What changes is what you do with it.
Trait rank order. You will probably always be more or less conscientious than your siblings.
Intelligence in the fluid sense (1.9), beyond removing what is suppressing it.
And basic preferences — what you find interesting, what tires you, what you enjoy. These drift and they do not reverse.
Accepting this list is not defeatism. It is what allows the effort to go where it produces something, and 9.1's conclusion applies: arranging your life around a trait is cheaper and more effective than trying to move it.
The mechanism
Everything that genuinely changes, changes by the same route, and the route runs in the opposite direction from the intuitive one.
Behaviour changes first. Feeling follows. Self-concept follows that.
Not: understand yourself, feel differently, then act differently. That order almost never works, and it is why years of insight can produce no change at all.
The evidence for the reverse order is scattered across this volume. Behavioural activation works by doing before feeling (8.8). Exposure works by acting against the prediction (2.6). Habits build identity by accumulating evidence (8.7). Opposite action changes the emotion by declining its instruction (4.6). Self-respect is built from kept promises (5.7).
In every case the leverage is the same: do the thing, repeatedly, and let the internal state catch up.
The four ingredients
Where change actually happens, four things are present. Where it does not, one is usually missing.
A specific behaviour, not an intention. "Be more patient" produces nothing. "When my son asks the same question a third time, I answer it once more before saying anything else" is a behaviour.
Repetition over months. Everything in this volume that works runs on months. Anything promising a transformation in a weekend is describing a mood.
Feedback. Some way of finding out whether it is happening — a log, another person, a count (3.5). Without it you cannot tell, and you will conclude either that you have changed when you have not, or that you have not when you have.
And a supporting environment. From 8.2: context is the strongest determinant of behaviour, and a change attempted inside an environment that supports the old pattern is being fought uphill continuously.
What genuinely changes people, in practice
Observed rather than theorised, and it is a short list.
A role with real responsibility. Consistently associated with trait change (9.1). Somebody depending on you changes behaviour before it changes anything else.
A relationship with someone who is different from what you are used to. The main documented route to earned security (9.5), and it works over years.
Therapy, which produces measurable trait-level change and is the most reliable deliberate route (Part 15).
Being in an environment where the new behaviour is normal. Underrated and powerful. You become like the people you are around, in ways you do not notice and cannot easily resist (11.1).
A serious enough consequence. The health scare, the marriage ending, the job lost. This works and it is a terrible plan, because it is unreliable and expensive and frequently arrives too late.
And sustained deliberate practice of a specific behaviour. The only one entirely under your control, and the reason this volume is written as it is.
What does not change people
Insight alone. The most important item here. Understanding exactly why you do something, in detail, with a full account of its origin, changes nothing on its own. A great deal of self-examination is a comfortable substitute for a difficult behaviour.
Information. Everybody who smokes knows.
A decision. New year's resolutions have a documented failure rate for a structural reason: a decision is a state (8.1).
Being told. By anyone, however correct.
Shame. Reliably counter-productive (4.8, 8.5).
And waiting. Age produces the maturity trend and nothing more specific.
A realistic timeline
For one deliberate change, honestly.
Weeks 1 to 4. You notice the old pattern after it happens. It feels like nothing is working. This stage is necessary and most people quit inside it.
Weeks 4 to 12. You notice during. You still mostly do the old thing, and sometimes you do not.
Months 3 to 6. The new behaviour happens sometimes without effort. Under load, the old one returns — and that return is expected, not a relapse into failure.
Months 6 to 24. The new behaviour is the default in most conditions. The old one appears under high stress, tiredness or in the original context.
Beyond that. It is who you are, and you will find it hard to remember being otherwise — which is itself the reason people underestimate how much they have changed.
And the old pattern is never deleted (8.4). It stops being triggered. Under the original conditions it is still there, which is worth knowing so that its reappearance in a bad month does not read as proof that nothing changed.
What to do with this
Pick one behaviour, not one trait. Specific enough to be observed by someone else.
Give it six months and expect the first month to feel useless.
Build the feedback in from day one — a log, or a person who will tell you.
Change something in the environment, because the environment is doing more work than your intention.
And stop waiting for the insight that will make it easy. From this whole Part: the insight tells you what to do. The doing is the change, and there is no version where the second part is optional.
Next: Part 10 turns outward, to the other people — starting with the uncomfortable question of how accurately you actually read them, and how confident you are about it.