There's No Such Thing as a Violent Person

There are only people in violent situations.

This is the line that took me longest to actually accept while reading Robert Sapolsky's Behave. It runs against everything we usually say about other people — that some are dangerous, some are kind, some are honest, some are not. We treat character as a stable property and then react with surprise when people turn out to be different in a different room.

Sapolsky spends 700 pages arguing that this whole way of thinking is wrong. Behaviour is the output of a layered, mostly invisible stack of causes — biology, situation, history, evolution — and the part we call "personality" is just the part we can see.

Once you take that seriously, almost everything you thought you knew about why people do what they do has to be rebuilt.

The time-scale stack

The book is organised around a single question: why did this person do that thing?

Sapolsky's move is to refuse a single answer. The honest answer is a stack:

A second before — what fired in the brain. Minutes before — what was in the room, what was the lighting, who was watching. Hours and days — hormone levels. Days and months — the neural plasticity from recent experience. Years — adolescence, childhood. Generations — culture and parenting. Hundreds of generations — gene-environment co-evolution.

All of these are running simultaneously when someone acts. Pick one and you've cherry-picked. Pretty much every confident explanation of someone's behaviour you've ever heard or made was a single layer of this stack, dressed up as the whole story.

The disturbing implication is that if you genuinely want to predict or understand behaviour, you have to look at the whole stack — and most of it is invisible to you, including in yourself.

What testosterone actually does

The most useful single chapter for me was on hormones, because the popular story about them is so wrong.

Testosterone does not cause aggression. It amplifies whatever behaviour the local culture associates with status. In a culture that rewards generosity, high-T individuals are more generous. In a culture that rewards aggression, more aggressive. The molecule is doing the same thing in both cases — making people care more about being seen as high-status. The culture decides what status looks like.

Oxytocin is similar. It's marketed as the "love hormone." It is — for your in-group. The same molecule increases hostility and suspicion toward your out-group. It's not pro-social. It's pro-tribal. The thing that makes you loyal to your team is the same thing that makes you unfair to the other team.

Both of these hormones are doing what evolution built them to do, which is to make us better at the social games our ancestors played. Neither is a "good" or "bad" chemical. They amplify whatever the context provides.

This connects directly to The Elephant in the Brain. Simler and Hanson tell you what's really going on socially — most behaviour is signaling, status, and tribal alignment. Sapolsky tells you the chemistry that makes those games possible. Read together they're brutal.

The brain decides before you do

The amygdala — your brain's threat-detection centre — classifies any new face as us or them in a tenth of a second. Before you know you've seen the person, the verdict is in. Conscious awareness arrives late, and your job is mostly to narrate a decision that's already been made.

The good news is that the categories are flexible. The brain isn't using "us" and "them" in a fixed sense. It's using whatever distinguishing marker is most salient — race, accent, team, profession, dress. Re-categorisation works. People you've been trained to see as them can be re-trained as us, with effort and exposure.

The prefrontal cortex is what does that re-training, and what overrides the amygdala's snap verdicts when the situation calls for it. The PFC is also what lets you do the harder thing when the harder thing is the right thing. It is metabolically expensive — it depletes with use, with hunger, with fatigue, with stress. Late-night decisions are worse than morning decisions for the same reason a worn-out gym muscle lifts less weight. The PFC is a muscle.

This means the most important calls of the day should happen when your PFC is rested. Not late, not after a long meeting, not after a difficult conversation. The version of you that's been working for ten hours is not the version that should be making the high-stakes call. Schedule against your own depletion curve.

This is also where Sapolsky maps onto Kahneman. Thinking, Fast and Slow describes the cognitive output of a depleted PFC — System 1 running unsupervised. Behave gives you the wiring. Same animal, different angle.

Childhood, in everyone

The Adverse Childhood Experiences (ACE) studies show that early adversity leaves measurable, persistent biological changes — altered HPA axis function, smaller hippocampus, larger amygdala, stress responses calibrated for an environment that no longer exists. These don't go away with age. They are baselines, not destinies, but they are real baselines.

This is one of those facts that becomes uncomfortable once you sit with it. Every adult you deal with came in pre-shaped by experiences you'll never see. You're never dealing with a person where they are now. You're dealing with the sum of where they've been.

It also means that children — yours, your team's, your friends' — are being calibrated, right now, by environments they didn't choose. The stress of a household becomes part of the architecture of the brain that emerges from it. This is one of the few cases where popular intuition (childhood matters) is exactly correct, and the science gives you the chemistry.

Hierarchy is biological

Rank in a hierarchy is not just a social fact. It alters biology. Subordinate individuals — in baboons, in humans, across many species — get more cardiovascular disease, more depression, faster cellular aging.

The pattern depends on the type of hierarchy. Stable hierarchies stress those at the bottom. Unstable hierarchies stress those at the top. There is no "stress-free" position; the structure produces the disease, and the structure is what changes who pays.

The practical use of this in business is large. People operating below their actual rank — informal subordination, unrecognised contribution, structural under-positioning — are paying a biological tax. Recognising rank, even informally, is one of the cheapest performance interventions available, because the cost of subordinate stress is so high.

The free will question

The most controversial thread, more developed in Sapolsky's later book Determined but seeded throughout Behave: free will, in the strong sense, doesn't exist. Every behaviour traces back to causes the person didn't choose — biology, upbringing, neural state, hormones, situation.

This isn't a counsel of despair. It's a counsel of mercy. The contractor who screwed up didn't choose, in the deepest sense, to be the system that produced that behaviour, any more than you chose to be yours. You still hold them accountable for outcomes — accountability is part of how systems learn. But you don't necessarily moralise about character.

Character is just the part of the time-scale stack you can't see.

This shifts how you respond to bad behaviour from outrage to curiosity about causes. Less retribution, more system design. Less "what kind of person does that?" and more "what would have to change for that not to happen again?" The first question doesn't help. The second one does.

The actual project

What I think Sapolsky is doing under the surface is more interesting than the catalogue of biology.

He's making the case that the way most of us think about other people — fixed personalities, moral characters, blame-deserving choices — is a bad model. It's a model that produced a lot of suffering historically (most of it in the form of harsh judgments and harsh institutions) and continues to produce it now in smaller, more personal ways.

The replacement model is more humble and more useful. People are biological systems running in contexts. Change the context and the behaviour changes. Change the system over time and the person changes. The visible behaviour is the tip of a stack you can never fully see, even in yourself.

Once that's the model, you stop trying to find good people and start designing good contexts. You stop trying to be a good person through pure willpower and start building structures that produce good behaviour whether or not your willpower shows up that day.

This is also the bridge to Peterson Academy. Most of what Peterson teaches in his lectures on personality and life-design is, in Sapolsky's terms, the project of training the PFC to override the parts of the limbic system that lead you astray. It's the conscious construction of a system that produces the behaviour you want from biology that wouldn't, by default.

Reading Behave didn't make me less judgmental. It made me less confident. There's a difference. The first is a posture you adopt; the second is what happens when you actually look at how complicated people are.

Sources: Robert Sapolsky, Behave. Also referenced: Robin Hanson & Kevin Simler's The Elephant in the Brain, Daniel Kahneman's Thinking, Fast and Slow, and Peterson Academy.

This post contains affiliate links. If you buy through them I may earn a small commission, at no cost to you.