There is a version of this you have already lived.
Something arrives — a number that is wrong, a decision made without you, a question in front of people that you did not want asked in front of people. And before the meeting is over you have done a thing you would not have chosen if somebody had given you ninety seconds and a whiteboard. You were sharper than you meant to be. Or you committed to something you had not thought through. Or you went quiet in a way that everybody in the room noticed and nobody mentioned.
Afterward you can see it clearly. You know what the better response was. You could have written it down at the time if you had been asked.
That gap — between the response you would have chosen and the one that showed up — is not a discipline problem. It is the visible output of two systems in your head that were never built to agree, running at different speeds, with different jobs, and a standing conflict of interest between them.
Leadership by Design treats that conflict as the second phase of the organizational lifecycle: Reaction. And it holds that most of what leaders try to fix at the level of behavior is being generated one layer beneath behavior, in a structure that does not take instruction.
What is happening in your head before a leadership decision?
Two systems are evaluating the same event, and they are not doing the same job.
The prefrontal cortex is the one that does what you consider thinking. Planning, sequencing, holding several possibilities open at once, weighing a consequence that has not happened yet against an impulse that is happening now. It also does inhibitory control — the work of not doing the first thing available. It is deliberate, it is flexible, and it is slow, and it is where every leadership skill you have ever been taught is stored.
The limbic system, and the amygdala in particular, does something else. It is a rapid relevance detector: it evaluates incoming information for threat and significance and it initiates a response, fast, before the slower system has finished forming a picture. It is not the seat of emotion in the crude sense — it is closer to a triage mechanism, deciding what matters enough to act on before full analysis is available.
These two are not partners. They are a structural tension — two systems built for different problems, running on the same input, arriving at different answers, with the faster one moving first.
That should sound familiar, because it is the same shape as every organizational tension in the framework. A fast local response against a slower central one. Discipline against drift. Control against expansion. The tension you have been trying to manage in your organization is already running inside the person managing it, one layer down, on the same architecture.
The conflict is not a defect in you. It is the design, and it is the same design your organization runs on.
Why is the fast system first?
Because being right slowly has historically been worse than being roughly right quickly.
Joseph LeDoux’s work established that threat information travels two routes at once. One goes from the thalamus directly to the amygdala — a rough signal, low resolution, and quick. The other takes the longer path through the cortex, where the input gets processed properly and arrives with far better information and considerably later.
The fast route wins on timing by a wide margin. And it wins for a reason that has nothing to do with your leadership: for most of the history of anything with a nervous system, the cost of reacting to a threat that was not there was almost nothing, and the cost of pausing to verify one that was there was everything.
So the architecture is not a bug. It is a speed-accuracy trade-off that was settled long before there were meetings, and the settlement favors speed by design.
What that means in a conference room is that your response initiates on the rough signal. The good information — the context, the history with this person, what you know about their quarter that would have changed how you heard it — arrives afterward, into a situation where something has already started.
What does pressure do to the balance?
It moves it, and it moves it in the direction you would least choose.
This is the part that surprises people, and it is well documented rather than speculative. When a threat response initiates, cortisol and catecholamines are released. Those chemicals do a number of useful things. They also impair prefrontal function — the working memory, the flexible thinking, the inhibitory control — while leaving amygdala-driven responding intact or enhanced.
Read that again with a leader’s eye on it. The same pressure that triggers your automatic response is the pressure that degrades your capacity to override it.
That is not a metaphor and it is not a figure of speech. It is a documented biochemical relationship, and it has a consequence that most leadership development never accounts for: the behaviors you have worked hardest to develop are least available in the moments you need them most.
Every skill you have built lives in the slow system. Every one of them requires the slow system to be operating well. And the situations that call for them are the situations that compromise it.
That’s the cruelty in the design, and it explains something leaders otherwise experience as personal failure — the sense of watching yourself do the thing you had decided not to do, with full awareness, and being unable to reach the alternative you can see clearly.
You were not weak in that moment. You were under-resourced, in a measurable way, by the same event that required the resource.
Does that mean I did not decide?
No, and the honest answer is more interesting than either extreme.
The research most often cited here is Benjamin Libet’s, at UCSF in 1983. Subjects were asked to flex a wrist whenever they chose and to note the position of a rotating clock at the moment they became aware of the intention to move. Electrical activity preceding the movement — the readiness potential — began, on average, about 550 milliseconds before the movement itself. Subjects reported becoming consciously aware of the intention roughly 200 milliseconds before it.
The activity that preceded the action started before the awareness of choosing it.
The interpretation of that finding is contested, and Leadership by Design does not attempt to settle it. Some researchers have argued the readiness potential reflects accumulating neural noise rather than a decision being made. The philosophical argument about free will has been running for forty years and it is not going to be resolved in an article about leadership.
But the framework does not need the strong version of the claim. It needs only the modest one, and the modest one survives every critique: initiation can precede awareness. Something can be underway before you know you have started it.
That is enough. It means the useful question is not whether you chose the reaction. It is what was available to you at the moment the reaction was already in motion — and what you had built beforehand, when the slow system was working, that could still shape what happened next.
Why doesn’t training hold under pressure?
Because most of it is installed in the wrong system.
The standard approach to leadership behavior is instruction: learn the framework, practice the technique, understand the model, commit to the intention. All of that is real and all of it is stored in the slow system. It is available when you are rested, prepared, and not currently being challenged in front of your board.
Then the quarter goes badly, and the training is where you left it — in the system that pressure just degraded.
This is why leaders keep concluding that they need more discipline. The evidence looks like a willpower problem: you knew the right response, you failed to produce it, therefore the gap is effort. But willpower is not a separate faculty sitting outside the biology. Inhibitory control is a prefrontal function, which means the thing you are trying to apply more of is made of the resource that the situation is currently consuming.
You cannot solve a resource-depletion problem by spending more of the resource.
Which points at something the framework holds and most leadership development quietly contradicts.
Pressure doesn’t create readiness. It reveals it.
Nothing gets built in the moment. What shows up in the moment is whatever was already installed, running on whatever capacity the moment has left you. The quarter that goes badly is not a test you can study for while sitting it — it is a readout of work that was either done beforehand or was not.
That’s the point where most leadership advice runs out, and it is where the framework starts being useful. If the moment cannot be fixed inside the moment, the work has to happen somewhere else.
What does this mean for the organization?
That your automatic response is not a private event.
Whatever fires in you before you have decided anything happens in front of people, and it happens fast enough to be read as truthful. A composed response comes with explanation attached, which gives the room something to weigh it against. A reaction arrives with no explanation, faster, and it is believed more — because it was not composed.
So your organization learns most efficiently from you at the moments you have the least control over what you are teaching.
And it learns something durable. One reaction is an event. A pattern of them, observed over a year, becomes the working model people navigate by — what happens here when things go wrong, what is safe to bring up and when, which conversations cost more than they return. That is the Reaction phase producing the Signal phase’s material, which is why the framework places them adjacent and in that order.
There are recognizable patterns in how leaders’ automatic responses take shape — the responses are not random, and they differ in ways that hold across situations and across years. Sorting out what those patterns are, and what each one produces in the formation around it, is the subject of Of Rabbits and Rhinos rather than of this article.
What matters here is the structural point: the response fires before you decide, and the organization is watching.
If you cannot win the tension, what can you do?
Move the work outside the moment, because inside the moment the resource is already spent.
What changes the outcome sits outside the moment, and none of it is trying harder while the moment is happening.
Widen the window. There is a real interval between a threat registering and a response committing, and it is not fixed. Anything that lengthens it — a stated practice of not responding to significant news in the first meeting, a standing rule that certain categories of decision wait twenty-four hours — buys time for the slow system to arrive. This is not composure. It is a procedure that does not require composure, which is the entire point.
Build the governor before you need it. A governor is a structure that holds when your judgment cannot: a person with standing to interrupt you, a decision that cannot be made alone, a threshold above which something has to go to committee whether or not you agree. Built in advance, running whether or not you have attention available. The one thing it must be is expensive to override in the moment, because in the moment you will want to override it.
Repair afterward, out loud. The reaction that already happened is not recoverable, but its teaching value is. Going back to the room — naming what happened, saying what you would do differently, and then doing that differently the next time — converts an event that taught one thing into a pattern that teaches another. This is the most underused move available to a leader (and the cheapest, BTW!).
None of this makes you a calmer person. It is not supposed to. The tension between your two systems is permanent, and the framework does not propose to resolve it — it proposes to design around it, which is what you would do with any other structural constraint you could not remove.
Where this leads you
To a different relationship with your own reactions than most leaders are offered.
Not what is wrong with me that I did that. Something fired that was built to fire, faster than the part of you that would have chosen otherwise, and it did so under conditions that had already reduced your ability to intervene.
The useful questions are structural. What does mine tend to do? What has my organization learned from watching it? And what have I built that holds when I cannot?
The first of those you can start on in five minutes. The second takes structured observation, because the people who could tell you have been managing around it long enough that it no longer registers to them as a point to raise. The third is design work, and it is the only one of the three that changes what happens next quarter.
What your organization has been learning from you → Signal integrity, and why watching beats listening.
Of Rabbits and Rhinos → Book Two of the series takes the patterns in threat response and works out what each one produces in the formation around it.
