The Carry Circuit

Most advice about anger fails because it treats one system as though it were three, or three systems as though they were one. Take the thing apart properly and the intervention point stops being a matter of temperament. It becomes a matter of design.

Call it the Carry Circuit. Three components.

Component one is the event. Somebody changes lanes without looking and occupies the space your car was going to be in. Note the property that matters for design purposes: the event has already terminated. It completed before you finished perceiving it. It has no duration going forward, no power source, and no capacity to act on you again. Whatever it did, it did once, in the past tense, and it is now a fixed historical fact with no ongoing draw.

Component two is the charge. This is the chemical cascade that fires on perception of the event. It is fast, involuntary, and physiological. It is also self terminating: the surge runs roughly ninety seconds to two minutes and then clears on its own, with no action required from you. That property is the important one and it is almost universally missed. The charge is not the problem, because the charge solves itself. Left entirely alone, component two exits the system in under two minutes.

Component three is the ledger. This is the narrative account you open on the event and then keep writing entries into. He did that to me. He saw me. People like that are why this city is what it is. The ledger has no natural termination condition. It does not decay, because it is actively maintained. And here is the load bearing property: the ledger is the only component in the circuit that you supply.

Now run the analysis, because a model that cannot be tested is decoration. Of the three components, which one do you actually own?

Remove component one and nothing further happens, but you have no control over component one. It belongs to the other driver, the weather, the day. Any strategy that depends on managing the events other people generate has a dependency you do not own, which makes it not a strategy.

Remove component two and nothing further happens either, but component two is autonomic. You do not get a vote on whether the cascade fires. Advice built on not feeling the surge is advice built against your own hardware, and it fails in the predictable way: under load, it inverts.

Remove component three and the circuit clears itself in ninety seconds. That is the finding. The only removable component is also the only one you own, and it is the only one nobody thinks to look at, because it does not present as a component. It presents as being right.

There is a compounding term, and it is what makes this worth engineering rather than tolerating. A ledger held open past the ninety seconds does not stay the same size. Held for a day it operates as a mood. Held for a season it operates as a temperament. Held long enough it stops being something you do and becomes something you are, at which point other people begin describing the system from the outside using words like bitter and resentful. That is not a character flaw arriving from nowhere. That is an accumulation curve, integrating ninety second intervals over years.

So, the intervention, stated as a specification. At the moment the charge fires, do not attempt to suppress it, because it is autonomic and suppression is not available. Instead, decline to open the ledger. Hold for the duration of the surge, which is under two minutes and generally far less at this scale, and take no narrative action inside that window. Do not compose the sentence about what kind of person he is. Do not rehearse the account you would give of it. The window is the whole intervention. Nothing else in the circuit requires your attention.

Now the strongest objection, which comes from control theory rather than sentiment and is therefore worth answering properly. A system with no feedback path cannot correct. If you never signal that a behaviour was unacceptable, you have removed the only mechanism by which the behaviour changes, and you have optimised your own internal state at the cost of the shared system everyone drives in. That is a serious argument.

The answer is a measurement, not a value judgement. Estimate the probability that a horn blast, a gesture, or a shouted sentence through two panes of closed glass modifies the future driving behaviour of an anonymous stranger you will never encounter again. It is not zero, but it is close enough to zero that the feedback path is nominal rather than real. You are not trading your internal state for system correction. You are paying the full price of the feedback path and receiving none of the correction. If the objection held, the design would change. It does not hold at this scale, which is precisely why it holds at other scales, and knowing the difference is the actual skill.

Where this model does not apply, stated plainly. It does not apply where the event is ongoing rather than terminated, because component one is then still live and the analysis above assumes it is not. A repeated harm from a person with continuing access to you is a different circuit with a different intervention, and declining to open the ledger there is not equanimity, it is a failure to respond to live input. It also does not apply where component two is not a spike but a sustained baseline, which is a physiological condition rather than a design problem, and belongs with a clinician rather than an engineer.

So: three components, one of which terminates instantly, one of which terminates in ninety seconds, and one of which never terminates because you keep it running. The leverage point in any system is the component nobody has classified as a component. Here, it is the one that feels like the truth.

Which transition in your day is currently running an open ledger you never decided to keep? Build the model into your week properly, and come to the workshop, where we take these circuits apart on paper and find the leverage points together.