THE STORMS · FOUR ALARMS
On four states that arrive as one — and on the caution that belongs before any of them.
These four are the states most often talked past. An athlete says something feels wrong; a worker says the situation is not safe; somebody says they are frightened. Each of these can be met with a protocol.
Alarm states are frequently accurate. Fear about an actual hazard, anxiety about a genuinely unprepared-for demand, anger at a real injustice — in each case the state is doing its job, and regulating it without addressing what produced it removes the signal and keeps the hazard.
So the order is fixed here more firmly than anywhere else in this collection: establish whether the thing is true before doing anything about the feeling. Somebody who is afraid of an injury they can feel does not need a breathing protocol.
Fear typically involves three things together: an identifiable threat, imminence, and doubt about coping. That is a common pattern rather than a definition — people fear vague and distant things too. Where the pattern does hold it is useful, because changing the coping term changes the state.
Anxiety more often runs without a specific object — same activation, nothing to address, which is part of why it persists. But plenty of anxiety is about one identifiable future event, so treat this as the usual difference rather than the boundary between the two.
Anger answers a different question: something is wrong and somebody did it. It raises energy and narrows attention onto an agent.
Frustration is blocked goal pursuit — the effort is going in and the result is not coming out. It is the one most often mistaken for anger, and it resolves differently: anger wants an agent addressed, frustration wants the block found.
Mine. The practical content is that you can take the information and decline the instruction.
Fear tells you accurately that something could hurt you and then proposes that you stop, which is often the wrong move. Anger identifies a real wrong and proposes an immediate response, which is usually worse than the delayed one. Take the reading, refuse the recommendation.
Fear and anxiety are the best-evidenced states in performance research. Attentional control theory is the sturdiest piece: anxiety consumes working-memory resources and shifts attention toward threat, degrading tasks that need those resources while often leaving effort and even outcome intact, because people compensate.1 That gap between effort and efficiency is the useful finding.
Intensity and direction are separable. The same level of activation can be read as facilitative or debilitative, and the reading predicts outcome better than the level does.2 That result is from self-report instruments and the causal direction is not established.
Reappraising arousal as excitement rather than calm has supporting evidence, though the best-known studies are small and the effect has not had the large-scale replication it deserves.3 Use it, do not lean on it.
On anger, Woodman’s work found it raised peak force and did not help a task requiring accuracy — the fit-to-task finding this library has used elsewhere.4 Small samples.
And the training finding worth having: practising under pressure transfers better than practising calm. Oudejans and colleagues showed skills acquired without anxiety degrade under it, and acquiring them under mild pressure protects them.5
The objection concerns what a career of using these states does.
The case against
The instrumental case is real — anger raises force, anxiety sharpens vigilance — and the cost is that the person is in those states while it works. Over a decade that is a great deal of time spent somewhere unpleasant in exchange for performance on tasks that may not have warranted it.
The Buddhist objection stated in this library’s Not-Two collection applies with full force here, and the performance evidence does not answer it. It answers a narrower question.
The limit worth holding: point an alarm that has already arrived, do not manufacture one. Those are different practices with different bills.
Establish whether the alarm is accurate. First, every time, before anything else.
Separate the four by their question. Is there a threat, is there no object, did somebody do this, or is something blocked? Four different answers, four different next moves.
Take the information and decline the recommendation. The reading is usually right; the proposed action usually is not.
And practise under mild pressure rather than in calm. Skills learned calm degrade when it matters, which is the best-supported practical finding in this family.
The sorting is most of the value.
Threat, no object, an agent, a block. Four questions that often separate states arriving as one indistinguishable lump, and asking them is worth more than any technique applied to the lump. They are sorting prompts rather than definitions, and a state that does not fit cleanly is common rather than a failure of the scheme.
The evidence is the strongest in the collection: anxiety degrades efficiency more than effort, direction beats intensity, and pressure training transfers where calm training does not.
And the caution stands in front of all of it. These states are frequently accurate, and regulating an accurate alarm removes the signal and keeps the hazard.
Before anything else, establish whether the alarm is accurate. An accurate alarm regulated is a signal removed and a hazard kept.
Then sort by question — threat, no object, an agent, or a block — because the four resolve in different directions. Take the reading; decline the recommendation.
Research is cited by author and publication, with replication status where it bears on whether to believe the claim. The full per-emotion reference lives in SportsFlow.