← upstream
RAW · OPEN · CLOSE TO THE SOURCE
COMPRESSED · SYMBOLIC · DOWNSTREAM
downstream →
upstream
01 · INPUT
Something happens
A change in the world, or inside your body. Raw and uninterpreted. The signal enters the nervous system before any meaning has been assigned to it.
🔥
02 · CAPACITY + THREAT
The body reads it
Two questions run at once: Do I have enough? (the tank) and Is this a threat? (the burners). These aren't thoughts. They're the nervous system's resource check.
03 · ALARM OUTPUT
Pressure
The alarm fires. Arousal rises, urgency is assigned, the system is mobilized toward something. This is felt in the body — tightness, activation, the pull to act — before any word arrives.
04 · PREDICTION ASSEMBLED
A strategy forms
The brain matches the signal to what it already knows. It assembles a prediction about what to do. This is built from every similar thing that has happened before — not from the signal itself.
downstream
05 · OUTPUT
The word
An emotion label. A feeling you can name. Compressed, symbolic, very far from the source. The word feels like the thing — but it's the most downstream translation of everything above.

Each stage compresses the one before it. By the time you have a word — "I'm anxious," "I'm fine," "I'm exhausted" — you're working with a summary of a summary of a summary. The signal that started this is still running. The word just arrived last.

Most of how we understand ourselves starts at stage five.
This model starts at stage one.

THE UPSTREAM SIGNAL MODEL · WHAT IT MEANS TO WORK CLOSER TO THE SOURCE

Why this matters
The word and the signal are not the same thing
"Anxious" might mean your body has no resources left. Or that a threat was detected. Or that a pattern from early in your life is running. The word collapses all of these into one category — and treating the word without knowing which one you're dealing with is why so much gets stuck.
Why it gets stuck
The system predicts what it already knows
At stage four, the brain isn't just reading the signal — it's matching it against everything it already expects. High-confidence predictions filter what gets through. Over time, the stream stops carrying new information. It just confirms the map it already has.
Start here — read the intro → Explore the full model