Most of us grew up learning that emotions are things that happen to us. Something occurs — a conflict, a loss, a surprise — and a feeling rises up in response. Anger, sadness, fear, joy. We detect these feelings, name them, and ideally manage them.
That picture is intuitive. It also isn't quite how the brain works.
The Upstream Signal Model is built on a different starting point — one that comes from research in predictive neuroscience, emotion construction theory, and stress physiology. The core idea is this: your nervous system doesn't react to the world. It predicts it. And everything you call an emotion is built from signals that existed long before the emotion word arrived.
This page introduces the foundational ideas behind the model. If you've ever felt like therapy got you to understand yourself but nothing actually changed, or like knowing what's wrong didn't help you feel differently — this is a map for why.
Here's the problem the brain is constantly solving: you have a body that needs to survive, and a world that is always changing. How do you know what's coming next? How do you prepare before things go wrong?
The brain's answer is prediction. Rather than passively waiting for information to arrive, the brain is always generating a model of what's happening — what's safe, what's threatening, what will happen next, what action is required. It builds this model from everything you've ever experienced, and it runs it continuously, beneath conscious awareness.
Incoming sensory information doesn't replace this model. It updates it. The brain asks: does this match what I predicted? If yes, nothing changes. If no — if there's a mismatch — the brain registers a prediction error, and adjusts.
The brain is not a camera that records reality. It is a prediction engine that is constantly asking: "Does this match what I expected?"
This matters enormously for how we understand nervous system responses. When your body braces before you've consciously registered a threat — when your stomach drops in an elevator, when your jaw tightens in a certain room — that's not irrational. That's your brain's prediction model running faster than language can keep up.
The researcher Lisa Feldman Barrett has spent decades studying how emotions are made. Her findings challenge something most of us take for granted: that emotions are universal, hardwired responses that the brain detects and labels.
What she found instead is that emotions are constructed. Your brain takes raw physiological signals from the body — changes in heart rate, muscle tension, gut sensation, energy levels — and builds an emotional experience from them by matching those signals to concepts it already has. "Anxiety." "Anger." "Grief."
The concept shapes the experience. If you grew up in a context where tight-chest-and-racing-thoughts got labeled "excitement," you experience it differently than someone who learned to call it "panic." The same signal. Different construction.
Emotions are not detected. They are built — assembled by the brain from body signals plus learned concepts, within a cultural context that taught you what those signals mean. This means the emotion label is an interpretation, not a raw fact.
This is not saying emotions aren't real. They are real, and they matter. It's saying they are the end product of a process — a construction — not the beginning of one. And the Upstream Signal Model asks: what happens if we pay attention to that process, instead of just its output?
Think of your nervous system as a stream. At the bottom, the source: raw body signals. Physical sensations. Changes in energy and tension. An undirected sense of something happening, before you've named it anything.
As that signal travels up the stream, it gets processed. It gets filtered through old memories — especially early, implicit memories about what's safe and what isn't. It gets matched against beliefs about how the world works. It gets shaped by strategies your system has learned to use when pressure rises. And eventually, far downstream, it arrives as an emotion word.
"I'm anxious." "I feel depressed." "I'm angry."
The word is real. But it's also the most compressed, most interpreted, most filtered version of the original signal. Working only with the word is like trying to understand a river by studying the water at its mouth — without ever going to see where it begins.
The emotion word is the end of the signal, not the beginning. The Upstream Signal Model maps the journey in between.
What happens between the raw body signal and the emotion label? The model identifies several distinct layers — each one a form of compression.
Each layer is a real part of the process — not a mistake, not a flaw. Compression is useful. It's how the brain moves quickly. The problem isn't that compression exists. The problem is when we try to resolve something at the narrative or label level that actually lives much further down the stack.
The dominant conversation about mental health has centered emotion for decades — naming feelings, validating them, building emotional vocabulary, processing through talk. This has been genuinely useful. Naming an experience matters. Validation matters. Feeling understood matters.
But there's a gap that many people have encountered: you can understand a pattern completely and still repeat it. You can name what happened, track where it came from, build a narrative that makes sense — and still find your body responding the same way in the same situations.
This model's position is not that emotion-centered work is wrong. It's that emotion labels exist at a particular layer of the stream — and that layer, on its own, cannot reach the patterns that live deeper. The signal isn't asking to be solved. It's asking to be met.
A decade of centering emotions. This model focuses on the signals that occur before the label arrives.
Working upstream doesn't mean ignoring the emotion word. It means using it as a starting point for a different kind of inquiry: what is this actually made of? Where does it begin? What layer does it actually live in — and what kind of response can reach that layer?
This site is a model — a conceptual map. It is meant as a teaching tool, a framework for curiosity, and a structure for understanding. It is not a clinical protocol, a diagnostic system, or a treatment approach.
The Full Stream visualization lets you see all thirteen steps of the model at once, toggle between states, and explore how different conditions map onto the architecture. The Compression System diagram shows the layered structure in a single image. The essays go deeper into each part — you can read them in order or start wherever your curiosity leads.
If you're a clinician, you may find the model useful as a conceptual frame for explaining mechanism to clients, or for understanding where different approaches actually work in the stream. If you're someone trying to understand your own experience, the map may help you locate things more precisely than "I feel anxious" — and point you toward what might actually help.
The Upstream Signal Model draws from predictive processing research, emotion construction theory (Barrett), allostatic regulation (Sterling & Eyer), and clinical observation. It represents theoretical interpretation and conceptual synthesis — not established clinical protocol or empirically validated treatment. This site is for educational and reflective purposes. It is not therapy, diagnosis, or medical advice, and is not a substitute for professional care. If you are experiencing significant distress, please reach out to a qualified mental health professional.