Movement is one of the richest signals biology gives us: how fast we move, how precisely, how much effort we invest, and how this shifts across a lifetime reveal the state of the circuits and body behind it, often before any symptom does. But that signal is only useful if we read it correctly: and composite scores routinely don't, collapsing distinct operations into one number that can hide the very effect we're looking for.

That is why we start from behaviour itself, decomposed rather than summarised, and why we test this across movement disorders, mostly Parkinson's disease, where dopaminergic degeneration reshapes movement in specific, trackable ways. Reading it well takes three things: a way to causally perturb the circuits producing movement (pharmacology, DBS), a way to measure it with enough precision to see the decomposition (wearables, computer vision), and a way to elicit it under conditions realistic enough for the perturbation to mean something (programmable immersive environments). The same logic extends to the cognitive dimension, since movement never happens in a vacuum. Movement unfolds within, and is shaped by, the space and circumstances around it.

This is the groundwork we need before building system-level therapies, from neuromodulation to rehabilitation, that target what is actually broken.