Human Motor Cortex Encodes Complex Handwriting Through a Sequence of Primitive Neural States

How the human motor cortex (MC) orchestrates sophisticated fine movements such as handwriting remains a puzzle1-3. Here, we investigate this question through Utah array recordings from human MC hand knob, during imagined handwriting of Chinese characters (306 characters tested, 6.3 {+/-} 2.0 strokes per character). We find MC programs the writing of complicated characters by sequencing a small set of primitive states: The directional tuning of motor neurons remains stable within each primitive state but strongly varies across states. Furthermore, the occurrence of a primitive state is encoded by a separate set of neurons not directly involved in movement control. By automatically identifying the primitive states and corresponding neuronal tuning properties, we can reconstruct a recognizable writing trajectory for each character (84% improvement in reconstruction accuracy compared with baseline). Our findings unveil that skilled, sophisticated movements are decomposed into a sequence of primitive movements that are programmed through state-specific neural configurations, and this hierarchical control mechanism sheds new light on the design of high-performance brain-computer interfaces.