This study investigates the dynamics of neural activity during reaching movements in motor and premotor cortex. The research challenges the traditional view that motor cortex neurons represent movement parameters, instead proposing that neural activity reflects a dynamical system generating and controlling movement. The study reveals that motor cortex responses during reaching contain a brief but strong oscillatory component, suggesting a mechanistic role for preparatory neural activity. These results demonstrate unexpected yet surprisingly simple structure in the population response, which explains many of the confusing features of individual-neuron responses. The study found that neural responses during reaching exhibit rotational patterns in a state space, similar to rhythmic activity in other motor systems. These rotations are consistent across conditions and are influenced by preparatory activity. The research also shows that these rotational patterns are not directly related to the curvature of the movement, but rather to the underlying dynamics of the neural system. The study compared the neural data with simulated models of motor cortex activity, including velocity-tuned and complex kinematic models, and found that only the generator model, which emulates a simple dynamical system, produced rotational patterns similar to the observed data. The generator model was able to accurately predict EMG data, suggesting that neural activity may be generated through underlying rotational dynamics. The study also found that the rotational patterns in the neural data are not simply due to multiphasic responses, but rather to a pair of multiphasic patterns with phases consistently 90 degrees apart. This suggests that the neural population contains a complementary pair of patterns that are not present in simulated or muscle populations. The study concludes that the rotational patterns in the neural data provide a natural basis set for generating non-rhythmic movement, and that motor cortex can be understood as an engine of movement that employs lawful dynamics. The findings suggest that the controversy over what single neurons in motor cortex 'code' or 'represent' may be resolved by focusing on the dynamics that generate movement.This study investigates the dynamics of neural activity during reaching movements in motor and premotor cortex. The research challenges the traditional view that motor cortex neurons represent movement parameters, instead proposing that neural activity reflects a dynamical system generating and controlling movement. The study reveals that motor cortex responses during reaching contain a brief but strong oscillatory component, suggesting a mechanistic role for preparatory neural activity. These results demonstrate unexpected yet surprisingly simple structure in the population response, which explains many of the confusing features of individual-neuron responses. The study found that neural responses during reaching exhibit rotational patterns in a state space, similar to rhythmic activity in other motor systems. These rotations are consistent across conditions and are influenced by preparatory activity. The research also shows that these rotational patterns are not directly related to the curvature of the movement, but rather to the underlying dynamics of the neural system. The study compared the neural data with simulated models of motor cortex activity, including velocity-tuned and complex kinematic models, and found that only the generator model, which emulates a simple dynamical system, produced rotational patterns similar to the observed data. The generator model was able to accurately predict EMG data, suggesting that neural activity may be generated through underlying rotational dynamics. The study also found that the rotational patterns in the neural data are not simply due to multiphasic responses, but rather to a pair of multiphasic patterns with phases consistently 90 degrees apart. This suggests that the neural population contains a complementary pair of patterns that are not present in simulated or muscle populations. The study concludes that the rotational patterns in the neural data provide a natural basis set for generating non-rhythmic movement, and that motor cortex can be understood as an engine of movement that employs lawful dynamics. The findings suggest that the controversy over what single neurons in motor cortex 'code' or 'represent' may be resolved by focusing on the dynamics that generate movement.