Watching how snails pace their eating may help explain how humans alter behaviour, such as walking faster or using longer strides, scientists say.
Researchers studied Aplysia marine snails
The Mount Sinai School of Medicine in New York team looked at how nerve cells dictate movement.
They say understanding how this process works could aid spinal injury care.
The study, in Current Biology, found "basic" neurons control the details of movements, but higher level neurons affect factors such as speed.
Most systems which control movement in animals and humans are organised into a hierarchy of at least two layers of neurons.
In humans, the "basic level" neurons are found in the spinal cord and the "higher level" in other areas of the nervous system, such as the brain stem.
Messages from the brain tell "pattern generator" circuits of basic neurons within the spinal cord to perform the required task, intervening as necessary to ensure precise control.
But scientists did not understand the details of how variations of acts, such as eating or walking, were generated.
The researchers studied a kind of marine snail called Aplysia to see how variations in its feeding behaviour were produced, focusing on how the snails bite when they sense food is near.
It was found the biting action was controlled by "basic level" neurons, but two higher-order neurons worked together to dictate how fast the snails performed the movement.
The researcher leading the study, Dr Jian Jing, told the BBC News website the team's findings could be used to help people with spinal injuries.
Experts are already looking at extracting signals from the cerebral cortex - which controls functions such as movement, vision and hearing - to see if they can control prosthetic devices in patients who are paralysed because their brains are disconnected from the spinal cord.
Dr Jing said: "Our work suggests that an alternative approach is to use these higher order signals to directly stimulate the spinal 'basic level' neurons in various combinations to generate a variety of behaviours.
"Of course, this requires more studies on how the 'basic level' neurons are organised in the spinal cord.
"The study may also help robotics researchers to design more intelligent robots that utilise a hierarchical controller where elements at different levels implement different functions."
John Cavanagh, head of research at the International Spinal Research Trust, said: "This work does have some relevance to understanding repair and control within the damaged spinal cord, although direct clinical application is a long way off.
"It shows that even apparently simple systems do have an amazing complexity of interaction between the cells involved, but that this is highly organised."
He added: "This kind of research involving intact systems of non-mammalian species also shows how much basic knowledge can be derived without the use of more controversial experiments in animals."