Your brain’s visual centers remain active when your eyes are closed and even when you sleep, studies have shown. But it’s a different type of activity, one not fully understood.

A new study sheds light. In both situations — resting with eyes closed or sleeping — electrical activity continues in the brain, but the activity is represented by slow electrical fluctuations, rather than the bursts of activity that occur when you’re awake with eyes wide open. The resting oscillations, as the scientists call them, were found to be most pronounced during deep sleep, as might be expected.

The slow fluctuation pattern can be compared to a computer screensaver, say the researchers at the Weizmann Institute.

Though the newfound activity’s function is unclear, the researchers have a couple ideas:

Perhaps neurons, like philosophers, must “think” in order to be; neuron survival, the idea goes, would require a constant state of activity. Or maybe the minimal level of activity enables a quick start when an outside stimulus is presented, something like a getaway car with the engine running, the researchers suggest.

These new ideas differ starkly from how scientists thought all this worked. “In the old approach, the senses are ‘turned on’ by the switch of an outside stimulus,” explained Weizmann Institute neurobiology student Yuval Nir, who worked on the study. “This is giving way to a new paradigm in which the brain is constantly active, and stimuli change and shape that activity.”

This different type of brain activity could also explain why most people don’t constantly experience hallucinations or hear voices while they rest, the researchers suggest.

Previous investigations of the resting brain typically relied on brain scans and told only part of the story. The new study used data collected from epilepsy patients who underwent extensive testing, including measurements of neuronal pulses in various parts of their brains during diagnosis and treatment.

The method suggests ways of probing the brains of young children or patients in a coma — anyone who is not alert or whose cooperation might prove challenging.

“The use of clinical data enabled us to solve a riddle of basic science in a way that would have been impossible with conventional methods,” said lead researcher Rafael Malach, a professor at the institute. “These findings could, in the future, become the basis of advanced diagnostic techniques.”

The research, funded by several institutions and foundations, is detailed in the journal Nature Neuroscience.