The video is beautiful. An octopus, dozing in a tank, continuously changes color like a tentacled kaleidoscope. While it’s impossible to prove, the snoozing octopus is most likely dreaming (what does an octopus dream about?). But whether octopus or human, dreams have one thing in common: their purpose is unclear.

In an extensive 2012 literature review, the psychologist Matthew Merced notes that, even though nobody knows for certain why we dream, advances in the technology and techniques of brain research have at least helped explain how we dream. Humans, at least, dream a lot, multiple times a night, and the brain is very active during dream periods. Dreaming must be important, even if it remains mysterious.

Early dreaming studies were, frankly, pretty primitive. Researchers would wait for the Rapid Eye Movement (REM) sleep to begin, then wake the subject up and ask about any dreams. Dreaming occurs during non-REM (NREM) sleep as well, but those dreams tend to be less vivid. Now, thanks to PET scans, it is known that large areas of the brain, covering such functions as motor control and sensory processing, all become active during dreams. Chemical changes occur as well, especially during REM: acetylcholine, a neurotransmitter that fires up the brain and forces muscles to contract, ramps up production. Unsurprisingly, areas of the brain controlling awareness and consciousness remain dormant.

Taken together, these chemical changes seem to stimulate a condition similar to wakefulness, but without any of the trappings of being awake. Images in dreams come entirely from the brain: visual sensory input is shut down; even open eyes will not process images during dreaming. Parts of the brain that control emotions are running full steam, perhaps explaining why dreams can be full of strong emotions, like terror or joy or sadness. Rational parts of the brain remain dormant, so insane events are accepted without question.

But back again to why. Is it possible there is no why? Many researchers think that dreams are simply the product of chemical changes in the brain during sleep. In this view, dreams are evolutionary hitchhikers: REM sleep is beneficial, and dreams tag along with it. The fact that we mostly forget our dreams lends some support to this view: Memory formation systems are mostly turned off during dreaming.

Assuming there is a purpose, natural selection suggests that dreaming must provide some sort of survival benefit. Why spend energy on involuntary movements and brain activation if nothing is being achieved? One possibility is that dreams are kind of a virtual reality world, a space where humans can safely practice coping with threats (being chased, for example, is pretty common). REM sleep and dreaming may also help process important or traumatic memories. Finally, there is a theory that dreams are a way for the brain to rid itself of information it isn’t using, a sort of “psychic disk cleanup.” According to Merced, any of these explanations could be correct. Sweet dreams to all, humans and octopuses alike!