Can We Control Our Dreams?

Photo by Илья Мельниченко on Unsplash

While we’re awake, we have conscious control over what we see, hear, touch, smell, and taste. In the world of dreams, we have an illusion of that control — a facsimile of it. In our dreams, we are simultaneously constructing our reality and controlling our actions within it. (If this doesn’t seem meta, then I don’t know what does.) Yet, despite the seemingly godlike control we have over our dreams, how come we struggle to shape it to our whims? I’m sure most of you have had the experience of waking from a dream to realize, “that didn’t make sense at all,” or “how could I have not realized so many things that didn’t make sense?” In essence, we don’t have full voluntary control of our dreams.

But what if we can control our dreams? In the famous fictional anime Sword Art Online, the NerveGear helmet can stimulate all five senses through the user’s sleeping brain so that the user can play a virtual reality massively multiplayer online role-playing game (VRMMORPG). In the show, the game is released in 2022. Of course, in reality, we would not have such a game by then even if we want to. But nevertheless, is dream manipulation, even on a smaller scale, still possible?

Researchers at MIT Media Lab’s Fluid Interfaces already figured out how to somewhat manipulate dreams using a technique called targeted dream incubation (TDI), which takes advantage of a sleep stage called hypnagogia, a transitional state between waking and sleeping states(Weisberger, 2020). Targeted memory activation makes use of sensory stimuli to reactivate memories during sleep. A sleep tracker, Dormio, communicates with an app that delivers audio cues to the subject at specific times to influence the content of the subject’s dreams. The researchers found that 67% of the dream reports reported dreams that integrate the prompted memory, so it seems possible that dreams can be influenced by some external manipulation using outside stimuli (Horowitz et. al, 2020).

If we ever want to have voluntary control of our dreams, then it’s useful to know what is being activated in our brains during sleep. The first sense we often think about when it comes to dreams is vision. Visual events are what we usually remember from our dreams. Surprisingly, the primary visual cortex is inactive during dreams. However, the visual association cortex that processes imagery associations is active and involved in the perception of dreams. There’s also activity in the temporal regions involved in facial recognition, auditory processing, and episodic recall (Hoss, 2013).

It’s important to note that dreams are not all visual, even though we don’t often remember the non-visual portions of our dreams. It is thought that all dreams experienced during REM are characteristically visual; however, other senses are employed as well, such as auditory sensations, feelings of movement (even though in REM sleep voluntary muscles remain paralyzed) , and to some extent even tactile sensory experiences (Dang-Vu, 2007). A person who is born blind does not have any visual dreams at all, but their dreams are representative of what they can experience in real life, which may still include sound, touch, smell, and emotion.

The “dream space” is formed via associations between emotions, memories, and conceptualizations. During dreaming, episodic memory is excluded, which is why we don’t replay events during our dreams. However, the limbic system along with the hippocampus remains highly active (this is not the case in non-REM sleep) which incorporates emotion into our dreams. Heightened activity in the limbic system is required for memory consolidation in the hippocampal regions and for emotional processing in the amygdala and hypothalamus. The high activity in these regions is a testament to the selective processing of emotionally charged memories and events of our dreams.

We hear about REM and non-REM (NREM) sleep a lot, but it is important to make some distinctions. There are multiple stages of sleep. Stage 1 consists of non-REM sleep, whereby we transition from wakefulness to sleep. Stage 2 is also characterized by non-REM sleep and is a transitive period between light and deep sleep. Stage 3 involves non-REM sleep as well, a period called “deep sleep”. Finally, the last stage is REM sleep which occurs approximately 90 minutes after falling asleep. Scientists have concluded since the 1950s that rapid eye movement (REM) sleep is the neurophysiological condition that underpins dreaming. When participants are awakened from REM sleep in laboratory experiments, dreams are recorded between 80% and nearly 100% of the time. These are usually the most detailed. Dreams have been found to be less common during NREM cycles, which is mostly composed of more thought-like sentiments, and a repetition of the previous day’s events.

Our vivid dreams — at least the dreams we seem to remember — occur during REM sleep. At this point in sleep, the brain undergoes unique emotional processing and makes connections in memory systems that make our dreams appear meaningful. In REM sleep, brain regions that are active, such as are those that are active in unconscious mental processing or the moments right before we are conscious of our actions. For this reason, dreams can be thought of as a representation of what is considered to dwell in our unconscious mind: feelings, moods, and attitudes that we are unable to control or access directly. During REM, brain areas such as the temporal lobe in combination with the cerebellum and pre-motor/sensory areas are active, resulting in dream-like sensory fusions, all of which are internally generated through mental representations.

Yet according to research, it is suggested that perhaps dreaming and REM sleep are not necessarily linked to one another (Solms, 2000). This challenges some of the previously accepted theories that dominated neuroscience for quite some time. It has been discovered that unique activity in the forebrain along with various other brain centers, such as the limbic system (hippocampus and amygdala), have an effect on how we dream. Furthermore, the irrationality of our dreams stems from a deactivated dorsolateral prefrontal cortex during REM sleep as well as the visual associations in the visual association cortex along with the right inferior parietal cortex. Interestingly enough, these regions are considered to be responsible for lucid dreaming. That’s probably why dreams don’t make a ton of sense to us once we wake up.

In a hybrid state known as lucid dreaming, whereby the dreamer becomes highly conscious of their dream, the prefrontal cortex (involved in decision making and normally inactive during REM sleep) becomes activated and EEG experiments show that this state is similar to a waking state. PET scans revealed that when the medial frontal cortex (involved in consciousness) was activated, there was a greater sense of control over the dream (lucidity), and when the amygdala (involved in emotional processing) was active, there was a greater sense of the dream being out of control. For example, Stephen LaBerge (1981) found ways for lucid dreamers to interact with researchers from the lucid state through shifting their eyes or flexing their muscles in programmed shapes, proving that conscious control exists.

Another interesting phenomenon involves the process of storing dreams. How are we able to recall the dreams that we dreamed? Increased activation in the temporoparietal junction (TPJ) can stimulate attention inclining towards external stimuli, allowing dreams to be encoded in memory. Along with the medial prefrontal cortex (MPFC), these regions are involved in dream recollection during wakefulness and outputting dreams and encoding them during sleep. Can you imagine a future where you can store your dream experiences in some sort of technology and continue dreaming it the next time you go to bed?

What is the future for dream control given what we already know about dreams? Indeed, sleep-controlling neurotechnology has a long way to go. There’s still a lot that we don’t know about the dreaming brain, so pushing into the boundaries of sleep science would only happen one step at a time. Neurotechnology startups are already working on bidirectional brain-computer interfaces that not only attempt to read and monitor the sleeping brain, but also try to manipulate it. However, we’re still far off from that possibility of voluntary and controlled dream manipulation that some of us dream of. More work still needs to be done both in sleep research and in neurotechnology development.

This article was written by Czarinah Micah Rodriguez, who is a senior undergraduate student at UC Berkeley studying Cognitive Science, Psychology, and Molecular & Cellular Biology, and Milan Filo, who is a senior undergraduate student at UC Berkeley studying Molecular & Cellular Biology. This article was edited by Abraham Niu, a junior undergraduate student at UC Berkeley studying Cognitive Science and Data Science.


Eichenlaub, J. B., Nicolas, A., Daltrozzo, J., Redouté, J., Costes, N., & Ruby, P. (2014). Resting brain activity varies with dream recall frequency between subjects. Neuropsychopharmacology, 39(7), 1594–1602.

Hoss, R. (2013). The neuropsychology of dreaming: Studies and observations. Published online by author.

Solms M (2000). Dreaming and REM sleep are controlled by different brain mechanisms. Behav Brain Sci 23: 843–850.

Weisberger, M. (2020, September 25). Dream-shaping tech from MIT channels suggestions into your dreams. Livescience.Com.

Haar Horowitz, A., Cunningham, T. J., Maes, P., & Stickgold, R. (2020). Dormio: A targeted dream incubation device. Consciousness and Cognition, 83, 102938.

We write on psychology, ethics, neuroscience, and the newest in neural engineering. @UC Berkeley

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store