Jaan Aru is a researcher at the Computational Neuroscience Research Group and at the Talis Bachmann Lab at the University of Tartu.
Anyone who suddenly gets an urge to attack a scientific problem that is interesting, intriguing, and ‘big’ might end up with the question ‘why do we dream?’. This question seems so simple, but as it stands we still do not have a good answer for it. As expected, this topic has attracted the attention of many scientists and thinkers, so that there are already several theories around about dreams and their functions.
As a great writer, Sigmund Freud managed to popularize the view that dreams represent unfulfilled wishes that creep up from our subconscious. Freud believed that there are special processes which try to disguise the unfulfilled wishes before they emerge in dreams. This would explain why our dreams are sometimes bizarre. However, in general these ideas of Freud do not have much credibility when the contents of dreams are subject to scientific analysis.
On the other end of the spectrum, Allan Hobson has argued that dreams do not represent anything at all – they just reflect the random activity patterns of the brain. This theory is probably also too extreme, as dreams clearly have a structure in space and time and the dream content often corresponds to the everyday activities of the dreamer.
What happens in our dreams? A painting by Franz Schrotzberg: “Sweet dreams dreaming of snow white and the seven dwarves”. Image credit: Wikimedia Commons
The fact that dreams are related to everyday occupations already suggests that during dreams daily events are processed. And a recent theory proposes that dreams are the way they are because they directly reflect the consolidation of memories (scientists use the complicated word ‘consolidation’ to describe the process which stabilizes memory traces in your brain). According to this theory, dreams provide a direct window to the memory consolidation process. If this were true, it would be pretty cool – while dreaming you actually perceive and feel your brain doing the hard work of storing your memories!
This theory is based on the observation that in the rat brain neural activity patterns that were elicited during learning in the awake state are replayed during sleep. This finding is quite interesting by itself: Your brain replays some events during sleep, although in a somewhat different manner than encountered during daily activities. Clever scientists have even shown that if external stimuli are used to elicit such replay processes, one can boost memory! But let us leave the intriguing questions regarding if and how we can enhance memory during sleep for another time. What matters to our current quest towards understanding our dreams is that such replaying of the learned material during sleep might be directly reflected in our dreams.
“Wait a second!” you might say, because if you have ever noticed or written up your dreams, then you know that one only seldom dreams about the exact events that happened during that particular day. So dreams cannot just simply reflect memory consolidation, can they? Should the dreams correspond one to one to the daily activities to bear any role in memory consolidation?
The fact is, dreams mostly do not match daily activities directly. Researchers have tried for decades to manipulate dream content and know how difficult it is – dreams are not a simple replay of what happened during the day. For example, even if subjects play a really engaging game of Tetris directly before going to sleep, they do not necessarily dream about playing Tetris, but they might rather perceive some elements of the game together with other content, such as Tetris bricks falling from apple trees!
But this is by no means a problem for the theory that dreams reflect the process of memory consolidation. On the contrary, this is what makes this theory elegant and interesting. Our current understanding of memory tells us that memories are stored in the brain in the connectivity patterns of millions of neurons. And this view implies that if new memories are stored, then these memories will use some of the neural connections that are also part of older memories. Therefore, it is to be expected that during the consolidation of new memories parts of old memories are reactivated. Hence, if dreams reflect memory consolidation, it is not surprising that the new memory contents are mixed with fragments of old memories. For instance, the Tetris players reported sometimes that they dreamed about playing the old version of Tetris and not the fancier version the researchers had used.
Thus, the fact that dreams never exactly replay events from daily life but rather combine them with other events actually supports the theory that dreams reflect memory consolidation. New memories are stored in the networks that already carry older memories, and the partial reactivation of them is responsible for our unique dream experiences that involve elements experienced in different epochs of our life.
So, it is a very beautiful theory about what dreams are. Yes, scientists also value the beauty of ideas – here, it would be really dreamy if dreams would give us direct insight into the memory consolidation process. It would not only be interesting for dream researchers, but even more for memory researchers, as it would give direct introspective clues about how our neural machinery is consolidating memories. However, scientists should value evidence even more than beauty, and, unfortunately, this beautiful idea does not have much direct evidence to support it. For now.
Thus, I myself would not yet take the bet that dreams indeed reflect the memory consolidation process, but this idea is intriguing and demonstrates that we still have a lot to learn about dreams and from dreams.
This example also illustrates that although scientists seem to be dull people wearing white coats while discussing in front of their blackboards, they are actually dreamers. They work on problems that penetrate the minds of all (Why do we sleep? Why do we dream? How can consciousness emerge from the brain? Just to name a few questions from my own field) and suggest all kinds of romantic solutions to them. For a moment or two, or sometimes for their entire lifetimes, these researchers think that they have indeed solved these grand problems – “Sleep is for consolidation of memories!”, “Dreams directly reflect the process of memory consolidation!”, “Consciousness is integrated information!”.
This is a great feeling – this feeling of having understood something is what drives scientists. But of course sometimes this feeling is illusory and the scientists are simply wrong. Other scientists will then quickly point out the mistakes and flaws in these theories, whose creators fiercely defend their theories as if they were their babies. All this belongs to the process of science – when we face such grand problems, we will come up with many wrong ideas, and we fight over them, sometimes an entire lifetime, but this is simply part of the game. If all fundamental questions could be solved with a snap of the fingers it would be a really boring universe.
