While you were sleeping: the dreaming and waking brain

• General anaesthetics do more than put you to sleep
• Podcast: General anesthetics and consciousness
• What sleeping animals can tell us about our sleep

Associate Professor Bruno van Swinderen studies sleep and consciousness. Here, he answers questions about dreaming, sleep and consciousness. 

Dreaming

Q: Are there any biological similarities between daydreaming and dreaming at night? Are the same connections happening in our brains?

A: I do believe serious research is being done on that. It would be great if we could accomplish some sleep functions while we are awake, and there is some research at QBI on this right now. Maybe that is what happens during meditation. Many people say that meditation is important for their wellbeing, making them feel refreshed, and it could be accomplishing deep-sleep functions. On the other hand, daydreaming could be accomplishing some REM sleep functions, which might involve emotional learning.

One aspect is of great interest to me: are sleep and wake really all that different, in terms of brain functions and processes? We think they are very different because when we wake up we have a conscious experience of changing states. But maybe the brain has been doing something very similar all along: blocking out the world, just a little bit more during sleep. Maybe we have been blinded by our fixation on our own consciousness, which is not the end-all of what a brain does.

Q: For infant animals (such as humans) prior to birth, I wonder if what is showing up as dreaming is actually the brain is writing instincts – an initial formatting of brain data with basic skills such as self-preservation, communication, emotions, and so forth.

A: With regard to emotions, think of what emotions do for you. They are a way to attach value to prediction errors – the signal that happens if a response is different to what your brain expected to happen, based on your prior experiences. That allows the brain to make new memories and prevent that error in the future. The first source of all prediction errors and internal models isn't the visual world or human interactions – it's all about motor learning in the womb: Where does my body begin and end? What is a consequence of my actions (sucking my thumb) or another's actions (mother rolling over)? I think this may be why infants 'dream' more – they are making those models of the world and assigning value to them. I think this is sort of what you are getting at as well, with 'instincts'.

Q: How did sleep come to be and evolve? Do you think it's kind of a big defence mechanism, almost like your brain can only handle so much consciousness and fatigue before it needs to shut down for a bit and rebuild? Or is has it evolved from species that were better at replaying their experiences when sleeping (REM sleep) to become efficient at responding to the environment?

A: I think that key to it is what we mean by ‘efficient’. What I think we mean is that we learn better while awake, so we don't make dysfunctional mistakes in the future. I think this is where REM sleep plays an important role with regard to emotional responses.

One idea I've been excited about recently is that dreaming keeps us surprised in the real world, so that we can learn better, because emotion associated with surprise consolidates our memories better.

When something unpredictable happens in reality, it is surprising to us, and this is often associated with a ‘prediction error signal’ (when the an outcome is unexpected, based on our experiences) and an emotional response. The emotion ensures that a new memory is formed so we can make better predictions. The waking brain basically wants to become a prediction machine.

The price to pay for this – there is always a price – is that the waking brain will become less and less surprised in that context. This erodes the very system that allowed that learning to happen. My idea is that this is where the 'alternate realities' of REM sleep step in, to curate the brain's valence (emotional) system, to ensure that the waking brain will always respond with surprise, so it can keep on learning effectively.

The function of dreaming, could be to basically keep us conscious so that we can pay attention and learn better. If we didn't dream, we would become maladaptive zombies.

The functions of deep sleep are becoming better understood, in comparison: clearing waste from the brain, maintaining synapses, moving memories around: basic janitorial stuff. Dreaming is different.

Q: Why do we sometimes not remember our dreams?

A: We don't remember most of our dreams – especially the ones at the beginning of the night after your first deep sleep cycle that you just slept through. I bet those were pretty interesting, but they are lost forever. One important thing to keep in mind about dreaming is that it probably has less to do with the narrative content you remember than the neural activity (electrical connections/signals of brain cells) that is happening at the time, which is probably maintaining some crucial function in the brain, like emotional balance.

By emotional balance, I mean the valence systems in the brain (e.g., neurotransmitters like dopamine) firing at just the right level to help consolidate a memory. It's known that neurotransmitters such as dopamine are important for memory consolidation. If those systems are out of whack, then you will not learn quite as well. But maybe if you learn too well, then you wouldn’t need those systems anymore, which would be maladaptive in a way. This is where I think REM sleep steps in, to make sure we remain surprised and keep on responding with the right amount of emotion. That's what I mean by emotional balance.

Q: When my son was less than three months old he was dreaming and bust out in pure joyous laughter at such a young tender age. I’ve raised three kids and never heard something so pure and raw come from a young infant. How would it be possible for an infant to dream and laugh? What could be funny to them at that age or was it more instinctual?

A: Laughing is the quintessential prediction error: funny jokes always have an expected twist, and it is a way for us to communicate that we have been surprised in a good, non-harmful way. Your child had a dream that evoked that, and he communicated that to you in his sleep. They do it all the time while awake; the easiest way of making my kids laugh was doing something harmlessly unpredictable, like putting a teapot on my head.

Sleep

Q: When I'm about to fall asleep, I think about the fact that I'm falling asleep, which takes me a few steps back into a more conscious state and I can’t sleep. Why does this happen?

A: Paying attention kills sleep. If you can just be aware of it without paying attention, you fall asleep.

Q: Why is it so important to sleep for so long?

A: That's one of the big mysteries. Functions of sleep have to be done in real time. The fact that a fly needs as much sleep as a human suggests that it has nothing to do with brain size or intelligence, and is all about conserved evolutionary functions. We are studying those functions.

Q: Can you explain what happens to the brain through the process of sleep deprivation? Would everyone eventually go crazy if they were not allowed to sleep? Would they die at some point or just pass out?

A: Rats have been shown to die of sleep deprivation before they die of starvation. I’m not exactly sure what they die of, and that says a lot about how little we understand the various functions of sleep. Cognition suffers first, but then cellular and stress pathways go off. This probably relates to both REM and deep sleep functions – you can't do away with them. You'd die.

Q: Why is it not possible to replace sleeping by taking a pill with all the hormones that sleep gives you, and maybe resting a bit? Does the brain really need to rest? Or is there something else that sleep gives us that can't be consumed by a pill?

A:  Sleep is all about alternating between stages, and pills usually tackle just one stage, such as deep sleep, without dealing with the other one. That's why these pills (mostly GABAergic drugs) never really quite work. Sleep isn't just one thing, serving only one function. Maybe we'll be able to recap the entire process when we know how it works, by using a slow-delivery cocktail of sorts rather than just one pill.

Consciousness

Q: How do you define consciousness?

A: For me consciousness is paying attention to your memories. Just like you can pay attention to representations of the outside world, you can pay attention to internal representations, which then trigger narrative sequences that are influenced by our history. Does it have to be more complicated than that? Finding those representations is the hard part.

Q: How does what we expect to perceive change how we direct attention? Is it more influenced by your internal model of the world or the sensory input the brain receives?

A: Great question. It probably evolves through life, starting with mostly sensory input and moving towards models as we gain experience. Some models are already there from the start. Is it a coincidence that developmental history of REM sleep shows this trend as well, changing through life? Maybe this is a clue as to what REM sleep does: curating our models of the world, and how we respond to prediction errors.

Q: What's the characteristic that separate us (consciously) from animals? What is it that we have, or can do, that they can't? For instance, a lot of animals can communicate but we seem to be the only animal that can tell stories.

A: I'd say the characteristic that separates you from animals is the same one that separates you from other humans: you have a different history. Your human history is one that taught you a language so you can tell us about your consciousness. Your personal history is a unique one only accessible to you, which no-one else in the world has. This is what makes you the conscious entity you are today, which is different than me and different than a dog or a fly. But why should the underlying mechanisms (e.g., selective attention and memory) be any different? You just have more to play with than a fly.

Q: Are we in control of ourselves, or are we just powerlessly observing the results of neuron circuits? For example I feel like I came up with a thought, but was it maybe just a result of another thought?

A: You are talking about free will, I think. I think free will is history. Both in the sense that it's an out-of-date view, and also in the sense that our history forms our choices. Even if consciousness is just an aftereffect of decisions of neurons, the emotional response associated with that will affect the next neural decision, so it's better to view it as a sequence through time, driven by accumulated history, rather than a magical thing that happens at any one moment. In that sense, it seems obvious that animals would have something similar.

Read more Q&A on this reddit AMA.