Dreaming....

[Cite134]

My first post in the Science section!
Anyways...

For a while, I've been fascinated with dreaming, and I used to believe that the particular events that occur in dreams had some profound meaning ( I don't anymore). Moreover, I've briefly looked up dreaming and there are different theories on the intrinsic function of it, such as: activation synthesis, or continual-activation. From what I know, I don't think that there is only one theory agreed upon.

MY question(s) is....what do you guys think about the subject? Why do we dream? Is there some type of biological benefit?  Is it the brain's way of maintaining some form of conciousness while we sleep?

[Thumpalumpacus]

I think it's the file-clean up preceding a defrag, done to maintain sufficient system memory.

[Asmodean]

Wonder what my PC dreams about when I /cleanup it...  

But I do agree in a way. I think it's a side effect of the brain doing a sort of self-analysis.

[The Black Jester]

My brother is a neuroscientist that specializes in sleep research, I should get him to post here.  In my discussions with him, he stresses the enormous amount that is still unknown regarding the purpose of sleep functions, but he speculates that part of it has to do with memory consolidation - specifically the pruning of weaker associative connections that are likely to be unused.  

As I understand it, there is a great deal of research indicating that the brain, during waking hours, accumulates far more information, in the form of memories, than it can feasibly store in the long term, and it must decide what to consolidate into long-term memory, and what to dispense with.  No one is yet sure how this specifically relates to dreaming, per se, but that is one of the known functions of sleep.  This is partly why your performance on memory and concentration tasks will be impaired without regular intervals of rest.

[Tank]

^^^ interesting. If your brother does have the time to contribute it would be much appreciated.

[The Black Jester]

^^^ interesting. If your brother does have the time to contribute it would be much appreciated.

I'm trying to bait him into posting....

[KebertX]

My brother is a neuroscientist that specializes in sleep research, I should get him to post here.  In my discussions with him, he stresses the enormous amount that is still unknown regarding the purpose of sleep functions, but he speculates that part of it has to do with memory consolidation - specifically the pruning of weaker associative connections that are likely to be unused.  

As I understand it, there is a great deal of research indicating that the brain, during waking hours, accumulates far more information, in the form of memories, than it can feasibly store in the long term, and it must decide what to consolidate into long-term memory, and what to dispense with.  No one is yet sure how this specifically relates to dreaming, per se, but that is one of the known functions of sleep.  This is partly why your performance on memory and concentration tasks will be impaired without regular intervals of rest.

Yes, most psychologists tend to believe it is a function of memory consolidation during REM Sleep.  When the brain is activating and deactivating different sections of the hippocampus to clean up our memory, we experience the activated memories as dreams.  The mind puts all the little things from our day (things you didn't know you were remembering, your mind picked up every little sight and sound from the day) into a coherent story of some sort, as a sort of test if the memory is functioning, and therefore, keepable.  That's also why we don't usually remember our dreams.  People dream 80% of all nights, but only remember about 15% of their dreams.  This lends support to the Memory Consolidation Theory.  But, when people are woken from REM sleep, they do remember what their dream was.

Fun Fact: Your dreams aren't chronological events.  Studies have shown that no matter how long a subject is in REM sleep, they always wake up remembering dreams of the same length and detail.  That's because after all the memories are 'tested,' and activated, the entire dream has been put together.  Experiencing a dream as a chronological story is an illusion, really, the whole dream is already put together from the start, and you are already experiencing the whole thing!  Have you ever been woken up from a dream before it ended, and been mad at the one who woke you up because you wanted to know how it ended?  If so, that was wrong, you HAD dreamed the dream to the end, from the very beginning.

I used to think dreams had meaning too, until I took Psych 2201.  I once had a very very long dream, that detailed me, being diagnosed with cancer, and slowly withering away (It doesn't mean anything when you die in your dreams.)  When I woke up from it I realized 2 things.  ONE: Even though I had perceived the event as taking place over a few days, it had only been 3 hours since I fell asleep.  TWO: Right before I fell asleep, there had been a commercial on for a Cancer Treatment Center.  I also realized I had an odd lump on the back of my head, maybe I should get that checked out...

[Thumpalumpacus]

I also realized I had an odd lump on the back of my head, maybe I should get that checked out...

... or at least avoid dark alleys ...

[Cite134]

My brother is a neuroscientist that specializes in sleep research, I should get him to post here.  In my discussions with him, he stresses the enormous amount that is still unknown regarding the purpose of sleep functions, but he speculates that part of it has to do with memory consolidation - specifically the pruning of weaker associative connections that are likely to be unused.  

As I understand it, there is a great deal of research indicating that the brain, during waking hours, accumulates far more information, in the form of memories, than it can feasibly store in the long term, and it must decide what to consolidate into long-term memory, and what to dispense with.  No one is yet sure how this specifically relates to dreaming, per se, but that is one of the known functions of sleep.  This is partly why your performance on memory and concentration tasks will be impaired without regular intervals of rest.

Interesting...I hope your brother does post though! . Yea, recently, I've been remembering quite a few dreams lately, I just wanted to know the function of it. I guess the brain consolidating memory makes sense. The brain is an amazing organ!

[pinkocommie]

My brother is a neuroscientist that specializes in sleep research, I should get him to post here.  In my discussions with him, he stresses the enormous amount that is still unknown regarding the purpose of sleep functions, but he speculates that part of it has to do with memory consolidation - specifically the pruning of weaker associative connections that are likely to be unused.  

As I understand it, there is a great deal of research indicating that the brain, during waking hours, accumulates far more information, in the form of memories, than it can feasibly store in the long term, and it must decide what to consolidate into long-term memory, and what to dispense with.  No one is yet sure how this specifically relates to dreaming, per se, but that is one of the known functions of sleep.  This is partly why your performance on memory and concentration tasks will be impaired without regular intervals of rest.

So he's suggesting that dreams are the product of the brain essentially defragging?  That's awesome!

[KDbeads]

MY question(s) is....what do you guys think about the subject? Why do we dream? Is there some type of biological benefit?  Is it the brain's way of maintaining some form of conciousness while we sleep?

So this is NORMAL dreaming, right..... not sleep disorder dreaming, like say night terrors or sleep paralysis?  

Cause my view on normal dreaming is pretty much what Thumpalumpacus said.  The other stuff.... well let's not go there.....

[LittleTheBlackJester]

Hello, I am the aforementioned brother of The Black Jester who posted on Dreaming.  He asked me to give my thoughts on the matter, so I will do my best.  First, I'd like to give a spoiler.  I am, indeed, a neuroscientist, and I do have a background in human sleep research.  But, I am not an expert in dreaming, per se.  That is another section of the sleep field, one which has a lot of controversy with it.  There are, as some have mentioned already, a lot of theories about the "purpose" of dreams.  They range from dreams as being an epiphenomena to being relevant to memory functions to relevant to emotional regulation to relevant to consciousness.  

Before I talk about my understanding of dreams as a sleep neuroscientist, I feel there are a few things I must clarify first.  One, dreams are very heterogeneous.  There are many kinds and many forms.  This may seem pretty obvious, but it needs to be stated.  Dreams can vary from brief transient thoughts during sleep to full fledged active dreams to experiencing passive scenes.  Many have strayed away from the word dreams towards 'sleep mentation', basically thoughts you have while asleep.  These thoughts can take shape as full fledged dreams, but also as passive thoughts.  What kinds of mentation you experience really depends on what time of  night and what stage of sleep you are in.  That brings me to the second point I want to make.  Dreams is not equal to REM sleep.  Not all REM sleep is full of dreams and not all dreams come from REM sleep.  They are related, to be sure, and dreams in REM sleep are the most active most intense and most memorable.  But there is more in heaven and earth than is dreamt up in REM sleep.  he he.  Sorry, I have a bad sense of humor sometimes.  

I would not say that dreams function to prune connections or to 'defrag'.  Nor would I say that they are principally functionless,  or that they are principally associated with memory consolidation...exactly.  The role of REM sleep in memory consolidation is a bit mixed at the moment.  Evidence is emerging and some much older work from the 60s support such a role, but there is still a lot of controversy.  I will get back to that later.  

Perhaps what I should start with is what I do know about dreams.  It has been established that dreams are affected by daytime experience.  One person posted their dream was meaningless because it was about cancer and cancer was in an ad they saw before bed.  There is a study that was conducted at Harvard by Bob Stickgold back in the late 90s.  He had a bunch of people play tetris for the first time (really hard to recruit for by the way, as most people had already played tetris).  He then looked at dream reports.  This all took place in the first 3 minutes of sleep, so this was definitely not REM sleep related dreams.  What they reported was seeing blocks moving down and manipulating them.  They didn't see the computer or the room they performed the testing in.  They just saw the pieces falling down.  Some of you may have experienced this at one time or another when you play a video game for the first time.  I know I have.  So we know that at least some dreams are affected by daytime experience.  The same experiment was carried out by waking people up from REM sleep.  The dream content was a lot more associative and a lot more disconnected from the tetris, but it was related.  So REM was affected by daytime experience as well.  Some people may think this means that dreams are meaningless.  I would argue just the opposite.  The same group, just recently published a paper showing that those who learned a 3d maze, much like those seen in first person shooters, if they would dream about it at all at night, they would perform significantly better the next day.  Those that did not dream of it, did not.  I haven't looked over the paper super closely, so I don't know if this was independent of sleep stage or dependent on sleep stage.  This is important, because if it was correlated with a sleep stage, like REM, for example, one could argue dreaming didn't cause the improvement, just the physiology of REM did.  If you spend more time in REM you are also more likely to dream about it.  So it would be more association than causation.  Whereas if it was not, whatever processes going on in the brain during 'dreaming' would then be necessary to improve navigation on this task.  But how would dreaming do this?  That just leads to speculation, but sleep and REM sleep in particular has been associated with more than just memory consolidation.  In fact, the relationships between sleep and memory is very complicated.  Probably because sleep isn't just about memory, and memories themselves are very complicated.

Lets look at memory as a whole for starters.  We have many different kinds of memories.  We have episodic memories: our experiences throughout the day, memories of space and time and association.  These are particularly hippocampally dependent.  And these memories are aided by sleep.  But the story isn't one of REM.  It is actually more of a NREM story, and there is a mountain of evidence in support of this now in animals and in humans.  Specifically, oscillations in NREM sleep, slow waves and sleep spindles, are associated with facilitating consolidation and encoding of declarative, hippocampally dependent memories.  These oscillations have been associated with being linked to hippocampal-neocortical crosstalk, and this cross-talk is thought to transform memories from being hippocampally dependent to being cortically dependent.  That is one theory.  Another theory is that slow oscillations globally downscale synapses, pruning away unnecessary ones and preserving strong ones.  The relative weights of synapses are preserved and signal to noise ratio is imporved since you knocked out all the weakest ones.  This is more of a memory side-benefit of a process more designed to manage the memory demands of the brain.  These two theories are not incompatible, and there is a lot of evidence to support both.  And both implicate NREM sleep, not REM sleep in this process.  

So REM isn't really as associated with hippocampal-dependent consolidation much anymore, though it was in the 60s with all the rat experiments.  The trick is, those experiments often had an emotional component or a reward component.  This makes it more like conditioning, which relies on a different neural network and might be more REM-dependent.  There is evidence in humans now that emotional memories are tied to REM sleep and not NREM sleep.  Additionally, other more procedural forms of memory that are less dependent on the hippocampus, such as learning to ride a bike or type on a keyboard or hit a baseball, are dependent on either NREM sleep or a combination of NREM and REM sleep.  So different forms of learning, which require different neural systems, will rely on different physiological brain states in sleep.  And, more specifically, they will all rely on different oscillations occuring in each of these sleep stages.  So, in short, it's complicated.  To complicate it further, sleep isn't just about memory, but is also involved with neurometabolic, immune, and hormonal regulation.  In the memory domain, sleep isn't just about consolidation.  Sleep also facilitates next day encoding ability, integration of memories, and abstraction of general knowledge from memories which facilitates insight into problems the next day.

Much of these latter abilities (integration, abstraction, emotional processing) are tied to REM in some manner.  Which makes me think that REM isn't important for weakening associations between unnecessary items, but instead would strengthen more weakly associated items that prove helpful.

But this isn't dreams, this is sleep.  And, as I have said before, REM does not equal dreaming.  There are a lot of theories about dreaming out there.  Some more psychoanalytic, some more memory related, some more emotion regulation, and some more 'its all garbage, and epiphenomenon with no meaning'.  I really don't think it is the latter.  I just don't think the evidence out there supports it (only some of which I have talked about in this post).  Dreaming is associated with eye movements, which are associated with brain activity in areas associated with mental imagery.  Dreaming is affected by previous waking experience and predicts next day behavior.  Dreaming, or at least REM sleep dreaming, has been associated with emotion regulation, consolidating emotional memories, and facilitating problem solving the next day.  In this way, psychoanalytic, memory, and emotion regulation theories could all be correct in some fashion.  In the end, all these functions serve to enhance adaptability, which is very important from an evolutionary perspective.  

My current advisor (I am a post doc right now) has a theory about REM sleep.  He calls it the 'sleep to remember, sleep to forget' hypothesis.  He is primarily a memory researcher, and he likes to think of emotional memories as containing multiple forms of information.  You have the visceral-physical, emotional response.  But you also have the details of the experience.  In the end, what is most adaptive is to actually remember the emotional experience, even if it was traumatic.  What you want to do is instead forget the emotional, viscerl, physical response, so that every time you think on the memory you don't regurgitate the emotional intensity you did the first time.  He posits REM sleep does just this by stripping away the emotional component and strengthening the memory component.  REM is then important to deconstruct experiences and extract meaning from them to remember what is most important.  Or perhaps to learn what is most important.  There is some evidence for this in that sleep will enhance specific components of a memory and not all of them.  Emotional pieces of a scene are more remembered than their neutral counterparts.  So sleep isn't like justice, it prioritizes based on what may be most adaptive.

My boss posits that this breaks down, or fails to happen in patients with PTSD.  You can see this in that their REM dreams are less associative and weird, and more like direct replays of the trauma.  He thinks there is some process in sleep that is failing to do its job.  When done correctly, the emotional side would be separated and forgotten, whereas the knowledge one could learn is consolidated and integrated into your general knowledge base.  Take, as an example, you walk down a new path and get bitten by a snake.  You want to remember that this path is bad in the future.  You don't want to shiver and sweat and feel stressed and scarred everytime you think of that path.  That is the essence of sleep to remember, sleep to forget.  

Completely separate from memory, sleep, and particularly REM sleep, is associated with modulating your reactivity to emotional events.  As the day wears on, you find things more negative and angry, and less happy.  If you are provided a nap, this bias goes away, so sleep is important for more than just memory.  Sleep, instead, seems to be important for regulating a lot of the functions in your body from immune, to metabolic, to cognitive and emotive (not just memory in the cognitive domain either).  So I don't know if I answered any questions or created any, but I did ramble on a bit.  I think that, more than anything, proves I am an academic.

[The Black Jester]

Sigh.  I love my brother....

[Cite134]

Hello, I am the aforementioned brother of The Black Jester who posted on Dreaming.  He asked me to give my thoughts on the matter, so I will do my best.  First, I'd like to give a spoiler.  I am, indeed, a neuroscientist, and I do have a background in human sleep research.  But, I am not an expert in dreaming, per se.  That is another section of the sleep field, one which has a lot of controversy with it.  There are, as some have mentioned already, a lot of theories about the "purpose" of dreams.  They range from dreams as being an epiphenomena to being relevant to memory functions to relevant to emotional regulation to relevant to consciousness.  

Before I talk about my understanding of dreams as a sleep neuroscientist, I feel there are a few things I must clarify first.  One, dreams are very heterogeneous.  There are many kinds and many forms.  This may seem pretty obvious, but it needs to be stated.  Dreams can vary from brief transient thoughts during sleep to full fledged active dreams to experiencing passive scenes.  Many have strayed away from the word dreams towards 'sleep mentation', basically thoughts you have while asleep.  These thoughts can take shape as full fledged dreams, but also as passive thoughts.  What kinds of mentation you experience really depends on what time of  night and what stage of sleep you are in.  That brings me to the second point I want to make.  Dreams is not equal to REM sleep.  Not all REM sleep is full of dreams and not all dreams come from REM sleep.  They are related, to be sure, and dreams in REM sleep are the most active most intense and most memorable.  But there is more in heaven and earth than is dreamt up in REM sleep.  he he.  Sorry, I have a bad sense of humor sometimes.  

I would not say that dreams function to prune connections or to 'defrag'.  Nor would I say that they are principally functionless,  or that they are principally associated with memory consolidation...exactly.  The role of REM sleep in memory consolidation is a bit mixed at the moment.  Evidence is emerging and some much older work from the 60s support such a role, but there is still a lot of controversy.  I will get back to that later.  

Perhaps what I should start with is what I do know about dreams.  It has been established that dreams are affected by daytime experience.  One person posted their dream was meaningless because it was about cancer and cancer was in an ad they saw before bed.  There is a study that was conducted at Harvard by Bob Stickgold back in the late 90s.  He had a bunch of people play tetris for the first time (really hard to recruit for by the way, as most people had already played tetris).  He then looked at dream reports.  This all took place in the first 3 minutes of sleep, so this was definitely not REM sleep related dreams.  What they reported was seeing blocks moving down and manipulating them.  They didn't see the computer or the room they performed the testing in.  They just saw the pieces falling down.  Some of you may have experienced this at one time or another when you play a video game for the first time.  I know I have.  So we know that at least some dreams are affected by daytime experience.  The same experiment was carried out by waking people up from REM sleep.  The dream content was a lot more associative and a lot more disconnected from the tetris, but it was related.  So REM was affected by daytime experience as well.  Some people may think this means that dreams are meaningless.  I would argue just the opposite.  The same group, just recently published a paper showing that those who learned a 3d maze, much like those seen in first person shooters, if they would dream about it at all at night, they would perform significantly better the next day.  Those that did not dream of it, did not.  I haven't looked over the paper super closely, so I don't know if this was independent of sleep stage or dependent on sleep stage.  This is important, because if it was correlated with a sleep stage, like REM, for example, one could argue dreaming didn't cause the improvement, just the physiology of REM did.  If you spend more time in REM you are also more likely to dream about it.  So it would be more association than causation.  Whereas if it was not, whatever processes going on in the brain during 'dreaming' would then be necessary to improve navigation on this task.  But how would dreaming do this?  That just leads to speculation, but sleep and REM sleep in particular has been associated with more than just memory consolidation.  In fact, the relationships between sleep and memory is very complicated.  Probably because sleep isn't just about memory, and memories themselves are very complicated.

Lets look at memory as a whole for starters.  We have many different kinds of memories.  We have episodic memories: our experiences throughout the day, memories of space and time and association.  These are particularly hippocampally dependent.  And these memories are aided by sleep.  But the story isn't one of REM.  It is actually more of a NREM story, and there is a mountain of evidence in support of this now in animals and in humans.  Specifically, oscillations in NREM sleep, slow waves and sleep spindles, are associated with facilitating consolidation and encoding of declarative, hippocampally dependent memories.  These oscillations have been associated with being linked to hippocampal-neocortical crosstalk, and this cross-talk is thought to transform memories from being hippocampally dependent to being cortically dependent.  That is one theory.  Another theory is that slow oscillations globally downscale synapses, pruning away unnecessary ones and preserving strong ones.  The relative weights of synapses are preserved and signal to noise ratio is imporved since you knocked out all the weakest ones.  This is more of a memory side-benefit of a process more designed to manage the memory demands of the brain.  These two theories are not incompatible, and there is a lot of evidence to support both.  And both implicate NREM sleep, not REM sleep in this process.  

So REM isn't really as associated with hippocampal-dependent consolidation much anymore, though it was in the 60s with all the rat experiments.  The trick is, those experiments often had an emotional component or a reward component.  This makes it more like conditioning, which relies on a different neural network and might be more REM-dependent.  There is evidence in humans now that emotional memories are tied to REM sleep and not NREM sleep.  Additionally, other more procedural forms of memory that are less dependent on the hippocampus, such as learning to ride a bike or type on a keyboard or hit a baseball, are dependent on either NREM sleep or a combination of NREM and REM sleep.  So different forms of learning, which require different neural systems, will rely on different physiological brain states in sleep.  And, more specifically, they will all rely on different oscillations occuring in each of these sleep stages.  So, in short, it's complicated.  To complicate it further, sleep isn't just about memory, but is also involved with neurometabolic, immune, and hormonal regulation.  In the memory domain, sleep isn't just about consolidation.  Sleep also facilitates next day encoding ability, integration of memories, and abstraction of general knowledge from memories which facilitates insight into problems the next day.

Much of these latter abilities (integration, abstraction, emotional processing) are tied to REM in some manner.  Which makes me think that REM isn't important for weakening associations between unnecessary items, but instead would strengthen more weakly associated items that prove helpful.

But this isn't dreams, this is sleep.  And, as I have said before, REM does not equal dreaming.  There are a lot of theories about dreaming out there.  Some more psychoanalytic, some more memory related, some more emotion regulation, and some more 'its all garbage, and epiphenomenon with no meaning'.  I really don't think it is the latter.  I just don't think the evidence out there supports it (only some of which I have talked about in this post).  Dreaming is associated with eye movements, which are associated with brain activity in areas associated with mental imagery.  Dreaming is affected by previous waking experience and predicts next day behavior.  Dreaming, or at least REM sleep dreaming, has been associated with emotion regulation, consolidating emotional memories, and facilitating problem solving the next day.  In this way, psychoanalytic, memory, and emotion regulation theories could all be correct in some fashion.  In the end, all these functions serve to enhance adaptability, which is very important from an evolutionary perspective.  

My current advisor (I am a post doc right now) has a theory about REM sleep.  He calls it the 'sleep to remember, sleep to forget' hypothesis.  He is primarily a memory researcher, and he likes to think of emotional memories as containing multiple forms of information.  You have the visceral-physical, emotional response.  But you also have the details of the experience.  In the end, what is most adaptive is to actually remember the emotional experience, even if it was traumatic.  What you want to do is instead forget the emotional, viscerl, physical response, so that every time you think on the memory you don't regurgitate the emotional intensity you did the first time.  He posits REM sleep does just this by stripping away the emotional component and strengthening the memory component.  REM is then important to deconstruct experiences and extract meaning from them to remember what is most important.  Or perhaps to learn what is most important.  There is some evidence for this in that sleep will enhance specific components of a memory and not all of them.  Emotional pieces of a scene are more remembered than their neutral counterparts.  So sleep isn't like justice, it prioritizes based on what may be most adaptive.

My boss posits that this breaks down, or fails to happen in patients with PTSD.  You can see this in that their REM dreams are less associative and weird, and more like direct replays of the trauma.  He thinks there is some process in sleep that is failing to do its job.  When done correctly, the emotional side would be separated and forgotten, whereas the knowledge one could learn is consolidated and integrated into your general knowledge base.  Take, as an example, you walk down a new path and get bitten by a snake.  You want to remember that this path is bad in the future.  You don't want to shiver and sweat and feel stressed and scarred everytime you think of that path.  That is the essence of sleep to remember, sleep to forget.  

Completely separate from memory, sleep, and particularly REM sleep, is associated with modulating your reactivity to emotional events.  As the day wears on, you find things more negative and angry, and less happy.  If you are provided a nap, this bias goes away, so sleep is important for more than just memory.  Sleep, instead, seems to be important for regulating a lot of the functions in your body from immune, to metabolic, to cognitive and emotive (not just memory in the cognitive domain either).  So I don't know if I answered any questions or created any, but I did ramble on a bit.  I think that, more than anything, proves I am an academic.

Wow . Insightful as well as informative. Thank you for that sir.

[Thumpalumpacus]

to transform memories from being hippocampally dependent to being cortically dependent.

I have no clue what this signifies.  Might you either explicate, or direct me to a reputable source?