Thinking During Sleep

Score another one for unconscious processing, which is especially prevalent during sleep. A new study in PNAS suggests that, as people sleep, their brains are forming relational memories, which require "the flexible ability to generalize across existing stores of information".

Earlier studies found that people appear better able to remember things they have just learned if they are able to sleep soon after. In effect, they found, the brain appears to use sleep time to consolidate memories.

This study suggests that the process is still more complex, and that sleep helps people make inferences from bits of knowledge that may at first appear random, said one of the authors, Dr. Jeffrey M. Ellenbogen of Brigham and Women's Hospital, the Harvard affiliate.

People make minor inferences all the time without sleeping first, of course, but those connections tend to be fairly straightforward. Sleep appears to play a role in helping people make "big picture" realizations, Dr. Ellenbogen said, like those involved in major scientific breakthroughs.

This elegant little study builds on a bevy of recent research. Last year, Ap Dijksterhuis published a paper in Science demonstrating that people often make better shopping decisions, at least when it comes to complex products, when they rely on their unconscious brain. Instead of consciously analyzing all of their options, consumers were most effective when they practiced "deliberation without attention," and let their unconscious brain digest the information while they were busy doing something else, like watching television or sleeping. Dijksterhuis summarizes the implications of his research:

Use your conscious mind to acquire all the information you need for making a decision--but don't try to analyze the information. Instead, go on holiday while your unconscious mind digests it for a day or two. Whatever your intuition then tells you is almost certainly going to be the best choice.

And then, of course, there's the work of Matthew Wilson of MIT. As I noted in a Seed article last year:

While a rat was running through a variety of different mazes, Wilson measured clusters of neurons in the hippocampus with multiple electrodes surgically implanted in its brain. As he'd hypothesized, Wilson found that each maze produced its own pattern of neural firing. To figure out how dreams relate to experience, Wilson recorded input from these same electrodes while the rats were sleeping. He was astonished by his results. Of the 45 rat dreams recorded by Wilson, 20 contained an exact replica of the maze they had run earlier that day. "During REM sleep, we could literally see these rat brains relive minutes of their previous experience," Wilson says. "It was like they were watching a movie of what they had just done."

But why would rats dream of running through a maze again? What's the advantage of replaying the activities of the day at night? According to Wilson, dreaming is a form of mental cleaning. The brain is figuring out what information it needs to keep. Since Wilson rewarded the successful rats with food, their brains were re-encoding the route, making sure they remembered how to find their way.

The moral of these studies is simple: the unconscious brain is our own personal supercomputer. While our consciousness can only process about 40 bits of information per second, the unconscious can process several million bits per second. As a result, it plays an essential role in helping us learn new information and properly assimilate it into our existing store of knowledge. It also happens to work best at night. It's ironic, but true: some of our best thinking is done while sleeping.

More like this

The purpose of dreaming is learning. While you are sleeping, your brain is digesting the day, deciding which new experiences to consolidate into long-term memory. That's the implication of Matthew Wilson's latest paper, which documented the neural activity in the brains of dreaming rats. Here's the…
There's a fascinating article in the NY Times Magazine by Sara Corbett on the publication of Jung's infamous Red Book, in which he attempted (often in vain) to map the infinite labyrinth of his unconscious: Some people feel that nobody should read the book, and some feel that everybody should read…
It's a shame that we stop encouraging naps once the preschool years are over. After all, there's a growing body of scientific evidence that the afternoon siesta is an important mental tool, which enhances productivity, learning and memory. (It's really much more effective than a cup of coffee.)…
Over the weekend, I had a little essay in the Times on some new research on why dream at night. When I can't sleep, I think about what I'm missing. I glance over at my wife and watch her eyelids flutter. I listen to the steady rhythm of her breath. I wonder if she's dreaming and, if so, what story…

You state above that "[w]hile our consciousness can only process about 40 bits of information per second, the unconscious can process several million bits per second."

Is there any evidence for this claim? What does it even mean? How can we tell that "a bit" has "been processed"?

This is an admittedly imprecise calculation, but the general comparison is valid. There is a large literature on the limitations of consciousness and working memory. You might as well begin at the beginning, and read George Miller's classic 1956 study "The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information".

Obviously, measuring the channel capacity of the unconscious is a much more difficult task. But just think of all the sensory data your brain is processing without your awareness. The world consists of lots of pixels, and those pixels don't become visual forms by accident.

As for your excellent question: How can we tell that "a bit" has "been processed"? I actually think some of the strongest evidence for this claim comes from neural recordings of the brain at work. Scientists can see the stimuli that different parts of our brain respond to. The V1, for example, is most interested in lines and angles of light. (See Hubel and Weisel). Obviously, our visual world doesn't consist of this elemental stimuli. Instead, this visual information is relayed from visual area to visual area, until the contrasting lines become a set of recongizable forms. The original bit of information - an activated photoreceptor in the retina - has been processed and processed (all of it done by the unconscious brain) in order to make our "reality".

This is an admittedly imprecise calculation, but the general comparison is valid. There is a large literature on the limitations of consciousness and working memory. You might as well begin at the beginning, and read George Miller's classic 1956 study "The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information".

Ah, yes. The Magic Number 7 (or 6, 5, 8, 9 or 4 or thereabouts.) I meant something serious.

Obviously, measuring the channel capacity of the unconscious is a much more difficult task.

As is measuring the number of angels who can dance in a given area.

But just think of all the sensory data your brain is processing without your awareness. The world consists of lots of pixels, and those pixels don't become visual forms by accident.

I doubt that the world consists of pixels and doubt that you really want to argue that it does. Generally, I'm quite aware of what I see, though I have no evidence that pixels become "visual forms". Do you?

Obviously, our visual world doesn't consist of this elemental stimuli. Instead, this visual information is relayed from visual area to visual area, until the contrasting lines become a set of recongizable forms.

I'm not aware of any evidence that such "visual forms" exist. Are you?

The original bit of information - an activated photoreceptor in the retina - has been processed and processed (all of it done by the unconscious brain) in order to make our "reality".

An activated photoreceptor is just that. It is not a "bit of information".

None of this is science. And it's not good poetry either.

Let me just cut to the chase here: Skinner proposed a selectionist account of language.

In the same way that Darwin accounted for species, Skinner accounted for language: verbal behavior is selected by the environment. Thus the only thing separating man from beast was eliminated.

Factions among the power elite were discomfited. It was too close to the truth. Chomsky was, as a result, trumpeted as the greatest thinker since Einstein, since he provided verbose arguments against Skinner. That none of his attacks were apropos was ignored. Their existence was held to be proof of their correctness.

He saved cartoon Gods. Our host is a supplicant.

I find it hard to believe that our host hasn't seen my previous posts. So i can only take his failure to respond as anything more than befuddlement. Still waiting. Tell us that Chomsky is God. You know that you want to.

We've waited 50 years for a reply. What are a few more hours? Do you have to check back with headquarters? Too busy to tend your blog? What better do you have to do? In the larger picture that is.

A couple of weeks ago, I posted some comments here on this blog about the nature of science, which I thought were fairly innocuous. To my surprise, I got quite a few hostile and sarcastic replies, which I really could not understand. My conclusion: there are alot of hostile people in the world. I put Mr.GI in this category. I don't understand his hostility; I don't understand the point he is trying to make. So far, I have noticed that Jonah makes occaisional replies, to clarify something, or to engage in interesting dialogue. Why would you expect him to reply to you, Mr.GI?

Mr. G:
You are clearly a devoted behaviorist. That's fine - I do think that cognitive psychology and cognitive neuroscience tend to underplay the seminal contributions of Pavlov and Thorndike. Modern neuroscience is very dependent on the Pavlovian ideas of the conditioned stimulus, unconditioned stimulus, etc. Look, for example, at the work of two titans of neuroscience: Eric Kandel and Joseph Ledoux. They found the neural source for the Pavlovian model. It's hard to imagine their research without Pavlov's drooling dogs.

That said, I don't understand your devotion to the Skinnerian model of language. I don't have any particular devotion to Chomsky - my favorite model of language is Terrance Deacon's version in "The Symbolic Species," which is a modified version of Chomsky's innate language module - but I've found the resulting 50 years of science, from the dissection of Broca's and Weirneke's area to the work of people like Steven Pinker, Ray Jackendoff, etc, to studies of deaf children in Nicaragua to studies of pidgin language to be thoroughly convincing. It's seem pretty clear that it's not enough for verbal behavior to be selected for by the environment of the child. The mind comes pre-equipped with innate knowledge that allows it to learn language. Recursion can't be taught.

But you certainly seem convinced that some grievious intellectual crime has been committed. Feel free to tell us why Chomsky is wrong and Skinner is right. But I insist that you keep the tone civil. I agree with Daniel that there are a lot of hostile people in this world. I don't want them on my blog.

"Earlier studies found that people appear better able to remember things they have just learned if they are able to sleep soon after. In effect, they found, the brain appears to use sleep time to consolidate memories."

I've been experimenting with polyphasic sleep and whenever I have an intense session of learning, I immediately want to take a nap afterwards to help solidify my understanding.

As far as processing while asleep goes, back in university I was once stuck on a problem for a programming task, and while I was asleep I dreamed I had discovered and implemented the solution. When I awoke I remembered the concept enough to perform it on my computer and it worked. Just my personal anecdote.

By a passerby (not verified) on 02 May 2007 #permalink

this is all making me very sleepy.

By alexander (not verified) on 02 May 2007 #permalink

quote: The mind comes pre-equipped with innate knowledge that allows it to learn language.

I would argue that's a bit imprecise. More accurate, perhaps, is "The mind comes pre-equipped with an innate structure that allows it to learn language."

Data comes into the brain with both temporal and spacial patterns. The incoming data activates cascading signals across the neurons; patterns of activation cause the involved neurons to form stronger connections, making them more likely to activate together. Re-activating these areas - such as during sleep - also strengthens them.

The brain is essentially a pattern storage and activation machine. Of course, you know this already, Jonah, I'm sure ... consider this primarily for Mr. G's edification.

[BTW - the 'magic' number 7 (+/- 2) comes from the literature. It's simply the number of distinct chunks of information that we can easily hold onto. If you're going to deride it, then you should at least look it up first, to see where it comes from and how the studies were done. Then, once you know the methodology and the definition, perhaps you can find fault from a more ... enlightened position.]

Fascinating information, particularly about the rats and their mazes. Thanks.

Cheers

Sorry. Meant to say, "The brain is essentially a pattern storage and recognition machine."

-Ted

When I was alot younger, in college, I was a very clever guy and could learn most everything that was thrown at me fairly easily, math, science, languages. Now, I am alot older. I notice that when I take a class, I have a much more difficult time learning, even though I feel about the same in other ways. I do, however, sleep much less at night.

Thanks for the clarification, Ted. You are right: referring to innate structures (as opposed to innate knowledge) is a much more precise and accurate way to describe the Chomskyan language module.

Hey, Daniel ... you should read "The Brain That Changes Itself." I'm about halfway through it myself, and there's a lot of fascinating real-world examples of people using targeted mental exercises to increase the brain's plasticity later in life. ("Plasticity" refers to the brain's ability to acquire new information, or - in a broader sense - re-organize its connections in sub-areas in order to change the way it functions in those areas. An extreme example of plasticity would be a person who goes blind, and then develops a stronger sense of hearing or touch. His visual cortex is no longer receiving signals from the optical nerves, so inputs from his other senses can sort of "take over" that space in the brain, thereby increasing the discriminatory power of those senses.)

It used to be thought that the brain becomes almost completely rigid after the formative years; however, that seems to not be the case, and this book highlights many studies where 're-mapping' has taken place.

Essentially, one school of thought is that, as you get older and your brain - as with everything else in our bodies - degrades, certain areas become too "noisy" ... that is, areas that used to work well now have too many divergent signals that interfere with the true patterns that make up cognitive abilities. These noisy signals make it hard to do the things that used to be easy. They may also accelerate the degradation, making things even worse.

However, some researchers have developed simple exercises that can re-optimize the degraded areas, vastly improving mental abilities and mental health across the board.

Anyway, that doesn't do it justice, but the book is great, and an easy read.

Cheers,

Ted

Daniel: Why would you expect him to reply to you, Mr.GI?

Because I've asked interesting questions (particularly the one about the fruits of cognitive psychology.) Questions that are still unanswered.

ted: patterns of activation cause the involved neurons to form stronger connections

Pretty efficacious, those patterns.

ted: 7 (+/- 2) comes from the literature

Though most recognize it as little more than idle party chatter, a popular factoid. (Yes, I know, Miller's paper. Ever read it?)

Jonah: It's seem pretty clear that it's not enough for verbal behavior to be selected for by the environment of the child. The mind comes pre-equipped with innate knowledge that allows it to learn language. Recursion can't be taught."

The authors of _Rethinking Innateness: A Connectionist Perspective on Development_ reach a somewhat different conclusion.

Chomsky's attack on Skinner's selectionist account of language preceded Intelligent Design's attack on Darwin's selectionist account of speciation by some decades, but the goals are the same: the preservation of comforting (and useful) superstitions from reason.

I do not know anything about Chomsky's attack on Skinner. To me, that is a somewhat obscure reference, which should probably be clarified.

Some authors of a book reached a somewhat differenct conclusion. So? What does that mean?

My last thought, so-called "intelligent-design" is speculative philosophical thinking which does not seriously challenge Darwinist evolutionary thought. If someone wants to promote a speculative point of view that is in conflict with established science, I would not call that an attack on science, since science operates with or without the approval of speculative religious and philosophical points of view. I would call these contrary points of view merely, "freedom of speech." There is no struggle between "evolution" and "intelligent design." Evolution is based on science and intelligent design is speculative philosophy based on wishful thinking. It is being overly-dramatic, to frame the speculation on intelligent design as an "attack" on evolution, as though real scientists are buckling under the pressure; they are not.

Daniel: I do not know anything about Chomsky's attack on Skinner. To me, that is a somewhat obscure reference, which should probably be clarified.

I'm not surprised that you haven't heard about it. Once behaviorism was "disproved" by Chomsky, the subject was skirted. Funding suddenly dried up. Behaviorism was pronounced dead for no valid scientific reason, but with the blessing of those who promote superstition in the furtherance of the interests of their clients. Hucksters of all stripes were relieved.

It's easy to find Chomsky's point of view. Just attend any Psychology 101 class, or Google "Skinner Chomsky". You'll hear all about the demon he defeated. The refutations of Chomsky's ill-considered claims are seldom mentioned and hard to find, but see this, e.g where it is said that

Finally, and it must be said, probably the strongest reason why no one has replied to the review is its tone. It is ungenerous to a fault; condescending, unforgiving, obtuse, and ill-humored. For example, the perfectly well-defined word "response" is consistently called a "notion" which creates, in time, an overwhelming atmosphere of dubiety with respect to the word. The review's one kind word is in a footnote (Chomsky, 1959, p. 32). It is almost impossible to reply to whatever substantive points the review might have made without at the same time sounding either defensive and apologetic, or as truculent.

Some authors of a book reached a somewhat differenct conclusion. So? What does that mean?

The link I provided was to the Amazon site, where you could read the publisher's blurb and readers reviews. Short story: Chomsky's claims about language and inateness are empirically false.

..."intelligent-design" is speculative philosophical thinking which does not seriously challenge Darwinist evolutionary thought.

If you think that a majority of Americans agreeing with ID is not a serious challenge, Im not sure what would convince you.

My point, however, was that Chomsky's arguments against Skinner's selectionist account of language are entirely analogous to ID's arguments against evolution.

I am not a psychiatrist, psychologist, or neuroscientist; that is why I find many of your references "obsure." Whatever it is you are for or against, it is above my head.

Daniel: Whatever it is you are for or against, it is above my head.

I doubt it. I assumed too much. Sorry.

Like what?

hi

By eddy l.boy ya (not verified) on 09 Oct 2007 #permalink

eternal drunken master hi im edward legacy junior im actualy able to invent anythingexpecialy when it come to video games toy music and movies there are just so many options left available but no one will listen to me i need to be help to be sertified ase as a master artist my mind is completely full of thought your spiritual leader
eddy l. boy ya

By edward legacy junior (not verified) on 09 Oct 2007 #permalink

Mr. G, I fail to see remotely where you are on-topic for this particular discussion. You seem to be almost autistically stuck on the fact that science cannot empirically prove or disprove elements of unconscious brain activity. The tone of the discussion, I believe, is one of hypothesis, hence your attack on the blogger being futile and pretty childish.