'Just Sleep On It' Solves Tricky Problems?

An anonymous reader writes "CBC news reports that the effectiveness of 'sleeping on it' when faced with a difficult task may have more than just anecdotal roots. 66 students were trained to perform a calculation on an eight digit number using two simple rules which would take seven steps to complete. A different method existed to perform the same calculation 'almost instantly', but was not shown to the students. After eight hours, where half the students were allowed to sleep and the other half remained awake, 60% of the rested and 22% of the wakeful students discovered the more efficient method."

Comments

This is good news

[gowen]

If nothing else, it means I've been thinking very hard indeed while at work this morning.

Telling the boss

[Anonymous Coward]

I've just forwarded this to my boss, sleeping on the job is now a good thing.

Re:Telling the boss

[gmuslera]

If your boss agrees, then that would be a truly dream job.

It's True

[CuriHP]

I know I've solved Calculus projects in my sleep before. The tricky part is trying to remember it when you wake up.

Re:It's True

[TopShelf]

That's funny, I've had the same issue with the "how to pick up gorgeous women for incredible one-night stands" problem.

Re:It's True

[Mork29]

That's easy. Quit spending all your time on /. !!!

Re:It's True

[the_mad_poster]

You've been modded funny, but there's more truth in your jest then the mods might've realized.

I'm sure most people realize it's very difficult to remember most dreams. People who say they "don't dream" are really just dreaming in deep sleeps and not waking up throughout the night. However, if you're startled awake for some reason (whether by the dream or external factors) during or shortly after a dream, your odds of remembering it shoot way up. Hence the advice to keep a notepad by the bed to write things down if you want to recall your dreams. I don't see why this wouldn't extend to dreams that may help solve problems.

One other thing I remember a professor telling us - If you're faced with a difficult problem of some sort, go do something else for awhile. Your brain will continue working out the solution while you do something else (sort of like './programming_problem &' I suppose with optimization for background processes). I do that at work all the time. I don't know if it would be more effective than sleep, but if I'm faced with a tough programming problem, I'll hit Slashdot or go take a walk. I recall working for hours upon hours once on a tough nested data structure for a custom search system. Finally, in total frustration, I got up and stormed out of the cube, went and sat in my car, and turned on a CD. After 15 or 20 minutes, I got up, came back in, sat down, and Hallelujah! I banged out the data structure and supporting code in about 20 minutes more. A few optimizations and tweaks later, and I was done. No clue where it came from. Wasn't thinking about it consciously in the car, but apparently the ol' brain was still churning and took advantage of the lack of stress from overfocus.

Re:It's True

[CrayzyJ]

True Story. I was stuck on a bug during my undergraduate work. In the middle of the night, my girlfriend tells me, I jumped out of bed, clicked away on the keyboard and climbed back in bed. The next morning, I found the solution to my problem, albeit ill-typed, on my screen.

After this occurred, I decided to look into it. Experts suggest purposely thinking of a tough problem *right* before you fall asleep. Your subconsious is a) much smarter than you conscious and b) never stops working. Giving it a job to do will result it working on it all night. Try it. If you remember your dreams, the results are cool.

This is why mental breaks, as the parent mentioned, work. Let your subconscious do the hard work. It's much better at it than your slow, cloudy, easily distracted, conscious thought. Another example, have you ever forgot something important, and then out of the blue while doing something else you remember? Your SC was working on the job the whole time. Ultra-cool, IMHO.

Re:It's True

[Communomancer]

Your subconcious is not much smarter than your concious. Your concious is not "slow", "cloudy", or any more "easily distracted" than your subconcious is. If you think that your subcon isn't easily distracted, think about when your dreams have gone from somewhat sensical to utterly non-sensical in one bewildering instant.

The only thing your subconcious has going for it is that it doesn't have to process the terrabytes of data that the outside world hurls at your concious every second. So yeah, you have a few spare cycles you wouldn't otherwise have. But don't mistake it for "superiority".

Re:It's True

[thelenm]

I've thought that I've solved calculus (and other) problems in my sleep before. Once I even forced myself awake so I could write down on paper whatever brilliant thing had come to me in my dream. When I woke up in the morning, the writing was mostly unreadable, but the readable words said something utterly stupid, like "half of pigs rolling will unite the values". Maybe dreaming is sort of like being drunk, in that you perceive yourself as much smarter than you really are.

Re:It's True

[skintigh2]

Being startled awake might work, but that doesn't happen very often.

If you stay in the exact position you were sleeping in, and slowly transition from sleeping to waking, you can often remember details more clearly.

Or, you can arrange to be startled awake. Supposedly Salvador Dali would hold a spoon in his hand while dozing off, and the spoon would drop and wake him up just after falling asleep (and entering REM) and thus inspired his paintings of melting clocks (and spoons).

Well established

[Space cowboy]

I thought it was pretty well-established that sleep plays a role in post-analysis of the days accumulatd information ? There have been too-many-to-count articles on the subject in New Scientist / Scientific American ...

There must be an evolutionary advantage to having a time when nothing else was going on to do something, and what else apart from the days events could occupy a brain if it has no external sensory input... I seriously doubt all the higher life-forms on the planet would do it if there wasn't a good reason....

Simon
Simon

Re:Well established

[jorleif]

Not using as much energy when asleep and rebuilding muscle and other tissue are probably also a factors, but perhaps orthogonal ones. It is imaginable that there could exist a lifeform which would rest (and thus save energy, rebuild) without putting its mind in a sleeping state.

Saving energy cannot possibly be the whole reason, because in that case you could compensate for lack of sleep by eating more, and you can to a point but after 48 hours or so of waking time you usually notice that it's not so much the lack of energy but the lack of ability to concentrate. So intuitively it would seem that the mind needs to do something that demands it to be in "sleep-mode".

Re:Well established

[Mwongozi]

Many aquatic animals do not sleep in the same way that we do. Dolphins have the ability to shut off parts of their brain when they're not using them, although for what reason I'm not entirely sure.

My CPU can do that too...

Re:Well established

[Beardydog]

I'd always heard you burn more energy sleeping than you do when you're awake, but relaxed (watching tv, etc) which isn't all the surprising if the brain is busy rebuilding itself and working on the problems of the day.

IIRC, going 11 days without sleep tends to kill people, so it's definitely doing things vital to your health.

Re:Well established

[Beardydog]

Explains dolphins a bit too. Dolphins don't get to sleep in caves and holes like surface mammals, they need to surface to breath, and holding still inthe middle of the ocean too long with your senses shut down makes you extremely edible

But with the ability to shut down one piece at a time, they can let sections rebuild without having to shut down the whole thing.

Like keeping a base system on one of my drives, so if one goes down, I can use the other to fix it without having to futz around with disks.

Re:Well established

[prockcore]

Dolphins have the ability to shut off parts of their brain when they're not using them

Humans have that same ability during election years.

The Great Brain

[Speare]

This was a common strategy used by the fictional middle brother in the series of books called "The Great Brain" by JD Fitzgerald. It tracked the deals and schemes of a wily kid in early 1900s Utah, as seen by his awestruck little brother. He'd think on his hardest mental problems just before going to bed, and would usually awake with an inspiration.

I often employ the same strategy, with mixed results, but it's better than not coming up with any ideas at all.

What kind of sleep?

[Gilesx]

It'd be interesting to see what type of sleep these students had. I regularly take 20 minute naps that leave me refreshed and able to better handle problems. Can I assume that traditional / deep sleep is better than light sleep / napping?

And what about induced sleep through alcohol or medication? Could it be beneficial to have the ability to "sleep on demand" to solve a tough problem?

Re:What kind of sleep?

[SamSim]

Some people have tried sleeping 20-minutes naps every four hours, and nothing else. Basically, by doing this, you trick the brain into getting all its REM sleep done right away instead of spending two hours sinking into it and two hours rising out in the morning. It's called Uberman's sleep schedule and people who've tried it seem to love it.

I'm hoping to try it myself over Easter.

Re:What kind of sleep?

[lawpoop]

I don't know about internal timers, but here's a story from my exchange in Ecuador.

For a Spanish immersion program, I stayed with a host family in Quito. About 20 miles from the equator. There, there is no dawn or dusk. At 6 AM, broad daylight. At 6 PM, night.

We had to get up early to catch a bus to take us to school. We were supposed to bring an alarm clock, but I didn't. The first day, I thought, "oh well, I'll just wake up tomorrow and see what happens." (There was no jet lag because we were travelling N/S). So, with my east-facing window, the sun woke me up the next morning. I got up, left my room, went in to the kitchen, and checked the clock on the microwave. It said, "6:21".

I never needed an alarm clock during the whole stay. The sun always woke me up at 6:21 or 6:22.

Re:What kind of sleep?

[mce]

Indeed. Experiments have shown that both types of sleep are needed. During the deep sleep phases, the brain processes the facts of the previous day and classified the things you have learned, thereby making sure they will be remembered. During REM sleep, it makes new mental connections between things it already knows.

Or to put it very simplistically: the deep sleep phase makes you smarter, the REM phase sleep makes you wiser. Your brain needs both for you to function properly on the longer term.

Re:What kind of sleep?

[poot_rootbeer]

Word of advice to anyone thinking about giving the Uberman sleep schedule a chance: check what actual experts think about it first. I recommend sleep researcher Dr. James Maas's book "Power Sleep".

Never take lifestyle advice from some guy who wrote an Everything2 article at face value. Do your research. Dr. Maas has.

Re:What kind of sleep?

[menscher]

I tried something similar... I slept for 2 hours every 12. So, 4 hours of each 24-hour period. It was pretty amazing. I was always alert. Never had the problem of being tired at the end of the day. And it works well with a schedule, since you can go to work in the day, take a nap, then work all night, take a nap, and be ready the next day.

Only one problem. After 2 weeks I realized my body wasn't keeping up with my brain. Even though I could think through things quickly (which is fine if you have a desk job) my body seemed to be physically deteriorating. So I went back to the "normal" sleep schedule.

Still, I'd recommend this if you ever have a "crunch time", like when an important project is due, or possible for finals week, or something.

Einstein

[pubjames]

I remember reading somewhere that Einstein would sleep for 14 hours or more at a time. And Margret Thatcher could get by on just a couple of hours.

Re:Einstein

[adrianbaugh]

Bush probably never sleeps at all ;-)

taking a shower works too

[Cederic]

I find I solve a lot of bugs in the shower. Or while out buying lunch. Or anywhere that my brain is not engaged in the current task, but where that current task is something other than the bug I'm trying to fix.

It's almost letting your subconscious thought processes work on the problem instead of trying to tackle it directly.

The upshot is that I feel no shame in saying "I'm not going to fix that bug today. I'll fix it tomorrow" when I'm stumped on something. Or a tricky design problem, etc - works for most problem solving situations.

Of course, this is all anecdotal..
~Cederic

Re:taking a shower works too

[kinema]

It's almost letting your subconscious thought processes work on the problem instead of trying to tackle it directly.

Isn't this kind of the basis of Zen? Letting your mind relax and revert to a almost child like state where you are "open" to most anything. They call this "the beginner's mind".

Re:taking a shower works too

[tinrobot]

The shower works for me, too...

I've requested my company install one in my cubicle.

Good to know

[nob]

I'm going to post something +5 Insightful, but I have to take a nap first. Check back later.

Isn't this somewhat obvious?

[RailGunner]

Isn't this a somewhat obvious result of the study? It makes sense that people who are well rested and refreshed from a good night sleep (or a timely nap) would be more alert and better able to think on their feet.

Anyone who has ever crammed all night for a final knows how your brain seems to turn to mush after the test is over from the fatigue of it all.

Who knows, maybe now that it's been scientifically proven, businesses will realize that people are actually more productive when not forced to work ridiculous amounts of mandatory, unpaid overtime.

Re:Isn't this somewhat obvious?

[1u3hr]

Isn't this a somewhat obvious result of the study? It makes sense that people who are well rested and refreshed from a good night sleep (or a timely nap) would be more alert and better able to think on their feet.

RTFA. They also tested a group that had slept, but not been exposed to the problem, and none of them solved it. Thinking about it before sleeping gave the best result.

The story as reported by BBC news

[cjellibebi]

http://news.bbc.co.uk/2/hi/health/3418017.stm

You wanna know what sucks?

[tuxette]

I'll tell ya what sucks. Waking up in the middle of the night with all the world's greatest ideas and solutions to all the world's problems, and 1) not having a pencil and some paper nearby to write these ideas down, 2) being too lazy/groggy to get up and look for a pencil and paper, and 3) falling back asleep only to have forgotten everything when the alarm goes off.

What's worse is when you actually do have pencils and paper nearby and you manage to scribble something down at 3 am., it's either illegible or utterly surreal at 8 am.

Re:You wanna know what sucks?

[Mandomania]

True story:

When I was in high school, I had a dream that I was in this ancient-looking stone room. There were books all over the room, some piled in stacks and some just strewn about. There was a desk in the far corner, with a bunch of candles on it and a large book in the middle.

Somehow, I knew that this was Merlin's study (I was in my King Arthur/Holy Grail/Knights Templar phase) and that the book held the answers to every question and secret I would ever have. I walked up to the desk and opened the book.

The writing was incredibly blurry; almost like I was trying to read it without glasses. I squinted to try and clarify things, but it only made it worse. Then, I remember thinking "Hey, my eyes are closed! I just need to open them and I'll be able to read it.".

So I opened my eyes and found myself staring at the ceiling of my bedroom.

That was teh suck.

Memory seems to work like this too

[Peter Cooper]

The other day I had to remember a name from ten years ago. I could picture the person but no name was forthcoming. Several hours later, while doing something quite menial, and not actively thinking about earlier, the name just suddenly appeared.

A couple of friends remarked that this was quite common for them, but I'd never really thought of it before. It seems some dark area of your brain remembers tasks you're trying to achieve, or things you're trying to remember, and sets about working on them in the background, while you get on with something else entirely.

This may be why people often come up with great ideas in the shower or while driving in the car, as their minds were 'set the task' earlier, and finally it's finished. Not too unlike a computer I guess, but certainly cool when you do it yourself. You realize that brain has a lot more tricks up its sleeve than are mentioned in the handbook.

Re:Memory seems to work like this too

[lithiumcloud]

You got a handbook? Bastard. I didn't even get a lousy pancreas manual.

A useful method

[HGWang]

This method was employed by many creative people over the years. A famous case is the discovery of the molecular structure of benzene by Friedrich August von Kekule after he had a dream about snakes whirling. He famously said "Let us learn to dream, gentlemen." I like to use this technique myself to boost creativity.

The method used to sleep actively on top is to slacken by using hypnosis, meditation, progressive relaxation or any other method which you know (a simple manner with the breath deeply several times all as affirming ). Now you speak the spirit without knowledge and ask for your spirit without knowledge to provide a solution during the night to you problem. Now let go from all the concern and go to sleep. It can be not also easy in the beginning to make this but with the practice you can become an expert with it.

Have a paper and a pencil with range of the hand so that you can write your thoughts and solutions with your problem as soon as you awake. Just continue to practise this and each time you have success by solving your problems by the sleep on top you will amplify your self-esteem and will increase the probability of success the next time.

I hate this...

[snatchitup]

Because of this, I now realize why all my "Natalie Portman Squirming Nude in Hot Grits" dreams abruptly end with a Light Bulb Over her head, and just below that, a Do Loop, an equation, or the name of some obscure (.jar) file that has a utility for the previous day's problem.

Yes, this works

[heironymouscoward]

This happens so often with me it's almost a standard procedure. If I'm working late trying to fix a problem, debug a difficult issue, or find that really elegant solution to a tricky problem, I leave it until the next morning. Almost every time, the solution is then obvious, clear, and works immediately.

Most likely it's because the unconcious mind needs space to work, and concentrating on the issue is counter-productive. Someone once wrote a nice article about why it helps to be stupid when you want to play football, because _thinking_ is not what you want to do when you're standing in front of the goal with an open shot.

Similarly in more intellectual challenges, the subconcious mind does a large part of the work but needs to be left alone to do its thing.

There are other ways to get the same effect:

- playing music while working
- going for a brisk walk (not heavy sport, because that tires you out)
- smoking a joint (depends on the person but for many people this does the trick)
- playing a game (solitaire?)

But sleeping is definitely the best way, probably because the brain is designed to do exactly this.

Incidentally, it works for social problems too. Having trouble with a colleague? Sleep on it, they say.

Re:Rubbish.

[worst_name_ever]

I think we can all learn a lesson from Dilbert:

"I've been working for five days without any sleep to finish this report. At first I had a mental block. But on the fourth day I was visited by an Incan monkey god who told me what to write. Now I just have to find somebody who can translate his simple but beautiful language."

Time does gently what you can't do by force

[FearUncertaintyDoubt]

I'm lazy. And I hate washing dishes. So what do I do with my dishes? I let them soak. You can replace a lot of scrubbing if you just let the dishes sit overnight in water.

I also like to let my brain do the same things with problems. You can sweat and fret over some technical or even personal problem, and hack out a solution that seems like it's the best you have (and yet seems inadequate), or you just let it soak for a day or two or even a week. You never force it completely out of your mind, and occasionally bring it back to mull actively, but not too much more than just running your hands over the surface of it. One of those times, you feel a flash of inspiration, or a depth of understanding that wasn't there before, and that's when you close in for the kill. Your mind gets it.

Anyway, that's the best I can do to use words to describe the proess I use to think and act creatively. I have found that my mind works in similar ways with regard to learning new things, such as juggling or a foreign language. I might practice my juggling for a week, and not see much progress, then take a week or two off. When I pick back up, lo, I've made a significant improvement.

Re:Rubbish.

[gowen]

Furthermore a single test with just 60 people is not enough to create a meaningful statistical evaluation of the experiment.

Says who?

Comparing two population proportions:
n1=n2=30
p1 = 0.6
p2 = 0.22

Null Hypothesis: Population proportions equal
Pooled proption = 0.41; standard deviation = sqrt(0.41 * 0.59) = 0.49

Z statistic = (p1-p2) / (sigma * sqrt(1/n1+1/n2)) =2.99

p-value = 0.0014.

That seems pretty significant to me. Go to the top of the class, and jump off.

Re:Rubbish.

[HeghmoH]

Let me see if I get this straight.

Sixty people in a controlled study is not enough to be 'meaningful'.

Yet a bunch of anecdotes coming from you and some of your coworkers is significant? Bizarre.

I have debugged in my sleep

[John Harrison]

In college I had a subtle bug in my solitaire game that was due the next morning. Finally I gave up and hoped that the grader wouldn't notice it. I went to sleep and woke up at 4am with the solution clearly in my head. I sat down at the computer and in 10 minutes the program worked flawlessly. This wasn't a one-line fix. It was something that pervaded most of the code. I was shocked that I had solved a problem this complex while sleeping.

I have also played some excellent games of Tetris in my sleep, but that doesn't seem nearly as interesting.

Best idea behind sleep I've heard

[Eldonv]

I've taken several classes on sleep, learning and how the brain works. It's really fascinating. However, the best reason I've heard for why sleep is necessary is to reprogram your brain. (It's not to recover the body at all; the body can function just fine, 24 hours a day, indefinitely.) If you think about it, a computer can run forever without needing to rest, but the brain is a lot different from a computer, it needs to process the activities of the day, needs to create new connections between memories and new pathways for various mental activities. Sleep allows the brain time to do this. Without periodic sleep the brain functions less and less effectively. (Think Windows with too many applications running.) If a person goes without sleep for too long the entire thing crashes. (Studies have shown that people can go without water longer than they can go without any type of sleep.)

Perhaps we're not too different from computers after all.

Coleridge was NOT asleep!

[Mark_in_Brazil]

Quoth the article:

Samuel Taylor Coleridge was inspired to write the epic poem Kubla Khan while asleep

Uh, no.

Coleridge was in an opium-induced stupor when he got the inspiration for the poem. Here are some sources that back this up (including comments from the poet himself):
You can read about the poem and its origins here, or you can read original notes on the poem from the author and others who knew him here. You can also read the original poem here.

--Mark

Records clerk of the brain

[Aexia]

The way I had it explained to me as a kid was that it's like asking the records clerk for your mind a question.

If you keep *trying* to remember something, it's like you keep calling the guy back to the counter and otherwise pestering him such that he can't actually do the thing you're asking of him.

But if you're patient and let him work back there, he'll find the answer. Usually.

The actual "problem" they "solved".

[Snags]

In Nature, volume 427 (2004), p352, they describe the actual problem given to the test subjects.

Take an 8-digit string formed from the numbers 1, 4, and 9. A series of two-digit comparisons is done. The result of the comparison is the same digit if they are the same, or the "missing" digit if they are different. That is, 1 1 -> 1, while 1 4 -> 9.

Start by comparing the first two digits, and from then on, compare your current result with the next digit in the string. Their example is 11449494, which leads to the results 1, 9, 1, 4, 4, 1, 9. The last result is the final answer.

The trick is that the original strings were "generated in such a way that the ... second [result] coincided with the final solution." People who found this trick were deemed to have gained insight into the pattern.

I think the study is bogus because of this. Sure, some people will notice the pattern, but careful people might choose to carry out the full calculation anyway, just to make sure. Any given string could follow the pattern or not. What they're demonstrating is how easily people can be tricked into finding patterns that may or may not be there. This kind of learning leads to racial profiling --- the result of the easy observation (race) implies the result of the more difficult one (criminality).

It has to do with Protein synthesis

[cjmckenzie]

It's not just a matter of relaxation, although that does have psychological effects. However, during REM sleep (when dreaming occurs) the brain synthesizes proteins that form long term memory. Long and short term memory are actually physiologically different. Short term memory are synapses created throughout the day on the fly, long term memory is created during REM sleep or during times when you may be zoning out (this is controversial). If finding innovative solutions can be statistically broken down as propotional to the amount of nueral connections involved in the processing of the question in the solution space, then it would make sense that people that sleep in REM sleep, which has a much higher level of brain activity, would be more apt to finding the innovative solution. However, it's nice to see that there are numbers that back this up.

Fuck it.

[LiberalApplication]

That's another corporate-work-place strategy for dealing with problems, and now it's been proved effective by actual scientists performing actual research!

"In the experiments conducted by Wagner and his colleagues, volunteers tackled arithmetic problems and then took twenty minutes to masturbate furiously while being lectured on mathematics. A second group was instructed to fornicate with a set of oversized-novelty-foam-numeral-character figures. Those who reached orgasm were twice as likely to realize that there was a hidden rule that substantially simplified calculations."

So the next time you can't figure out how to solve a problem, just fuck it!

PS: I refuse to cite the source of this quotation.

Tetris Experiment

[RealRav]

I remember reading about an experiment a couple of years ago. A group of people who had never played Tetris before were asked to do so, in a controlled environment, every day for a period of time. The people who had vivid dreams about the game showed a marked improvement where the others did not.

I believe that dreaming is a way of working through our problems and possibly indexing our memories.

Dreams are better as dreams than reality.