Contribute to Something Great

[DoubleHelix]

Hey guys,

Something great is on the horizon, and you can be a part of it. The idea has been discussed for a long time, but amazingly, there is no research to show whether or not it is possible. I am talking about learning while you sleep. It is possible.

More specifically, I should say that your brain can receive the effects of practice while you sleep. In recent research, sounds were paired with a spacial task. Those same sounds were replayed to experimental subjects during slow-wave sleep (deep sleep; stages 3 and 4). They remembered the spacial orientation of the objects in the task significantly better than than the control group, who slept just as long but did not have the sounds replayed to them during sleep.

The first thing that crossed my mind when I saw this was, "this needs to be done with the olfactory sense - if sounds work, smell will definitely work." Very recently, the same research team did just that. The task was different than in the first experiment, but sure enough, the repetition during slow-wave sleep of the smell administered during the task led to increased performance of the task on the after-sleep trial.

This is truly amazing stuff. The implications are huge. Soon, I will be volunteering to assist a neuroscientist specializing in sleep and memory. This is not my area of expertise (yet), so I will be learning the ropes of sleep research while voraciously studying the brain and memory. It should not be too long before I propose my own design and we carry out the experiment that I'm dying to do.

We can use these findings as a springboard to figuring out how to enhance our vocabularies while we sleep. Someone WILL succeed in this endeavor, and it won't be long until that happens. I will succeed, I think, but I may not be first. That doesn't matter. What matters is designing an experiment just right - so the effects of the sleep-practice will be most visible. The following are some notes on such an experiment:

(I do not have access to masks and gases, so I will work with sound.)

1)What should the sound be? Is playing a recording of someone saying the actual word the best option? Perhaps, if the word is an obscure synonym for "dog," then a dog barking might be better than the word being read. If the word being read sounds like the best option, would it be more effective if the subject, while studying the word, read it aloud into a tape recorder, so his own voice could be replayed while he slept?

2a)What words should we choose? They must be so obscure that not a single one of the subjects knows even one of them. Thus, a foreign language is an option. However, a foreign language that one is unfamiliar with can appear so unfamiliar upon first sight that it may be very hard to learn. English is probably better, because the appearance of the words, and the phonemes within them, strike a familiar tone. If a subject wouldn't learn the words well after two exposures, (the sleep-practice is after all merely a second exposure, and it is probably weaker than an exposure while awake), then both the control and the experimental group will have very low scores, and thus there will be little variability in the scores. With so little variability, the results wouldn't be significant. Here, the number of words comes into play. If there are only ten words, there is little room to prove that real improvement has occurred because all of the subjects will learn a very high percentage of them. Thus, the balance of the number of new words, as well as the difficulty of learning those words, is of paramount importance.

2b) My best suggestion for balancing these factors would be to have a very large amount of words that could each correspond to a one-word definition. Any single word would thus be easy to learn, and a one-word definition is so simple that a second-exposure would certainly increase the amount of words recalled. However, the sheer amount of words ensures some amount of error on the part of the subject.

3) The words should be nouns accompanied by a picture. Nouns are easiest for a child to learn. They are most basic, and correspond most directly to photographs.

4) Since learning novel vocabulary is so hard after a single session, we must employ techniques such as the above to increase recall. However, if methods unrelated to repeatable sound are used, such as accompanying pictures, they should not be so effective that the effects of the sleep-practice are overshadowed. Ideally, the most effective technique used for remembering the words should be the technique that involves a repeatable sound. That would make the practice most valuable.

5) The best idea I have to accomplish this is to accompany each word with a relevant, comedic video with a terse sentence describing the word. The content of the sentence should be memorable, and I stress that the sentence must be very short - able to be uttered within 3 seconds maximum. For instance, I could film myself just bawling, tears streaming down my face, while the word on his current flashcard is "lachrymose," and I would just say in a loud, yet strained, lugubrious tone, "I'm so lachrymose!!!" Maybe then he could write on a piece of paper, "Lachrymose means weepy or crying," and possibly write a sentence of his own. Maybe that 5-second movie could be shown before and after he writes his sentences.


Notice I have a hell of a lot of ideas, but not enough knowledge to tell which ones are good. You don't need the apposite knowledge to offer any ideas either. I am going to conduct this experiment after learning the ropes and writing a lit review; this is not just a flight of fancy. Anything you say will be taken into serious consideration and combined with what I learn. Brainstorming has always been useful for me, so do your worst.

Thanks.

 

[zelo1954]

My suspicion is that by attempting to learn during sleep we will negate the value of the sleep. But, there again, I always preferred the black hat in those silly corporate games.

 

[Ism]

My thinking is that obscure words are often psuedo-synonyms to words we already use, like the word ingurgitate versus the word drink. That could make it difficult to reconcile one-word definitions and accuracy, since though there could be a very simple synonym (ingurgitate means drink), it would still be a shade off, and not totally correct. Would this be a problem?

I think, though, that with one of the solutions you had (to pair these words with sounds), the word can still be portrayed accurately. The listener can get a better idea of what it means by letting them hear what ingurgitate sounds like, so that they can hear that the audio clip is one of someone greedily gulping down a drink, rather than just casually sipping from their drink. In that way they can also learn the definition unconsciously, without really realizing it.

My words aren't much more than thinking out loud, though. I guess I just want to say that it sounds cool and that I'm curious to know how you're going to condense a bunch of descriptive definitions into single words.

 

[DoubleHelix]

My suspicion is that by attempting to learn during sleep we will negate the value of the sleep. But, there again, I always preferred the black hat in those silly corporate games.

I have no idea what that means. Maybe I am fortunate in that I haven't had to play many silly corporate games.

It is an interesting thought though, negating the value of sleep. I have thought about that, but I think that if you reinforce the connections that are lighting up anyway, it probably shouldn't be a problem, however, listening to a tape during all stages of sleep really could pose that risk, I guess. Curious.

My thinking is that obscure words are often psuedo-synonyms to words we already use, like the word ingurgitate versus the word drink. That could make it difficult to reconcile one-word definitions and accuracy, since though there could be a very simple synonym (ingurgitate means drink), it would still be a shade off, and not totally correct. Would this be a problem?

I think, though, that with one of the solutions you had (to pair these words with sounds), the word can still be portrayed accurately. The listener can get a better idea of what it means by letting them hear what ingurgitate sounds like, so that they can hear that the audio clip is one of someone greedily gulping down a drink, rather than just casually sipping from their drink. In that way they can also learn the definition unconsciously, without really realizing it.

My words aren't much more than thinking out loud, though. I guess I just want to say that it sounds cool and that I'm curious to know how you're going to condense a bunch of descriptive definitions into single words.

In response, to your last statement, I don't know what gave you the idea that I needed to condense descriptive definitions into single words. I don't think I do. Did I say that? If I did, I agree that there's no way to do it while retaining accuracy. There are a couple of things this brings to mind, though. First, even if the proper and full definition is not learned, it would still be a very significant finding if even a simple definition can be learned. After all, we are only 'practicing' and 'reinforcing' during sleep, rather than 'learning' from scratch. It would even be practical to know simple definitions - it would give you something to anchor the more complex and accurate definition to whenever you go back to study them.

Another possibility is that the full definition can be learned, and only the word needs to be repeated during sleep for it to be reinforced.

Or, the whole definition could be read.

However, none of the above will be the case. I came up with the design tonight, and I'm very proud of it. The details are not totally fleshed out, but that hardly matters - the foundation is solid. I was able to come up with it earlier than I thought. I still don't know a lot, but I know the right things.

On another note, over the past week, I have been reading up on memory, especially spatial memory, sleep, and vocabulary acquisition. I came across a little thing called the method of loci. 'Loci' as in the Greek (possibly Latin) "place," like the English word, location. It is an ancient method of memory in which memory of all kinds is translated by analogy into spatial memory. Orators used to use it to remember speeches, and it is where we get the phrases "in the first place, in the second place..." It is effective for memorizing really ANYTHING - anything that you can sit down and willfully try to memorize, you can memorize with this method. After learning how to use it to work with numbers THREE days ago, I can already recall a list of 200 digits, after studying it for 45 minutes (I am not NEAR my potential - I'm a n00b as you know - which means that I will get much, much, faster. Within a week, I expect to be able to do it in ten minutes). I can not only write it, but recite it, or recite it backward, or tell you any number that is in the digit's place you choose. If you studied it as hard as I have, you could do the same. It's the method, not me. And if you don't want to study hard, then study just a little bit a day for two weeks. Same thing. Actually, that method's probably better. I'm not kidding, guys...if you really want to know how it's done, P.m. me and I'll tell you what I know. It's more than you could get by scouring the web.


Anyways, I'm really proud of what I have. I really, really think it will work. If it does, it will be very important, whether or not it is recognized. But I don't have the time to share it all right now, I'm tired, and plus, I'm going to talk to the professor tomorrow or the next day and see what ideas she has to improve it.

Cheers

 

[Ism]

In response, to your last statement, I don't know what gave you the idea that I needed to condense descriptive definitions into single words. I don't think I do. Did I say that? If I did, I agree that there's no way to do it while retaining accuracy.

You didn't, but I wrote that with the assumption that any obscure word you could try and use would have a more complex description. My reasoning behind why is kind of silly, now that I think about it, but that's where I was coming from in regards to that.

 

[WALMART]

 

[zelo1954]

I have no idea what that [the black hat] means. Maybe I am fortunate in that I haven't had to play many silly corporate games.

The Devil's Advocate. Always looking for the snags.

 

[DoubleHelix]

Any of you scientists out there could really give me a hand on this one. But first a side note:


The method of Loci has been serving me well. I just memorized and repeated digits 201-600 of Pi. I made a couple errors, but they are the kind of errors that will go away with practice. I also invented a secondary system to use along with it that will allow me to hold ten digits per file instead of only four, and it should involve equal or less effort, but it will take practice to get used to. That means if I practiced it, I could do the same work that I did today, but hold 1000 digits instead of 400, with no extra effort. So, as my skills do increase, I think that 2000 digits should not be out of the question in about a month. If I had more time to devote to it, that would be a near certainty.

And yet, nobody has sent me a private message. Frankly, it's surprising that anyone would pass up this power. Perhaps you think it is too good to be true? Or maybe you think it is only a parlor trick? (That couldn't be farther from the truth - It'll be some time before I ever use a grocery list or forget an acquaintance's name.) Maybe you are too lazy - or too busy to incorporate something new? Or could it really be true...that you find this ability uninteresting...

The experiment needs refinement badly. Here's how it stands for now:

Previous research shows an increase in recall of spatial position of objects after hippocampal reactivation during SWS through familiar auditory stimuli. I need to use this fact as the crux of the experiment. Spatial recall of objects implies some recall of the objects themselves. So, the objects' content (what the picture is of) may be recalled with greater accuracy as an incidental effect of spatial recall, then the experiment should be a home-run. It seems like it would, but learning and memory can be counter intuitive.

The previous experiment had a 'memory match' style game, with digital cards in fixed positions. You could select two at a time, and you try to pick the matching ones. They had a few tries at this so the learning would be "robust." The relevant sounds the cards made when selected were replayed during SWS for the experimental group. Recall improvement was significant.

I'm making the assumption that with an weak an inefficient learning method, with a learning value of "1," each rehearsal is worth "+1" or something close to that. It's an awful scale, but bear with me. A method six times as effective may have a learning value of "6." Thus, the difference between initial learning and initial learning + 1 rehearsal is larger with the more effective method. SWS reactivation runs through the learning session in a linear fashion, so its efficacy is dependent on the efficacy of the original learning method. Thus, for an experiment, the most effective learning method should be used. Furthermore, if the SWS reactivation is roughly similar in learning value to a practice session, then the fewer practice sessions there are, the less its effects will be obscured. Thus, effective learning with minimal rehearsal is optimal.

Loci works because it ties a mnemonic image to a space, and other mnemonic or analogical images nearby. Spots in a path around town or in a path through your house work well, but it might also work in a 2d landscape. It also involves motion, but that might be hard to recreate, so we can leave it out. Already, our recreation is considerably inferior to loci, but let's move on for now. An excellent mnemonic image to use for a vocabulary word is an image phonetically similar to the actual word but literally representative of the definition. For ambuscade, I imagined an ambulance on a skate with a tank cannon on top, waiting behind my wall socket, and then creeping out to attack. The definition I had was "secret or hidden attack," if you didn't know. I didn't. Because I can picture all of the images at 50 predetermined spots in my home, that go in order as you walk through it, I know that ambuscade was the 9th word on the list. Knowing WHERE the word is is of utmost importance - the ambulance skate cannon would be helpful, but it wouldn't permanently sink in unless I had it anchored to my wall socket. If more effective learning makes for more post-SWS reactivation recall, I'm taking it all into consideration.

So let's say we wanted to test SWS on learning German words. German words are good in that they have familiar combinations of phonemes for an English speaker. When taught, they should be accompanied by a mnemonic picture, or just a pair of pictures, one representing the word, and the other representing the phonetic similarity. Here's a potential list of some interesting ones I thought up. I could grab a dictionary and get more/better ones. This similarity allows for the audible enunciation of the German word to carry meaning with it, after the picture has been seen. That's good.

Opposite = Gegenteil – black and white checkered tile
Beans = bohnen – a bone (among beans)
To iron = bügeln – a bugle (near an iron)
Candy = Suessigkeiten – Dr. Seuss (near candy)
Guy = Typ – (guy sitting at) typewriter
To destroy = zerstoerren – store (being demolished)
Clothes = Klamotten – clam (on or near clothes)
Lightning = Blitzen – blitz (football players struck by lightning)
Cake = Torte – tortilla (near a cake)
Dryer = Trockner - Truck (Hauling a dryer)
Pot = Topf – a top (in or near a pot)
Left = links – chain links (near left arrow)
Fast = schnell – snail (moving quickly)
Head = Kopf – cop (with a large head)
Paintbrush = Pinsel – pencil (in a cup with a paintbrush)
Office = Büro – Burrow (desk underground; burrowed holes in desk or office)
Saturday = Samstag - (Sam’s, the store, near a checked calendar)
Brush teeth = putzen – Zen (Buddha holding toothbrush)
Coat = Mantel – coat hanging on mantle
Relaxed = locker – lock or locker (near man reclining on beach)
Knife = Messer – messy room (near a knife)
Pineapple = Ananas – bananas (Pineapple among them)

There would have to be a lot of words, otherwise I fear even the control group would remember them all. First, familiarize the subjects with the spatial orientation of the pictures. Before the subject begins, she is shown an example of what the pictures are for and how they work. Split the words into groups of 8. We could use a computer program with sixteen cells. The subject is shown the 8 pictures in order. They are in the slots numbered 1 through 8, so each slot has an 'identity.' The images are scrambled, now on the bottom of the screen, and the subject must put them back in order. If they drag to the wrong spot, there is a buzz and the image goes back to where it was. When they drag them to the correct spaces, there is a brief tone and the audible German word sounds. This will pique curiosity and attention. Next, they are given each word and its single word definition in each of the bottom 8 cells, with the 8 images in the appropriate top 8 cells. They are told to drag each word/definition to the correct image. Buzz if incorrect. If correct, a tone and then the audible German word sounds. In the next phase, the German word is superimposed upon the images in the appropriate top 8 spaces. The scrambled definitions at the bottom must be dragged up to the proper pictures and words. Either buzz or tone + audible word. Finally, the German words remain in place, but the pictures vanish. The scrambled definitions on the bottom cells must be matched with the words in the top 8 cells. These four phases should instill a spatial sense of the words' positions, and each word will have been heard four times. This process should be relatively quick, so it could be repeated with many groups of words. The largest amount possible before fatigue starts playing too large a role. The next group of words should be in cells labelled 9-16, to create the illusion that these cells exist "somewhere else."

Perhaps turn off the lights, leave the room and tell them to relax or whatever - facilitate the learning with a few minutes of wakeful-rest. Remember, I am going off of my assumption that the better the words are learned in one session, the stronger SWS effects we'll see. The whole thing depends on there simply being too many words to remember. In SWS, tones and german words are played in order for the experimental group. After waking, the subjects are administered a pencil and paper test. They must write the correct English translation of each of the ~48, randomized German words. Their memory will be pulling information from several angles, and knowing the position of a picture entails knowledge of the picture itself, which is a dead giveaway for the definition. Thus, the greater spatial recall, the better retention of definitions.

Let me know if you need clarification.

Let me know how the experiment could be improved. I'm really looking forward to any feedback.