temp to perm memory

June 29, 2005 1:44 PM

For some research I'm doing (in order to help actors), I'd like to know what science exists about moving data from (human) temporary memory to (human) permanent memory.

I DO know that our temporary memory can only hold about seven items at once. But what is the most efficient way to move those items from temp to perm? Let's say someone told you to memorize three phone numbers so that you could recall them next week. Obviously, this would involve some repetition. But how much? Can we say something like, "for most people, saying each number 60 times will be enough to move it into permanent memory?" Should you repeat it 10 times, take a break, and then do another round? Surely there have been experiments.

posted by grumblebee to science & nature (13 comments total)

The experiments form nearly the entire basis of the field that is named "Cognitive Science." The questions you ask have hugely complicated answers.

I have an introduction to cognitive science text, but it's sort of dense. If you found one of these (by searching amazon for 'introduction to cognitive science', for example), you would learn a lot about experiments, and not very much about practical things to do to improve memory.

The practice of improving memory is called 'mnemonics'. I don't know of a good book about mnemonics offhand, as my freakish memory makes them unnecessary to me, but they are certainly out there.

A girl once dared me to memorize her phone number, which was 280-4376. Here is the process I used:

An old girlfriend of mine once owned a Datsun 280Z, of which I was very fond. I imagined this car.

My grandfather used to live in an apartment building whose street address is 4076, which I happen to own.

I formed a visual image of the 280Z, lodged in the side of that house 3 floors above my grandfather's bedroom. That house did not in fact have 3 floors, but the image simply contains a car sticking out of the top of the house with a sort of mental note that is 'CAR IS THREE FLOORS ABOVE BUILDING.'

The trick is to find things to put in the image that have some emotional connection to you. I loved the girl with the Datsun and I loved my grandfather, so that works. If love is not a big emotion in your life, pick things that remind you of anger, jealousy, sexual attraction - any of the amygdala functions.

There you go. I don't actually recommend this method for 'next week' use, because it is really permanent - I did the above example in 1999. Luckily she retains the number and I call her from time to time to chat.

I've slightly altered the number so my friend won't get a bunch of calls from you freaks.
posted by ikkyu2 at 3:02 PM on June 29, 2005

There are tons of mnemonic approaches to memorizing things. For example, I read this guy's stuff. He is a Guinness record holder for memorizing 800 digits after hearing them once. Im assuming you don't read hebrew though. You might check out this book which tries to be more scientific about various techniques than most memory books. It is a good starting point.

The 7+-2 is the number of things we can be consciously aware of at once. It is a slightly deceptive concept as consciousness is not something that neuroscience is even marginally close to explaining or pinning down. There are many ways to hold more things in your awareness, for example by grouping things.

In terms of the memory techniques. Repetition is not the issue, it is building strong enough associations. If you build them strong the first time, you do not need to repeat them at all.
posted by blueyellow at 5:39 PM on June 29, 2005

Here is where 7+2 comes from - the original essay by George Miller.

I have to disagree that 'repetition is not the issue'; building strong association is part of it, but so is repetition - although not rote. Use the information in conversation (or if you don't want to bore your friends, in imagined conversation). Follow the 'active learning' approach. Actors do this through rehearsal, whereby the physical actuality of the performance reinforces memorisation of the script.
posted by TimothyMason at 10:25 PM on June 29, 2005

Oh - and here's a chapter from a psych text book on memory and memorization. You might want to check out the SQ5R memorization programme, which is recommended on a number of university study-guide sites. Further web leads here
posted by TimothyMason at 11:41 PM on June 29, 2005

For a completely non-scientific subjective answer, the thing that has worked best for me is associating something with a rhythm. That means either finding rhythms in the language (rhyme, alliteration, meter) or memorizing it to a tune or a beat. The tempo acts like a prompt.
posted by cali at 12:13 AM on June 30, 2005

You should look at learning styles. I would bet a disproportionate number of kinesthetic learners (usually a small minority of adults, but a the majority of children) are actors, so the type of excercises that would be most beneficial to actors are not what would be most beneficial to adults, generally.
posted by raaka at 4:02 AM on June 30, 2005

Check out books by Tony Buzan on improving or learning to harness your memory. There are plenty of other authors. Find a Buzan book on Amazon and look at the "other authors" list at the bottom, or the recommendations. Then Google em. Oh, I love the interweb.

The basic techniques seem like "tricks", but they work. They also seem to be how people with powerful memories actually use their brains, without even realising. The substitution techniques (like the car/house/phone number example that ikkyu2 mentions) seem to be the most popular, and are certainly effective. They definitely achieve long-term storage and retrieval. Buzan describes a few systems to get yourself started. He also recommends incorporating emotions into the memory - so the image of a car sticking out of your grandfather's house may seem hilarious, or frightening, which will heighten the importance the brain attaches to the knowledge, making storage and retrieval easier.

There have been recent studies (no idea where) showing that improving the short-term "scratch-pad" memory can improve performance in intelligence tests. This is the part of the memory that can remember and manipulate tokens. For example, when you're calculating 7 x 13 + 98 - 4 x 3 you have to remember the answer from each step, then manipulate it according to the next operator. I don't know what sort of puzzles would exercise and improve this part of the brain, but I've recently started playing su doku. The harder versions involve remembering a few numbers and options whlie trying to fill in the grid. It seems to me that improving the performance of this part of the brain is will make your short-term memory tend towards 7 or more objects, rather than 7 or fewer. And if you work on memory systems such as Buzan's, you ought to improve your general memory performance.

All the above is IMHO and I have little evidence other than anecdotal :-)
posted by ajp at 4:08 AM on June 30, 2005

Having re-read your question, I think I got a bit side-tracked in my response.

One way that the ancient Greeks famously used to memorise long speeches, in the days before the autocue, was to think about a journey with which they were very familiar. Say, from home to work. At key stages on this journey (waking up, switching off the alarm, getting out of bed, having a shower) they would associate sections of the speech. The mnemonic/substitution technique would be very useful here.

This is also a technique often cited by performers who manage tasks like remembering a deck of cards within a minute. But it boils down to the same stuff mentioned here.
posted by ajp at 4:33 AM on June 30, 2005

building strong association is part of it, but so is repetition - although not rote. Use the information in conversation (or if you don't want to bore your friends, in imagined conversation)

What you are describing is a process of familiarization and generating new associations by application in new contexts. Thus repetition in this case builds associations and understanding.

You can not remember things you are not familiar with and do not understand. By understand, I mean, have a useful way of representing in your mind. For example, you could remember a particular equation as a sequence of symbols (which unfortunately is how many students do it), but you would be much more successful if you actually understood what the equation means. If you didn't find any way to express some structure of the equation in your mind, you couldn't remember it.
posted by blueyellow at 5:42 AM on June 30, 2005

Thank you for all the responses. They are all interesting, but most don't specifically address what I'm asking (or more likely, I wasn't clear enough).

If you repeat something enough times, you WILL eventually lodge it in your permanent memory. There may be more efficient methods, and there may be less painful methods, but repetition does work.

My question is, on average, how many repeats do you need? Maybe this varies from person to person, but I bet it's pretty close to the same number for most people.

Here's why I'm asking: memorization for actors is different than most other uses of memorization, because it has to be EXACT. If you're making a graduation speech, no one cares if you substitute a word or two. But if you're playing Hamlet, you must be word perfect. And most mnemonics tricks don't help with that.

So if your line is, "I'm coming too," you can't say, "I'm coming also." (Some would disagree, but in my theatre company, exactitude is important.)

At times, the only way to deal with this issue is through repetition. Actors sometimes try and give up, saying they'll never get it. I think they're wrong. I think they just haven't repeated it enough times. It's not fun to repeat things over and over. But I think if they had a magic number, they'd be more likely to try. Because there would be a light at the end of the tunnel -- even if it's a long tunnel.

Hopefully, they won't have to use this method to learn all of their lines, but they do need to use it on problem words and phrases.

Now, a little personal history: I have a HORRIBLE memory. Or rather, I have a really hard time moving things from temp to perm. I have tried every mnemonic trick, and none work for me. The ONLY way I can memorize is by repetition.

I recently had to learn a huge part -- probably about 20 pages of speech if you run all my lines together. Here's how I licked it:

If my speech started like this...

Now is the winter of our discontent
Made glorious summer by this sun of York;
And all the clouds that lour'd upon our house
In the deep bosom of the ocean buried.
Now are our brows bound with victorious wreaths...

I took the first five words, "Now is the winter of" and repeated them over-and-over, 50 times. I chose five, because that seemed to be about the number I could hold in my TEMPORARY memory at one time, without making any mistakes.

Now is the winter of
Now is the winter of
Now is the winter of
Now is the winter of
Now is the winter of
etc.

When I got finished with this, I took the last word of the first five, "of," and added on the next four, and repeated these five words 50 times:

of our discontent Made glorious
of our discontent Made glorious
of our discontent Made glorious
of our discontent Made glorious
of our discontent Made glorious
etc.

People told me I was insane to work this way, that I needed to give the lines MEANING, and, of course, by breaking them up this way, they had no meaning. But I had already done all my meaning research prior to memorization. So I already knew what they meant.

In the end, though it took a REALLY long time to work this way, and though it was hateful work. I knew my lines perfectly long before anyone else in the play did. And I had more lines than any of them. Some of them never learned their lines word-perfect. But I did. I would say a word, and the next (correct) word would just pop into my head with no effort. And my confidence on stage went through the roof.

Sorry for the LONG response. I wanted to explain, in detail, what I'm trying to find out and why I'm interested. I know this is the only method that will work for me. But I think people with better memories should use it to solve problems. The question is -- how many repetitions. I settled on 50, having first tried and failed with 20.
posted by grumblebee at 8:06 AM on June 30, 2005

Not sure if it even comes close to answering your question, but I've seen acting books recommend learning lines backwards, especially for monologues. Learn a chunk at the end first, then the next chunk back, etc. It supposedly give you more confidence, since you're not hitting that "Oh my god, what's my next line!" panic so often, so that you can relax more and concentrate better.

But I'm not sure that I agree with there being a "magic number" of times. I know that I learn much better by staring at lines on a page, while I've worked with actors who can only memorize lines that they've heard over and over (this would be why, I think, some people are good at quoting movies they've seen, others not so much).

I think it has to do with learning styles, and ability to concentrate, and general constituional attention to details, and all sorts of variables. And I would worry that saying "You just have to do this X times" would end up being more frustrating to those folks who still didn't get it after X times.

Possibly off-topic: I've also noticed a weird phenomenon with every play I've worked on: Right after everyone goes off book, stumbles for a week or so, and then seems to get all the lines perfectly, everything suddenly falls apart. (It's always been after two weeks off-book for me.) People miss lines, screw up cues, etc, and everyone's frustrated as hell.... and then it suddenly all magically comes together. I've always figured that stumbling period was the "transferring to long-term memory" period.
posted by occhiblu at 8:53 AM on June 30, 2005

Slightly annoying PDF that might help. It gets into more of what I was talking about with different learning styles, and what techniques might work for those styles. (It's not particularly in-depth, and might be stuff you already know, though. The memorizatin section starts on the brochure page 27.)
posted by occhiblu at 9:03 AM on June 30, 2005

(Because why post only twice when you can make it an even three?)

All *that* was to say -- someone who learns well by hearing things out loud will probably have to repeat lines fewer times to memorize them than someone who learns them by staring at the page. That's why I don't think there's a universal "magic number." If it takes the first person 20 times and the second person 100 times, you can tell them that it takes an average of 60 repetitions, but you'll end up with the first actor VERY bored and annoyed at having done three times as much work as necessary and the second actor frustrated and annoyed at being forced to do something that's not effective and still not getting it.
posted by occhiblu at 1:03 PM on June 30, 2005