with a different sort of demonstration.

I want to just show you one of the change-blindness studies

that has been done in the real world.

And these videotapes are not available publicly.

We get them from the web and see them as little Java scripts.

So, this is one of the first studies done by Dan Simons when

he was at Cornell. And his adviser at the time was

our Frank Keil, who's now in our department.

So, here's the study.

[laughter] And you don't notice it.

Change blindness is one of the more striking phenomena

discovered by laboratory scientists and by psychologists.

But it's important to realize, to get away from the sort of

surprise of the gorilla and the fact that it's hard to see the

flickering--the object that's flickering,

and appreciate the big moral of this, because the big moral of

this is actually, I think, striking and quite

important. You think right now that you're

perceiving the world. I look down on you and I think

I have a whole sense of where everybody is.

I can't see everybody perfectly in back.

You're kind of far away and blurry but there's a sense in

which I have a world around me. Similarly, if I'm to close my

eyes for a second, everything just remains and I

could sort of remember some of the things that are there.

That's really good sound localization by me .

So you're looking up and you think you have a sense of the

world both in perception and memory.

The change-blindness experiment suggested this isn't true.

The change-blindness experiment suggests that if you look at me

for a second and during that second all of your classmates

change positions, including those next to you,

you are extremely unlikely to notice.

The change-blindness experiment suggests that if you turn your

eyes away from me towards there for a second and turn back,

and I'm dressed entirely differently, you wouldn't

notice. The exceptions would be if you

told yourself consciously, "Remember what this guy is

wearing; he's wearing this,

that and the other." But if you don't do it

consciously you'll lose it, and usually this is okay.

Usually, it's okay because your memory and your visual system

exploits a basic fact about the universe,

which is that most things stay the same most of the time.

I don't have to explicitly remember that you're over there

when I turn my head for a second because you'll be over there in

any case. You don't need to hold precise

representations of the world. And so you only notice it in

certain clever circumstances. One sort of clever circumstance

is when psychologists change reality as in the

change-blindness studies. A second sort of circumstance

is in movies. So, one of the big surprises

when people started making movies involving cuts was it is

extremely difficult to get everything continuously right.

And you need to work very hard to notice.

So, there's all of these continuity errors that creep up

into movies and you have to be a film buff or writing it down to

even notice this. And the overall moral here then

is that your perception of reality is a lot more sparse,

a lot more limited, than you might think it is.

So, this is where we were at the end of last class.

We were talking about the different sorts of memories:

Sensory memory, which is the sort of fraction

of a second of sensory residue of what you're hearing and what

you're seeing, working memory,

short-term memory, and then long-term memory.

And we talked last class about how things get into sensory

memory, into working memory, the role of attention.

And in fact, the change-blindness studies

are actually just studies of how something gets from your senses

to your consciousness and what does and what doesn't.

Now I want to move to the distinction between working

memory, short-term memory, and long-term memory.

Now, the obvious distinction is actually just in fact--is

storage differences. So, long-term memory or "LTM"

has a huge storage capacity. This is your memory like the

hard drive of your computer. This is the memory you walk

around with. It includes all the words in

English, just for example, 60 to 80,000 words.

It includes everybody you've ever met, languages,

faces, stories, locations, nursery rhymes,

songs, TV programs. Nobody knows the storage.

It is not true that you remember everything that has

ever happened to you. There's no reason to believe

that this is true. At the same time though,

you have a huge amount stored in your brain in long-term

storage and nobody actually--It has to be limited because it's a

finite, limited brain.

But nobody knows how big it is. Nobody knows how many terabytes

you carry around in your brain and--but it's a lot.

Compare this to working memory – the short-term memory,

which is actually very limited. Your memory of what you could

store on--in--where you could hold in consciousness right now

is quite limited. Here is an exercise.

Do not write these things down. I want you to remember them.

I'm just going to give you a few numbers: 14,59,

11,109, 43,58, 98,487, 25,389,

54. Please write them down.

View this as an IQ test if that would relax you.

How many of you who decided to participate in this experiment

got three or less? Good.

Good. Four, five, six,

seven, eight, nine or more?

Anybody get all eleven? This is a particularly

difficult memory task. The numbers are meaningless.

And I told--and I forgot to tell you to get your pen and

pencil ready, so some of you just glared at

me. But [laughter]

under normal circumstances the cognitive psychologist George

Miller said that this sort of suggested that the standard

memory storage of short-term memory is seven,

plus or minus two. And what that means is anywhere

from five to nine roughly. Some of you,

I bet, can beat that. Some of you on a not-so-good

day maybe won't make it that much.

Now "seven plus or minus two" is what you--;so,

that's what you hold in consciousness.

I can tell you 14,21. You walk around,

"Oh, yeah, 14,21." You hold that in consciousness

with no problem. But I throw eleven numbers at

you, you can't. Some dribble out.

You can't hold that in your conscious window in your

short-term memory. Now, this raises the question

"seven plus or minus two" what? And the answer seems to be what

George Miller calls "chunks." And a chunk is a basic memory

unit, something you think of as a single, individual entity.

So, suppose you see the string of letters "L,

A, M, A, I, S, O, N."

If you don't know--If you can't form these into words and you

have to remember them, these are eight chunks.

You have to just pick them up separately.

On the other hand, if you break them up into four

words you could just remember it as four chunks.

And if you break it up into two words in French,

"la maison," "the house," it could just be one or two.

How much you know depends--affects how much you

memorize--how much you could store in memory because it

affects what counts as a basic unit of memory.

And there's all sorts of examples of this.

If I tell you "1,1, 0,1, 1,0, 0,1,

0,1, 1,0," those of you who don't know binary numbers might

have to remember that as "1, 1,0, 0," whatever I said.

Those of you who are computer scientists or mathematicians or,

for whatever reason, know binary numbers could

convert it into a single binary number.

Anybody know what the number is? No, I cannot say it again.

[laughter] Some number,

24, or not 24--to some number, 24, and then you remember "24."

It's easier. Suppose you see a chessboard

and the chessboard is set up and you don't know how to play

chess. It is murderously hard to

remember that. They've done the experiments.

They've taken people in a lab who don't know how to play

chess. They set up a chessboard and

then they say, "Okay.

Look at this for five minutes." Then they take it away,

set it up again, and it's murderously hard.

"There is a horse-y thing on the side there and everything."

But if these chess pieces are set up in some way that's

logical for a chess player, then a chess master could look

at it and remember it in a glance, "Oh.

It's the Fibonacci defense" or something like that [laughs],

and then immediately recover it.

Similarly, football coaches have been tested on their

memories of football diagrams. And they have a photographic

memory for football diagrams because it corresponds to things

that make sense. Architects could have a

photographic memory, a perfect memory for floor

plans because it makes sense to them.

They understand it