Memory

Dr. C. George Boeree


The traditional approach to memory divides memory into three parts. The first part is called the sensory register (or "very short-term memory"). It is an experience that lasts for a very short time because it takes a second or two for the sensory neurons and the neurons that come right after them to recover from stimulation. If you look at a window, then shut your eyes tight, you will still see the window for a moment, and then see a reversed image of the window (where white becomes black, red becomes green, blue becomes yellow, and so on). Then it all fades back to black.

This visual sensory memory is also called iconic memory, and last less than a second.  Do you remember when you were little and you played with sparklers on the fourth of July?  Remember making circles or writing your name in the air?  You could actually see the circle or the name!  That's because the information was still in your neurons.

The auditory version is called echoic memory, and it lasts three or four seconds.  Did you ever have someone tell you something while you were distracted, turn to them and say "excuse me?" only to "hear" what they said?  Of course you have!  That's echoic memory.  Other senses have similar forms of sensory memory.  This is not, of course, what most of us think of as memory at all.  It is really more an aspect of perception.

The next kind of memory is called working memory (or "short-term memory"). It is the aspect of memory that you are aware of, or can bring back very quickly and easily.  It is where you do your thinking and imagining.  A decent analogy is to the working memory (RAM) of a computer.

The aspect of working memory that has been most studied is nowadays called the phonological loop.  If you read a telephone number in the phone book, you can keep it in your short-term memory for about 10 to 20 seconds. If you repeat it over and over, you can hold on to it longer. If you are anything like me, you will forget it before your reach the phone anyway!

Besides being brief, this kind of memory is very limited in capacity. It can hold “7 plus or minus 2” chunks of information, that is, from 5 to 9 “things.” Notice that phone numbers are 7 numbers long -- just about the right length to keep in mind for a bit. Social security numbers are 9 numbers long, and that is a strain for many of us.

One thing that helps is that the “chunks” can be of different sizes themselves. So, while a series of random letters, like IBMFBICIAJFKLSDNOW..., are pretty difficult to remember, they become much easier to remember if you “chunk” them: IBM FBI CIA JFK LSD NOW....

Another part of working memory is called the visuo-spatial sketchpad.  As you can easily figure out from the name, it is like the phonological loop, only it is devoted to the visual sense.  "Sketchpad" is in fact a nice name for it, since it functions like one:  You "draw" things on it as you work with images.  When you imagine a multicolored cube and turn it to see the different colored faces, you are doing that in the sketchpad.  Or when you are counting the windows in your house.  Or picturing a friend smiling at you.

Presumably, other senses have similar "work spaces".  Even motor activity, such as going through a dance move or a gymnastic move, probably has a specialized area.  But all of them, just like the phonological loop, are rather limited in size and the things you hold in these kinds of memory can only be held for a matter of minutes at a time.

There are particular parts of the brain, near the respective sensory or motor cortices, that appear to be associated with these different kinds of working memory.  But working memory isn't really a place where things are temporarily stored.  It is actually more a matter of temporary cycles of neural excitement that, if repeated often enough, eventually leave their mark as more permanent memory.

This last kind of memory is long-term memory. As the name implies, this contains the memories that we hang on to for a long time - our whole lives, often enough. It is also enormous in capacity - more room than we will ever need! Long term memory is a little like the hard drive on your computer. And you may have noticed that, once you are talking about terabytes, you rarely have to worry about running out of room.

Long-term memory clearly involves physical changes in the brain - most likely some kind of facilitation at the synapse: Repeated experiences increase the efficiency of certain synapses by “pulling” receptor sites and axon endings closer together, as well as making the receptor neuron more sensitive over the long run. This is called long-term potentiation.

Moving things from short-term to long-term memory is called encoding.  One thing that helps encode verbal information, for example, is repetition or rehearsal.  Repeat that poem or list of  French words over and over, and eventually, the short-term cycles become permanent.

Three additional significant things that help us encode verbal information are organization, meaningfulness, and imagery. It helps, for example, to put things into headings and subheadings, or to use outlines to organize difficult material. Likewise, you may have noticed that it is easier to remember things that are relevant to you, that have meaning for you: stuff about your family rather than others’ families, or classes in your major rather than “gen ed” classes, and so on. And if you can visualize the information - using your mental sketchpad - you will remember it even better. It seems that human beings are very visual creatures.


Cognitive Structure

We often think of memory as involving storing stuff in the brain or mind, but that's just a metaphor. In a very real sense, there is "nothing" in the mind, no thing, that is, except relationships (as the linguist Hjelmslev once said).

It is difficult to describe something so complex, fluid, and ever-changing as these knowledge-relationships we have. We psychologists have drawn a little inspiration from linguistics, and as long as we remember that we are talking metaphorically, we have a nice model of the mind...

When we use these linguistic models, we talk about our cognitive structure (or our construction system), and the components or “basic building blocks” of this cognitive structure we call concepts (or constructs, contrasts, dimensions, categories...).

Concepts are ways we have of organizing what we have learned from our experiences. Concepts treat a variety of experiences as equivalent in some way:  It could be features or qualities these experiences have in common, or their general similarity to some prototype, or some way in which we, the conceptualizers, relate to the experiences - something like Gibson's affordances.

Birds’ feathers are an example of a feature. Coins are gold or silver colored - this is a quality of theirs. A robin or a sparrow are more prototypical birds - ostriches and penguins are not so obvious. A chair is anything we use to sit upon (something that affords sitting-upon).

Please note that these concepts need not be verbal: A cat knows the difference between the expensive cat food and the cheap stuff, yet can't tell you about it; An infant knows who mommy is, long before he or she can say the word; Wild animals contrast safe areas and dangerous ones, etc. Even adult humans sometimes "just know" without being about to say: What is it about that person that you like or dislike?  It may be quite difficult to put into words.

Concepts don't just float around independently, either. We interrelate and organize them.  For example, we can define some category of things by combining various concepts: "Women are adult female human beings." Or we can go a step further and organize things into taxonomies, those tree-like structures we come across in biology: A Siamese is a kind of cat, which is a kind of carnivore, which is a kind of mammal, which is a kind of vertebrate.... Both of these - definitions and taxonomies - are contained in what is called semantic memory.

Or we can put contrasts into structures that involve steps one must take, as well as definitions, like rules. These are often called schemas or scripts. You can find explicit examples in books about card games, etiquette, or grammar; but you know quite a few rule systems yourself, even if they have become so automatic as to be nearly unconscious!

Not all organization of contrasts are so tightly structured. We can describe something: "Women are delicate." As the example is intended to suggest, descriptions, as opposed to definitions, don't necessarily have to be true! Beliefs are similar to, but looser than, taxonomies.  They don't have to be true, either.  Whereas birds definitely (i.e. by definition) are vertebrates and have feathers, it is only a belief that says they all fly - you could be wrong! Stereotypes are examples of beliefs; so are opinions. But some beliefs are so strongly held that we see them as definite.  This is a common cause for arguments - and wars!

Another kind of memory is called episodic memory, which include our narratives - the stories we have in our minds. These are temporal, like rules, but are amazingly flexible. They can be a matter of remembered personal experiences, or memorized history lessons, or pure fiction. I have a suspicion that these contribute greatly to our sense of identity, and that animals don't have them to the degree we do.

Both semantic and episodic memory involve the hippocampus. The hippocampus seems to be involved in working memory, and seems to be responsible for the transfer the information to the cerebral cortex, which is clearly the locus of most of our long term memory. If there is damage to the hippocampus, we have trouble developing new semantic and episodic memories. If the damage is to the left half of the hippocampus, it is verbal memory that is most affected. If the damage is to the right half, it is spatial memory that is affected.

The hippocampus is also a part of the limbic system, and as such it shows why there is such a powerful relationship between emotional situations and strong memories. Exceptionally clear memories of emotional events, ones that have a near-photographic quality, are know as flashbulb memories.  The amygdala appears to be especially involved in these emotional memories.

The last kind of long-term memory is called procedural memory.  You know how to drive a car, play rugby, ride a bike, hammer a nail, and so on, without giving them much thought. Procedural memory involves the cerebellum (as well as the motor cortex and the basal ganglia) which makes sense in that most of our "procedures" involve movement.


Remembering and Forgetting

Remembering (often called retrieval in research literature) comes in two forms: recall and recognition. Recognition is the easier one: We recognize our friend when we see him coming down the road. Recall is more effortful, and involves mentally rebuilding the experience. It is a myth that we have everything in our heads like a motion picture. Really, we only have a certain amount of  “information” in the form of neural connections, which we use to reconstruct our memories.

There is a degree to which we tend to forget things as we get older, and there is some loss of neurons as we age. And there are drugs (such as alcohol) and diseases (such as strokes and Alzheimer's) that can speed that loss along. Amnesia is what we call the more sudden loses of memory, whether temporary or permanent.  The most dramatic examples occur after serious trauma to the head such as sometimes occur with car accidents or gun shots to the head.  The usual kind of amnesia is called retrograde amnesia, where you can't remember past events.  It is usually episodic memory (memories of events in your life, or even of your identity).  We seem to retain things like our skills, the ability to speak, definitions of words, and so on.

Anterograde amnesia, on the other hand, means you can't make new memories.  As mentioned above, the hippocampus plays a big role here: if the hippocampi are badly damaged, such as by a stroke, the person cannot transfer their experiences into long-term memory.  A person with anterograde amnesia remembers their past, but will lose his or her experience of all new events in a matter of minutes.  If you introduce yourself and have a nice conversation with such a person, then leave and come back ten minutes later, they will act as if they had never met you.  In their minds, they never have!  A good movie that plays on this is Memento.  Another is 50 First Dates.  But there is nothing amusing about this disorder.  Most of these people wind up in an institution, living each day as if it were the first since their accident.

Most of our day-to-day forgetting seems to be a matter of interference. In other words, there is so much stuff in your head that it is hard to separate one thing from another.  It's like trying to find something in a particularly messy attic:  It's not that the stuff isn't there somewhere, it's just that you can't access it.

Proactive interference is when an older memory interferes with remembering a newer memory. If you take a French lesson after taking a Spanish lesson, some of your Spanish may creep into your French. It is harder to remember that the word for man is now homme, not hombreRetroactive interference is when newer memories interfere with older ones. When you talk to your friend Juan right after your French class, you may tend to say homme instead of hombre. Interference is really a simple idea - sort of like how its hard to find things when your hard-drive is stuffed full of files, or your room is filled with junk.

One of the biggest controversies in psychology concerns repressed memories. This is the idea, promoted by Sigmund Freud, that we push painful memories out of our awareness and into a deep, dark place called "the unconscious mind." This is why, traditionally, we talk about going to a therapist to try to recover these traumatic memories so we can deal with them. There have even been some therapists who use hypnosis to recover repressed memories.  (Note that "repression" is a broader idea of Freud's which includes not acknowledging your needs or desires as well as repressed memories.)

Unfortunately, some of the people who remembered terrible things like being abused as children were discovered to have created these memories under pressure (unintentional, we hope) from their therapists! Some parents were even sent to jail because of their adult children's “recovered memories.” But research indicates that not only is there very little evidence of repressed traumatic memories, but trauma - with its emotional intensity - actually makes memories harder to forget!  Remember "flashbulb" memories?

Of course, people really do get abused, and other traumatic things do happen to people. There have been people who have recovered memories and whose memories have been confirmed. So it is a difficult issue that has yet to be decided.

But, as I said earlier, memories are not like films. Outside information may alter our memories as we reconstruct them. Some people are easily manipulated, and everyone can be manipulated to some degree. This happens, for example, when a lawyer asks you what happened when you saw the accused’s car “crash” into his client’s car - when in fact it merely bumped into it. Hearing the word crash tends to subtly alter your recollection in the direction the lawyer wants it to. Hypnosis is especially powerful when it comes to altering memories. So are drugs. And children are very susceptible to manipulation. This is why children’s testimony in court is rarely accepted.


© Copyright 2002, 2009 by C. George Boeree