There has been much research into the ways which our "memory"� actually stores information in

our brains. One of the original theories is Crain and Lockhart's Levels of Processing Model,

which represented the memory of a series of stores and boxes with a fairly linear progression from

sensory information to short term store to long term memory. They proposed that the different

levels of memory were due to the different levels of "processing"�, but found it very difficult to

quantify what "deep processing"� is. A more complicated, yet more plausible theory is the Parallel

Distributed Process model which suggests that memory operates on many different levels at one

time, connecting many different stored items, which have connections to other items. The evidence

for this is seen in our ability to generalize and remember information that we have not specifically

"stored"�.

But these theories do not answer the basic question: How exactly does memory work?

There have been many experiments conducted to try to determine the answer. Hermann

Ebbinghaus was one of the first (1830s) to experimentally pinpoint factors that aid in memory. He

proposed that distributed practice (items learned over a period of time rather than all at once) was

most effective for long term memory. His findings were later confirmed by an experiment run by

Bahrick and Phelps which tested the participants on Spanish vocabulary they had learned eight

years prior. They found that those who had learned over longer periods of time remembered more

than those who had learned through massed practices. Of course, there is an alternative theory

called the Total-Time Hypothesis. This hypothesis states that learning depends on the amount of

time spent altogether, not the way it is spaced out. This theory has several drawbacks in an

experimental setting because it is difficult to control the attention span of participants and would

have no external validity in the real world. The impact of these theories on the idea of "learning"� is

enormous. The evidence suggests that the most effective way to learn a piece of information would

be to rehearse it repeatedly over time, not "cram"� study in a short, but intense, time periods.

One very interesting experiment that was performed by Karni and Tanne tested the effects

of REM sleep on learning capability, a cognitive function that is difficult, if not pointless, to

separate from memory. It was found that disruptions in REM sleep patterns reduced the amount of

improvement the participants showed on visual-discrimination tasks. The connection was further

evidenced by the finding that there were increases in REM sleep in proportion with exposure to

learning situations. It was concluded that REM sleep aids in memory consolidation, but the full

extent of these findings cannot be understood unless they are viewed in the context of another

experiment conducted on rat hippocampus cells. It was found that cells in the hippocampus, which

has been determined to play a role in memory storage, that were activated during a learning task

were reactivated during subsequent periods of REM sleep. It seemed a form of "replaying"� for the

purposes of consolidation. These results also shed light on the theory of distributed practice,

because by spreading a learning task over time the brain has more opportunities of REM sleep for

rehearsal, a major factor in the conversion from short-term to long-term memory. Many real-world

situations provide evidence that rehearsal, not parrot repetition, secures information in the long

term memory, such as ad campaigns that run constantly but fail to register in the minds of the

public receiving them. Of course, one must take into account that the experiment was run on

laboratory rats which have a much simpler level of cognition than humans. Perhaps the rats

"dreamed"� of the learning tasks because they were all that concerned the brain at that time.

Another issue of memory that must be considered is the effect of different forms of

retrieval. What good is it to have information stored in long term memory is one is unable to

access it? Tulving and Pearlstone conducted an interesting experiment that showed how the way

we retrieve information is almost as important as how we memorize it in the first place. The

experimenters gave the participants a set of information that was organized into categories to

memorize. They then separated the participants into two groups and asked them to recall the lists.

One group was asked to recall as many items as possible in a simple free recall and the other was

given category labels as a cues and then asked to recall as many items from each category. It was

found that participants given the category cues performed much better than those asked for free

recall. Had the experimenters not had the cued participants results to compare to the low scores

gathered through free recall would simply be attributed to forgetfulness. These findings suggest

that our memory works through a form of categorized schema and, like a library index, the best

way to retrieve information is to look in the proper categories. When applied to real-life learning

situations one can see how it would be possible to have information stored that will not be usable

unless the correct methods of cueing are utilized.