Memory can be described as the ability to retain past experiences and the ability to recall these experiences. In many circumstances, memory is important in the professional and private lives of every living person. Remembering the skills needed to play football, remembering a loved one's birthday or other significant social event or, arguably even more important, remembering which side of the road to drive on and the importance of the laws of gravity.

This entry is designed as a short explanation of the psychology of memory and forgetting.

Memory

There are three main types of memory: iconic and echoic memory, short-term memory; and long-term memory. They can be categorised by their respective methods of encoding, storing and retrieving information.

Iconic and Echoic Memory

Our iconic and echoic memory acts as a buffer for input from the senses¹ before the information passes on to the short-term memory. The encoding in iconic and echoic memory is precategorical: the information is stored for a very short time then retrieved serially, by going through every piece of information in order, and transferred to short term memory. Iconic and echoic memory lasts for about one second, before starting again. We are usually unaware of such memory because the information is constantly lost and replenished.

The most important point to remember about iconic and echoic memory is that it acts as a store for holding information while it is being put into categories, before moving on to short-term and long-term memory. Iconic and echoic memory decays very quickly if it does not enter short-term memory.

Short-Term Memory

Short-term memory (STM) is our 'working memory' during the processing and learning of information. It encodes information categorically and is largely verbally-based. STM stores information for around 20 seconds and this information can be maintained by rehearsal. It also has a limited capacity of (7 +/- 2) things — such as letters, words or numbers.

Information is retrieved from STM by means of a rapid serial readout, which goes through every piece of information in STM and then comes back again with an answer. In STM information is not always being recalled from STM alone. Some previously-learned information may be coming from long-term memory.

The Suffix Effect is very important in STM. This is when a piece of information which is learned at the end of a learning session replaces a piece of information learned at the start. For example, when learning a simple list of numbers (1, 2, 3, 5, 6, 7, 9, 4), the number nine or four might eventually replace the number one.

Long Term Memory

Long-term memory (LTM) differs from STM in two key ways: first of all, it is permanent and durable. Secondly, LTM does not decay over time, unless through a condition such as Alzheimer's. Information in STM is either replaced by new information coming in or is permanently stored in LTM — nothing is permanently in STM.

Information is encoded in LTM based on meaning (for example, the meaning of 'dog': animal, canine, perhaps pet, pet names). Information is stored in LTM based on the time the information has spent in STM (the longer information stays in STM, the more likely it is to get into LTM) and is stored in the context in which it occurred (say you learned information in a class, taught by a teacher — this information would be grouped with the information learned) with no time or capacity limit. Information is retrieved from LTM by some sort of search, finding a path to the memory via associations.

However, there are some problems within LTM. Not all LTM memories are stored with context; information is not always stored relating to time in STM and some information gets lost (for example, there is some evidence that people can, over time, forget the associations but retain the actual information).

Information in LTM can be split into two sections: episodic (autobiographical, stored relating to time and place) and semantic (general conceptual knowledge, such as the meanings of words, not stored relating to time and place).

Forgetting

A common belief is that our memory decays over time. This is called decay theory: the breakdown of information unless it is preserved by practice and reiteration. It is the reason for loss of information in iconic and echoic memory, but only partially in STM and not at all in LTM.

One theory of forgetting is that the information is simply 'displaced' by new material coming into STM, due to its limited capacity — and because of this, the information didn't make it into LTM. Waugh and Norman (1965) explored this through a 'serial probe task'. The participants were shown a sequence of 16 numbers at either four or one per second and were then shown one of the digits (the 'probe') and asked to say which number followed. The idea was that if the number was at the start of the task it would have been displaced by a later number, but if it was at the end then it should have been remembered. The results supported this theory. However, in the four-per-second task, the numbers were remembered more easily — this also ties in with decay theory.

Retrieval failure theory suggests that information is not lost, but that the right cues are not being used. An example of this is childhood amnesia. This is when an adult cannot remember events before their fourth or fifth birthday, which can be explained by understanding that things around them and the way they see things would be very different to a young child, compared to when they became an adult, and thus they have no cue to remember the information. A 'tip of the tongue' study was done by Brown and McNeill (1966). In this, they read out dictionary definitions of unfamiliar words to participants and most either knew the word or knew that they did not know the word. However, there were some who did not know whether they knew the word or not. Some of these people could give the word's first letter and even how many syllables it had, but could not remember the word. It can be seen from this that the words were in their memories, but they had not used the correct cue to access it. This cannot be explained by other theories, such as decay theory.

In interference theory, forgetting occurs because of what has happened before or after learning. Retroactive interference is when information learned after learning interferes with recall: for example, if Jim learned about the D-Day landings and then learned about the First World War, the information about the First World War might interfere with his fully recalling the information about the D-Day landings. Proactive interference is when information learned before learning interferes with recall. If John learned ten numbers and then learned ten completely different numbers, the first ten numbers might prevent him from fully recalling the second ten when asked.

When the information being recalled is similar in each trial, interference can be prominent, but when the topic of the experiment changes (for instance, if the questions in the first three trials were on recalling dog breeds and then the fourth trial changed to recalling numbers) there would be, as in the first trial, a high rate of recall.

These reasons for forgetting, of course, do not apply to conditions such as Alzheimer's and amnesia, which are separate, biological reasons for forgetting.

One very important thing to remember is that information never leaves our brains after being lodged in LTM (except, of course, for biological reasons). However, retrieval cues can get harder to find the further we are from them. An old woman in a nursing home is less likely to remember her childhood than an old woman in her childhood home.

Remembering, Learning and Improving our Memory

The best ways to learn and remember information are to:

• Keep it organised
• Associate it with something funny, elaborate or easy to remember

One way of doing this is by using mnemonics or memory stories, particularly when revising. Rehearsal (just going over it, over and over again) is not a good way of learning.

Another way of ensuring recall is by reinstating the conditions of learning. You might go to where you last remember having your towel to recall what you did with it, or take an exam in the room in which you studied. Another example is if you were drunk or underwater when you said or did something, you are more likely to remember what it was if you got drunk or went underwater at the same place and time.

Hypnosis has also proved to be a useful way of remembering information lost in LTM. However, the person doing the hypnotising must know the correct retrieval cues and hypnosis does not work on everyone.