Thinking : Memory, Forgetting


Thinking about the way we think can be a difficult pursuit. We often make the statement that 'I have forgotten'. This is interesting because to know that you have forgotten means that you still have some memory. Something that you had truly forgotten would have no meaning for you. It is difficult to say whether things are lost from memory or simply not accessible since the effect of the two is the same to us.

It is thought however, that some forgetting does indeed occur. It appears that we remember skills longer than we remember facts or cognitive knowledge. The mechanism of forgetting is not really clear. Forgetting could simply occur through gradual fading of memory or memories could be displaced by new learning.


There is great confusion between the conflicting explanations for failure to recall. It is of course possible that people never actually acquired the knowledge and only think that they did. Perhaps more commonly, the conflict arises between what is failure to remember and what has really been forgotten. Some believe that nothing, once learned, is ever forgotten. It may be that reorganisation of our memory makes some things inaccessible but these memories are nevertheless still there if the pathway to them once again becomes active. It is just as likely however, that knowledge is indeed lost as part of the normal process of updating memory. If we never actually forget anything, this has serious implications as to the capacity of human memory.

Is the truth then, that things are forgotten or is it that access is denied but the memory is still there. How could we know which of these applies to our failed attempts at remembering something?


Ebbinghaus was one of the first to carry out an objective study of the rate of forgetting. He conducted experiments by trying to remember sets of nonsense syllables.

When testing his recall he found that he could remember after intervals ranging from 21 minutes to 31 days. He used the time it took to re learn lists as a measure of how much had been forgotten. His experiments showed that forgetting occurs rapidly at first (say up to 1 hour) and then levels off to a more steady decline (experiments show up to 31 days). About 56% of the information was lost in the first hour and about 20% still retained after 31 days. (see diagram in book 2 page 106)

Warrington and Sanders showed that there is significant forgetting for memory of popular news items. They also showed that younger people had better memories for distant as well as recent events.

A study by Bahrick, Bahrick and Wittinger, using 392 high school graduates looked at their ability to recall names and faces of old class mates. The study showed that even after 30 years, there was good memory for faces when asked to pick out a class mates face from a set of unfamiliar faces. However, memory for names was much worse. The study also showed that significantly more forgetting occurred after 50 years.

Another study looked at people who had learned Spanish in college but not used it since (or hardly used it). The forgetting curve (Book 2 page 108) showed that most forgetting occurred after 5 years and that there was hardly any more forgetting until about 30 years when a steady but shallow decline in memory occurred.


Dali's Persistance
              of MemoryOne theory is that memories simply fade with time.

Experiments with insects have shown that information is remembered for longer if the insect is immobilised for the period between learning and testing. Insects which carry on as normal forget more quickly. Unfortunately, it is not easy to extend this to humans. Sleep in humans is more complex than immobilisation in insects. Several researchers have attempted to show that sleep and dreams are necessary to organise or consolidate memories but their results are not conclusive.


Experiments have been more conclusive in showing that interference does indeed affect memory. The more similar are items to be remembered, the more forgetting occurs. This may be why people who are learning are often advised to switch topic regularly in order to maximise their learning effort.

Two types of interference are known to have an effect. These are:

  • Retroactive Interference is where old information is forgotten through interference from new information.
  • Proactive Inhibition is where a new memory is disrupted because of an older,similar memory.

The existence of Proactive Interference implies that we store more than we can recall at any particular time. The phenomena of knowing that you know something but not being able to bring it to conscious thought is a common one. In general, people are correct when they believe that they know something, even if they cannot remember it at the instant.

Wilder Penfield, an eminent neurosurgeon, reported the evocation of memories following direct electrical stimulation of the cortex during operations when patients were conscious. A total of 40 patients reported memories of previous incidents, often in great detail, when areas of the temporal lodes were stimulated. It was suggested at the time, that these flashbacks were accurate and complete memories of earlier events, brought fully into view through direct brain stimulation.

It is possible however, that patients (relatively few reporting detailed flashbacks) were perhaps cobbling memories together from various other incidents. There is little evidence to justify the complete recall interpretation from the study. It is known that human memory is not always accurate and true and that people can easily put memories together from separate experiences.


Skills are learned differently from cognitive knowledge. Skills require constant and regular practice until they are learned. Driving a car is very difficult at first but is performed effortlessly when mastered.

There is much anecdotal information about memory for skills such as you never forget how to ride a bicycle. How true is this? It appears that where skills involve closed loop learning where actions depend on current sensor states such as balance, then very little forgetting occurs. However, for open loop skills such as typing, forgetting does occur and regular re-training may be required if the skill is not employed on a frequent basis.

One study showed that training for 215 shop and office workers in the skill of cardiac resuscitation declined rapidly in the first 12 months and reached a very low level after 3 years. This is a complex and partially open loop skill. In such cases, re-training programmes are required.


Your Memory: A users Guide: Alan Baddeley Penguin 1993

Human Memory: Theory and Practice Alan Baddeley, IEA 1993