Memory is defined as the acquisition, storage, and retrieval of information. While many animals have instinctive behaviors, the human brain is most unique in its ability to quickly form memories. This development of memories provides for maximum behavioral flexibility and allows humans to most efficiently adapt to their environment. This occurs because the brain not only stores information about the environment but is able to modify these memories as circumstances change. Simply put, memories are not permanent but are capable of changing as associated stimuli in our environment change. This occurs when new or different information is presented.

This change can occur even from information from the individuals themselves in the form of thoughts (Loftus, 1996). The major point is that memories are formed as relatively permanent but are designed for change.

In spite of a long history of data supporting this fact, most lay people and many professionals think that memories are permanent. The brain is simply not designed to remember all events. Surprisingly, in the one case of documented total recall, the individual was dysfunctional. An example of total recall is reported by Luria in a case called S. He reports that S could not categorize information and was forced to deal with each event as separate entities.

While the exact nature of memory formation can't be observed in the brain, the general consensus of both cognitive psychologists and neuropsychologists is that memories are established in the brain as patterns of neuronal activity. The first stage of memory formation is sensory registration. This is a brief period of receptor activity. Certain events in our environment are registered as a pattern of neuronal activity that is longer lasting and referred to as short-term memory (STM). Critical or important memories, (those of emotional significance) in STM are consolidated into more permanent memories referred to as long-term memory (LTM). It is interesting that memory consolidation is controlled by structures in the brain that are primarily concerned with emotion (see figure 1).

One type of memory is rote memory. Rote memory is memory that is learned in isolation with practice. That is, the information to be remembered is not linked to existing memories but individual words are linked to themselves. The associations are formed between stimulus items. It is fairly limited and requires many repetitions for formation sufficient for later recall. Rote memories are more difficult to recall. Everyone who has had the experience of learning a poem or passage in school can recall the difficulty in learning the passage and if one segment is not recalled recall of the rest fails.

Association is important in all memory formation, however, in most cases new material is linked to old information by association. Stimuli that are presented in close approximation of space and/or time are associated with existing memories. This also occurs with meaningfulness rather than simple spatial and/or temporal summation. In this way, a new object or term is best remembered if the perceiver can question what it is or what it is like. In doing so, they are establishing neural connections with the Nervous System and existing memories. This association or categorization allows to not only modify memory but to group events and/or items as categories. This can be thought of as similar to constructing a brick wall. The individual memories are the bricks and each new one is laid down on the existing ones. Memory, like the brick wall, is constructed from the bottom up and loss occurs from top down.

The brain is confronted with too much information to remember and the important associations relate to meaningfulness. We don't try to remember all of the stimuli or words presented, rather, we attempt to extract the salient features from the story. This is referred to as gist memory; or remembering the gist of the story (Loftus, 1995).

While the term forgetting is useful, in most cases information is not forgotten; rather, new information is acquired which interferes with the old. This interference is greater in some areas than others. For example, if we perceive a visual stimulus, our visual system forms a pattern of activity for this stimulus. Little memory can occur in the primary visual system, however, because the same neurons must process all visual input. Visual memories, then, are formed in more remote association cortex where the pattern of neuronal activity is less likely to have interference. The same is true throughout the nervous system. Research has shown that information for which there is little or no interference can be retained almost undisturbed indefinitely. In our day to day activities such is seldom the case.

Memories are not permanent but change as a function of retrieval and association. If memories are analyzed in a different context, association with stimuli in this new context may change them. This is how psychotherapy works. It engages the individual in recalling early memories which are stressful and reanalyzing them in a low stress environment. In so doing, the memories are gradually changed by their association with new information or new interpretations. Likewise, we deal with stress by the use of defense mechanisms. Defense mechanisms work by changing our perception of stimuli by our own thoughts. An example would be rationalization. In this case, some condition occurs which is threatening to us and we reinterpret the situation. A specific example would be turning in a product and have the evaluator inform you that it is inferior. This is threatening to us and rather than accept this we might use rationalization and conclude that the problem is not with our product but due to inadequacies in the evaluator. This doesn't occur with one thought but over time. We think a great deal about this and tell ourselves about the evaluator's problems. We also tell others. Gradually, after enough thought and discussion we come to believe this. In effect, we have changed our memory about this episode.

Cognitive research has clarified the complexity of memory by indicating that memory is not some single unified phenomenon, rather, they have identified different types of memory. In addition, neuropsychologists have been able to associate these different types of memory with different areas of the brain.

Implicit memory is defined as remembering how to do things (function of the basal ganglia and sub-cortical systems). With aging and brain injury, this type of memory is the most stable, less likely to be lost and/or the first to recover.

Explicit memory refers to remembering facts (function of the hippocampus and cortical systems). It is most likely to be damaged with brain injury. Source memory refers to the location of facts or the source of the information is the weakest due to limited repetition.

Mechanical force to the brain can cause damage to neurons and impairs the ability to form new memories causing post traumatic amnesia (PTA) (Benson, 1967). In addition, this damage can impair the retrieval of memories causing retrograde amnesia (RA). The extent of disruption of either type of memory is a function of the severity of the force. In most TBI the injury is minimal to moderate and the memory impairment is short (PTA <7 days; RA <5 min.) As the severity of the injury increases (PTA 7 days), the RA increases proportionally from 5 hrs. to 12 hrs. in cases of very severe injury. In some cases of very severe injury the RA has been many years.

PTA and RA may extend over many months. This is further extended in cases where focal injuries to the brain occur with subdural hematoma and increasing areas of severe focal brain injuries. In severe cases, not only will retrieval and consolidation be impaired, but the neurons forming the basis of memory stores will be destroyed.

Memory consolidation (the ability to form new memories) increases with recovery of the diencephalon and is reflected in a gradual decrease in the level of PTA. RA does not decrease in the general level but appears to change or shrink from early memories to later ones or from the past forward. Ribot (1882) first reported this shrinking RA. Again this appears to be based on the fact that early memories are more established and represent the foundation for subsequent memories. Thus, once foundation memories are established, old memories may be recalled but more recent memories may remain inaccessible.

It is always difficult to determine what is being recalled and what is being reconstructed since during early stage recovery the individual has both impaired PTA and RA. They may repeatedly ask what happened and be informed by others what was thought to have occurred. Gradually they recall or reconstruct a memory of relevant events. It is difficult to determine whether information is recalled or whether it is reconstructed from information provided, but the speed of recovery suggests that both factors are operating. (Loftus, 1996, 1992 for misinformation studies and Read & Lindsay, 1994 for suggestibility studies).

During the process of recovery we attempt to interpret or give meaning to fragments of memory just as we do with dreams. We attempt to reconstruct memories from external cues and in the process the external cues become part of the old memory. Much of our memory is dependent on external cues (Vigodsky). An example of external memory aids would be remembering something, going into another room to get it, being unable to remember it, then returning to the original room where we got the memory. This often results in remembering and it is due to exposure to the stimulus that casued us to remember in the first place.

RA shrinks during recovery from PTA and is thus related to consciousness. Recovery is linked to feedback provided and this is difficult to distinguish between what is reconstructed and what has been told. This is a function of stability. Research has shown that children and older adults have weaker memories and are thus more influenced by what is told (memory distortion). While specific research is not available regarding TBI, it is clear that they are highly suggestible since their memories may be much weaker than either young children or older adults.

Memories are formed by the limbic system which is concerned with emotion. It is perhaps for this reason that highly emotional events are remembered best: A birthday or special event. With the recovery of post-traumatic amnesia, there may be islands of memory related to such emotional events. However, there is no evidence of islands of memory associated with shrinking retrograde amnesia.

It is clear from most memory research that RA is a shrinking process. The closer to the point of injury the weaker the memory. Things assumed to be remembered in detail are most likely a combination of repeated events and external input, not actual memories.

Bell, B. E., & Loftus, E. F. (1985). Vivid persuasion in the courtroom. Journal of Personality Assessment, 49, 659-664.

Benson, D. F., & Geschwind, N. (1967). Shrinking retrograde amnesia. Journal of Neurology, Neurosurgery, and Psychiatry, 30, 539-544.

Blomert, D. M., & Sisler, G.C., (1974). The measurement of retrograde post-traumatic amnesia. Canadian Psychiatric Association Journal, 19,185-192.

Corkin, S.H., Hurt, R.W., Twitchell, T.E., Franklin, L.C., & Yin, R. K., (1987) Consequences of nonpenetrating and penetrating head injury: retrograde amnesia, posttraumatic amnesia, and lasting effects on cognition. In Levin, H.S., Grafman, J., & Eisenberg, H.M. (Eds) In Neurobehavioral recovery from head injury. New York: Oxford Univ. Press.

Garry, M., Loftus, E. & Brown, S. W. (1994). Memory: A river runs through it. Consciousness and Cognition, 3, 438-451.

Goodman, J, & Loftus, E. F. Judgment and memory: The role of expert psychological testimony on eyewitness accuracy. In Suedfield, P., & Tetlock, P. (Eds.) In Psychology and Social Policy. New York: Hemisphere Publishing Co., 1992.

Kapur, N. (1993). Focal retrograde amnesia in neurological disease: A critical review. Cortex, 29, 217-234.

Loftus, E. F. (1996). Memory distortion and false memory creation. Bulletin for the American Academy of Psychiatry and Law, 24, 281-295.

Loftus, E. F. (1992). When a lie becomes memory's truth: Memory distortion after exposure to misinformation. Current directions in psychological science, 1, 121-123.

Loftus, E. F., & Loftus, G. R. (1980). On the permanence of stored information in the human brain. American Psychologist, 35, 409-420.

Loftus, E. F., & Rosenwald, L. A. (1995). Recovered memories: Unearthing the past in court. The Journal of Psychology & Law, 23, 349-361.

Markositsch, H. J., Calabrese, P., Liess, J., Haupts, M., Durwen, H. F., & Gehlen, W. (1993). Retrograde amnesia after traumatic injury of the fronto-temporal cortex. Journal of Neurology, Neurosurgery, and Psychiatry, 56, 988-992.

Read, J. D., & Lindsay, D. S. (1994). Moving toward a middle ground on the 'false memory debate': Reply to commentaries on Lindsay and Read. Applied Cognitive Psychology, 8, 407-435.

Richardson, J. T. Clinical and Neuropsychological Aspects of Closed Head Injury. London: Taylor & Francis. 1990.

Russell, W. R., & Nathan, P.W., (1946) Traumatic Amnesia, Brain, a journal of neurology, 67, 280-300.

Conclusions