[excerpt from Schacter's Searching for Memory] 

The intrustive recollections that plagued Melinda Stickney-Gibson, Jadzia Strykowska, and numerous other people have experienced profound traumas underscore that emotionally overwhelming events are frequently our best-remembered experiences. And the research I considered in the last chapter is beginning to uncover the reason why this so. But under special circumstances trauma is associated with far-reaching amnesias. What accounts for such a seemingly paradoxical state of affairs? This question has assumed paramount importance during the past few years because of the controversies concerning recovered memories of childhood sexual abuse (which I examine in the next chapter).

Heated debates have raged about whether or not people can temporarily forget terrible traumas. Students of memory mostly clinicians, have reported instances of trauma-related forgetting for more than a century. Unlike the material I have considered in previous chapters, these examples of traumatic amnesia have not played a major role in shaping the new conceptions of memory that have unfolded in recent years. When we enter the world of trauma and amnesia, we encounter exotic cases of fugues and multiple personalities that populate the edges of memory research. We do no yet understand such cases very well. But some of the techniques and ideas I have considered are beginning to illuminate these strange manifestations of memory’s fragile power. Given the urgency surrounding questions of trauma and amnesia in our society, we must examine carefully some of the peculiar instances in which a person’s past seems to vanish without a trace.

I became involved in such a case back in September 1980. I was then a graduate student at the University of Toronto when a call arrived from Dr. Paul Wang, the clinical psychologist who just months earlier has arranged months earlier for me to test a fascinating head-injured patient who had give me a firsthand glimpse of implicit memory. Dr. Wang informed me that a most remarkable new patient had just been admitted: a young man who did know his name, where he lived, or just about anything else about his personal past. About all he could say about himself was that he had once been called by the nickname Lumberjack. Lumberjack, Dr. Wang continued, had approached a policeman in downtown Toronto two days earlier complaining of excruciating back pains. He was taken to the hospital, but when he arrived the young man was not able to identify himself, nor was he carrying any identifying information. A local newspaper printed his photo the next day in an effort to locate family members. Dr. Wang wondered whether I might be interested in testing Lumberjack’s memory.

As a graduate student with interests in both normal and abnormal memory, I found such an invitation irresistible. Lumberjack appeared to be suffering from psychogenic, or functional, amnesia.: a temporary loss of memory precipitated by a psychological trauma. Psychogenic amnesias involving loss of personal identity have been reported in the psychiatric literature for at least a century, and they are often depicted on television and in the movies. But in reality they are quite rare. Psychiatrists have estimated that extensive functional amnesias that cover large sectors of the personal past occur in less than 1 percent of psychiatric patients. The incidence rate may be somewhat higher during wartime, when combat-related stress can produce temporary amnesic episodes in traumatized soldiers. One study of soldiers who were hospitalized during World War II , for example, reported a 14 percent incidence of various forms of psychogenic amnesia. Soldiers who had been subject to heavy fire were more likely to present with amnesia than those who had no combat exposure.

Motivated by pure fascination with the subject, I had spent hours hunting through remote library stacks of musty, turn-of-the-century journals in which Pierre Janet, Morton Prince, and other psychiatrists had described their impressions of patients who responded to overwhelming stress or unbearable disappointments by blotting out much of the past. Some wandered for days in what psychiatrists call fugue state, in which a person is totally unaware of having lost all knowledge of personal identity. Patients in fugue states often focus their attention exclusively on achieving a specific goal, such as finding their way to a particular destination. In one case of a wartime fugue, an Australian soldier serving in Africa during World War I1 became traumatized when a German fighter plane came swooping down at him. He remembered trying to fire at an approaching dive bomber, then he "blacked out." Thirty-two days later he "came to" in a Syrian hospital, hundreds of miles away. After the bombing incident, he had become totally focused on seeking refuge near a camp he had heard about in Syria. He wandered in a fugue for over a month, not knowing who he was or what he was fleeing, until he became aware of his memory loss in the hospital)

Patients in fugue states are generally oblivious to their disconnection from the past until a situation arises that requires them to identify themselves or to provide information about their background and experiences. Lumbejack was in just such a state prior to entering the hospital. He had been wandering the streets of downtown Toronto for more than a day. It was only when hopital personnel asked him to identify himself that he realized, much to his surprise, that he could not.

Back in 1980, there was not a single controlled investigation of memory retrieval during an amnesic episode of the sort that Lumberjack was experiencing. Here was an opportunity to break new ground by carefully studying Lumberjack with scientific procedures. But I would have to move quickly, since psychogenic amnesias frequently clear up within a few days, and Lumberjack had already been in the hospital for more than two days by the time I learned about him.

When I met Lumbeuack the next day in his hospital room, I encountered a quiet young man with stringy blond hair who appeared mildly embarrassed by his inability to recall his past. His IQ was in the normal range. He had some difficulty recalling stories and pictures that were presented to him, but his memory for ongoing experiences did not seem to be seriously disrupted. Lumberjack could recognize faces of famous people and use vocabulary normally, showing intact semantic memory.

To find out more about his ability to recall episodic memories, I used the Crovitz technique described earlier: I read out to Lumberjack a series of common words, such as tale, hurt, and run, and asked him to try to think of a particular experience from a specific time and place that was triggered by the cue word. Normal young adults retrieve memories ranging from the immediate past to the early years of childhood, but over 90 percent of Lumberjack's memories came from the two days since he had been admitted to the hospital. He could remember little else.

I noticed one intriguing feature of the few memories that Lumberjack was able to recall from his prehospital life: they were largely restricted to a time period about a year earlier, when he worked for a courier service. When I probed, Lumberjack provided detailed recollections of specific incidents that had occurred during his time there, and he remembered a great deal about his fellow employees and what they did. I had apparently managed to stumble upon a preserved islandof memory in a vast sea of amnesia.

This memory island turned out to be a key feature of Lumberjack's amnesia. I contacted the courier service to confirm Lumberjack's memories and learned that employees there had come up with the name "Lumberjack"--and that this had been the only time in his life he had been called by that nickname.

Why was Lumberjack able to remember this particular period and no others? His time at the courier service, he said, was one of the happiest in an otherwise difficult and sad life. It emerged later that Lumberjack had been abandoned by his parents as a young child and had been raised almost singlehandedly by his grandfather. Lumberjack's life appeared to consist of a series of disappointments, rejections, and failures. At the courier service, however, he was liked, accepted, and successful. The one happy period in his life seemed somehow immune from the amnesia that had hidden just about everything else.

Lumberjack's amnesia cleared up the evening after I tested him. While watching the television rendition of the novel Shogun, Lumberjack began to recall during an elaborate funeral and cremation scene that he, too, had recently been at a funeral: his grandfather had died a week earlier. He then remembered his real name and, during the next several hours, managed to recover and piece together the rest of his past.

The death of Lumberjack's grandfather--the single most important person in his life--had apparently triggered the amnesia. Lumberjack eventually recalled going to his grandfather's funeral, and leaving it in a state of shock and grief. He recalled nothing else after that until twenty-four hours later, when he approached the policeman. Even after the amnesia cleared up, Lumberjack did not recollect anything that had transpired during the day or so when he walked the streets of Toronto in a fugue state--and it is unlikely that he ever will. In most cases of psychogenic amnesia, patients eventually recover their entire personal past with the exception of what happened during the fugue state.

When I saw Lumberjack again several weeks later, he said he had felt "stupid" about being unable to remember his name and so much else about his past that day. Now that I could compare Lumberjack's performance in his normal state with his performance during the amnesic period, the results were clear-cut. His IQ and recognition of famous faces remained unchanged, but he was now able to recollect episodes and experiences from many parts of his life. He could barely contain his happiness as he showed me that he, like everybody else, could travel in time and tell the story of his life.

WHEN THE MIND FORGETS ITS SELF

Beyond Lumberjack

Because there have been so few controlled studies of memory, during episodes of psychogenic amnesia, it is difficult to say whether the constellation of features that characterized Lumberjack--loss of explicit memory for individual episodes and other personal information, a preserved island of autobiographical recall involving a specific lifetime period, and excellent retention of nonpersonal, semantic memory—is typical of other patients. The neurologist Marc Kritchevsky and colleagues recently reported that ten patients with psychogenic amnesias involving loss of personal identity performed just like Lumberjack on the Crovitz task, recalling many episodes since the onset of their anmesias and virtually none from before. But only half of these patients recognized famous faces normally; the other half, in contrast to Lumberjack, performed poorly on this test of semantic memory. All of them had problems recalling specific public events (for example, Who killed John Lennon?), a task that probably draws on both episodic and semantic memory. In contrast to Lumberjack, some of the patients remained amnesic for weeks and months. These findings indicate that no single profile characterizes all patients with psychogenic anmesia and loss of personal identity. This should not be surprising, because such amnesias are no doubt influenced by idiosyncratic features of each patient's psychological history and present conflicts (See figure 8.1.)

These points are illustrated by one of the most bizarre cases of amnesia ever reported. I first learned about K. in April 1986, when I received a letter from one of his physicians. This fifty-three-year-old married man had been found sitting on his kitchen floor, silent and dazed· He held in his hands a defective electrical element from a 220 volt oven, but his body was neither scarred nor burned. In an ambulance on the way to the hospital, K. began to speak. He was confused and said that he had a terrible headache from being hit on the head with a baseball bat. It was later revealed that he had indeed been hit on the head with a bat in a Little League game when he was fourteen years old. K., however, believed he was still fourteen years old and had no memory for anything that had happened to him after that age. He failed to recognize his wife and children. He believed he was still living in his childhood town, felt shock and dismay when he learned that his father had died, and was taken aback by how old his mother looked in photographs· He was equally surprised by the sight of his own face in the mirror and was amazed that he needed to shave every day."It was as if," wrote his physician, "Rip Van Winkle had awakened."

This extraordinary amnesia was unlike anything I had ever heard about before. A team of researchers at Johns Hopkins University who tested K.'s memory found that his amnesia was not confined to autobiographical recollections. He showed no memory for famous people or public events after 1945, although he easily remembered those before 1945. He was unfamiliar with, and amazed by, electronic devices such as televisions and VCRs. Incredibly, K. had lost the ability to execute skills he had acquired after 1945, including such basic ones as driving and shaving. Yet he had no difficulty remembering ongoing, day-to-day events. Extensive neurological examination failed to reveal direct damage to the brain. It turned out, however, that K. had been under severe job-related stress. He had been experiencing breathing difficulties and chest pains, and was on a disability leave for these psychosomatic problems at the time his amnesia occurred.

Why did the amnesia affect the lifetime period beginning abruptly in 1945? K. experienced several salient life changes shortly after that year, including a family move, change of schools, death of a close grandparent, and a fire that destroyed his family's uninsured house. Moreover, World War II ended in August 1945, signaling the end of one lifetime period and the beginning of another. K.'s amnesia provides further evidence that lifetime periods serve as organizing structures in memory, helping to separate different constellations of experiences from one another. Although K. differs from Lumberjack in many respects, in both patients some lifetime periods were resistant to anmesia-

The prospect of faking must be taken seriously in certain cases of psychogenic amnesia; indeed, some warning signs often point toward a simulated amnesia. If a patient has been charged with a criminal offense, is attempting to escape financial difficulties or some other legal obligation, or stands to benefit in some material way from the·amnesia, then there is certainly cause for suspicion.

In genuine cases, psychogenic amnesias may result from the combined effects of brain damage and emotional trauma. A considerable proportion of patients have a history of prior head injury or some other kind of brain abnormality. Lumberjack, for example, was involved in a car accident when he was four years old that damaged his right temporal lobe. This damage was serious enough that a brain scan administered during his hospital stay clearly revealed it. K. was hit over the head with a baseball bat at age fourteen. Several other cases of functional amnesia reported during the 1980s also include a history of head injury or brain damage. In fact, nearly all early psychiatric studies of psychogenic amnesias and fugues published from the 1930s through the 1950s indicate that many patients had once suffered brain trauma or disease.

Knowing that does not tell us much about why psychogenic amnesias present in the way they do. Psychogenic amnesias are rare, and it is safe to assume that most people who once suffered brain damage do not end up like Lumberjack or K. And we have already seen that memory is affected in specific ways by damage to different regions of the brain; the general notion of "brain damage" is too coarse to be helpful in Illuminating psychogenic amnesias. But further on in the chapter, l will discuss how damage to specific parts of the brain could alter a person's response to psychological trauma occurring later in life to produce a highly unusual outcome: extensive amnesia

Alan J. Parkin and Hans G. Stampfer

[excerpt from Broken Memories]

The study of disordered memory has been primarily confined to people who have suffered irreversible deficits following some form of brain injury or disease. However, a much larger number of individuals suffer temporary memory impairment which is presumed to be 'psychogenic' or 'functional' in origin, and is invariably associated with wider psychiatric disturbance. This group has been relatively neglected in formal neuropsychological studies, although there is a considerable clinical literature on individual cases and a large body of 'psychodynamic' theory about the nature of such impairment and the mechanisms that might be involved (Kopelman, 198'7; Parkin, 1987; Schacter and Kihlstrom, 1991).

There is a close association between memory impairment and psychiatric illness. In day-to-day practice, complaints of 'poor memory' often emerge spontaneously or in response to specific inquiry (for example, in different forms of anxiety and depression). Intermittently, 'amnesia' is the presenting complaint (for example, dissociative reaction, fugue state and multiple personality).

After orientation in time, place and person, the testing of memory is probably the most common 'bedside' cognitive assessment undertaken by any duly qualified clinician. In the main, the primary concern is to distinguish 'organic' from 'functional' impairment. 'Organic' means that the memory impairment is due to some 'physical' fault caused by fixed or reversible damage to brain regions directly involved in mediating memory - and the most immediately important reason for distinguishing 'organic' from 'functional' memory impairment is to diagnose reversible or treatable brain damage, caused, for example, by vitamin deficiency or hypothyroidism.

The term 'functional' could simply be taken to mean the converse of'organic'; namely, that the memory impairment is not due to some 'physical' brain damage. However, more complex issues are involved and this is also true of the seemingly straightforward concept of 'organic' impairment. For example, there may be 'functional' impairment of memory which is a secondary manifestation of 'organic' damage to parts of the brain not directly involved in mediating memory. The primary problem is physical or structural brain damage, but this has led secondarily to functional system impairment in certain parts. Similar difficulties are involved in the concept of 'functional'. The primary implication of 'functional' is that the impairment is not due to physical damage. This usually implies (although not necessarily) that the impairment is reversible, however, the nature of 'functional' impairment may depend on when one looks and what coexisting factors are relevant at the time. More specifically, if it is recognized that long-term motor immobility (including voluntary and enforced 'functional' immobility) can lead to physical disuse atrophy and irreversible fibrosis, is this also possible in the realms of emotional and higher intellectual functioning? We think not, but we don't know.

In this chapter we describe a case of presumed functional memory impairment associated with an atypical psychotic disorder characterized by prominent 'negative' symptoms that are usually associated with schizophrenia (Andreasen, 1982). The impairment was presumed to be 'functional' in origin, in that thorough clinical assessment with appropriate laboratory investigations failed to demonstrate any organic cause. The patient's illness and memory impairment improved and there was evidence of premorbid personality vulnerabilities linked to an emotionally traumatic upbringing and adult life.

Serial neuropsychological testing was undertaken on our patient Elizabeth during her three-month admission, and we believe that the presented findings show the potential benefits of incorporating experimental neuropsychological test data in psychiatric re-search and day-to-day clinical practice.

Elizabeth

Elizabeth was born in Western Australia. Her birth was unproblematic and she achieved developmental milestones normally. Although of good intelligence, she left school at fifteen and obtained a trade qualification. Her home life was dramatic. Her parents separated when she was three due to problems with her mother's alcoholism and her Father's persistent gambling. Her mother remarried, but this was to another alcoholic. At eighteen she had her first serious relationship. It was a traumatic time which culminated in her being badly stabbed. In her early twenties she married an alcoholic. A child was born but she had a rare congenital illness. Although given only three days to live, the child lived to be 2.5 years during which time Elizabeth had to resuscitate her a number of times. Finally she left her husband to live with another man and they now have a daughter.

Elizabeth was referred by a psychiatrist as someone who 'couldn't move.., couldn't cope'. She gave a history of 2-3 months of increasing disorganization, anxiety and depressed mood She felt her problems began after seeing a hypnotherapist in order to give up smoking. Since then she had never really come out of being hypnotized. Initially she had experienced racing thoughts and had felt more alive, but very soon she became unable to complete daily tasks, had difficulty thinking and making decisions, began to feel frightened to leave the house, and was losing track of time. She would wander around the house not really knowing what she was doing, and had been found on numerous occasions by her partner lying in the foetal position. On one occasion following hypnosis, she reported seeing black smoke coming out of the 'pores of her body' and had been told that depression was black and would come out of her. On admission the diagnoses considered were a dissociative state, an organic brain syndrome, or a delayed grief reaction. Over the next four weeks she reported feeling progressively worse, with feelings of hopelessness. She wanted to be in a long-stay mental hospital and began requesting electroconvulsive therapy. Four weeks after admission, while on weekend leave, she attempted suicide, slashing at her arm with glass and also being found with a blow-dryer and a bucket of water. She said she wanted to commit suicide because she did not think she would get better. Little improvement was noted over the next five weeks. Again on weekend leave, she attempted suicide and continued being preoccupied with suicidal ideation. Ten weeks after admission she began to feel subjectively better. This improvement coincided with her return from three weeks' holiday with her partner. Earlier in the admission her relationship had been strained. A steady progress ensued. There was no more suicidal ideation and she was noted to be caring well for herself, her daughter and partner while on leave. She was discharged twelve weeks after admission and remains well.

Neuropsychological Investigations

One feature of Elizabeth's case was particularly fascinating. She claimed to have great difficulty remembering the past. On one occasion she remarked that everything in her life had been fine until her present illness developed - a somewhat odd statement, given the range of negative life events she had been through. She also remarked that even people well known to her seemed odd: she knew who they were, but they somehow did not seem familiar. This feeling of alienation caused her particular problems with her family.

We decided to examine her impairments more closely, but all we had available were the tools of the neuropsychologist. None the less these provided some useful insights into her difficulties and, possibly, some pointers towards the organic basis of her psychogenic disturbance.

A Neuropsychological Examination

Elizabeth's estimated premorbid IQ_(Nelson, 1982) was 106, but on the Weschler Adult Intelligence Scale (Wechsler, 1981) she had an IQ_of only 78. Impaired performance was widespread, with performance normal on only two tests. On the Weschler Memory Scale (WMS: Wechsler, 1945) she also showed some impairment - particularly on those subtests most sensitive to amnesia: logical memory, paired associate learning and visual reproduction - where she scored 9, 13.5 and 6 (corresponding control scores are 12, 16 and 15). She was given the Rey Figure (Rey, 1964) to copy (see figure 6.1). Copying was accurate but her execution was bizarre in that she failed to take account of the overall organization of the picture. Furthermore she could remember very little about it after five minutes. Figure 6.2 shows Elizabeth's attempt to draw a bicycle: while not all the bicycle drawn by controls might be considered roadworthy, Elizabeth's problem with this task is self-evidant.

Elizabeth's unusual copying ,of the Rey Figure is considered indicative of frontal lobe function (Ogden ct al., 199(I) and her difficulties with the draw-a-bicycle test also merit a similar interpretation. Other testing confirmed this. Her word fluency, the number of words she could generate beginning with a given letter, was substantially impaired (a total of 23 words compared with a control average of 44) and she had very marked problems with the Wisconsin Card Sorting Task in that she was unable to understand what was required.

Other tasks emphasized her substantial memory deficit. The Brown-Peterson task involves the presentation of target stimuli followed by a distracter interval, then instructions to recall the target. Figure 6.3 shows Elizabeth's immense difficulty with this task. She was also markedly impaired on the immediate free recall of simple twelve-word lists, scoring 4, 2, and 0, and performance did not improve much when the words in the list were drawn from obvious categories. She was also given a number of tests of recognition memory and here a different picture emerged. On the Warrington Test (Warrington, 1987) she scored 82 per cent correct on the words and 84 per cent correct on the faces version of the test. On a 36--item test of yes-no recognition (Gardiner, 1988) she scored 56 per cent correct with two false alarms (control range 52-92 per cent), and on an additional 30-item yes-no test of recognition designed by John Dunn and Klm Kirsner (personal communication) she scored 77 per cent correct. Overall, therefore, her recognition performance was quite good especially when one considers how poor her recall performance was

As we noted earlier, Elizabeth complained of being unable to remember her past, and we investigated this formally using the autobiographical cueing technique devised by Robinson (19761. Elizabeth was given a number of cue words and asked to recall a specific incident in her life in response to each one. This revealed some access to her turbulent past the cue 'break' brought recollection of an incident in which she broke a pizza over her husband's head. However, the number of memories produced, 58 per cent, was well below the control rate of 96 per cent. There was also evidence of repressed memory. Just two days earlier Elizabeth had slashed her wrists but, in response to the cue word 'cut', she related an incident about cutting her thumb when she was eight.

Elizabeth's remote memory was also evaluated using the Famous Personalities Test devised by John Dunn at the University of Western Australia. In the first phase Elizabeth had to identify which of a group of four names was that of a famous person. Next she had to explain why that person was famous. Figure 6.4 shows that Elizabeth was as good as the controls (mainly library staff of her age) at identification (note that this was not because the test was easy - some of the names were relatively unfamiliar, for example, Sally Ride, Sidney Nolan, Amelia Earhart. However, she was much poorer at retrieving the occupations of correctly identified people.

The WMS revealed that Elizabeth bad a marked problem with learning unrelated pairs of words. We explored the possibility that this deficit might be overcome by teaching her to use images as links between the words. Instructions to use imagery greatly improved her ability to learn the word pairs and, more remarkably, she was able to retain them for twenty-four hours. However, it was disappointing that on a subsequent test using a similar task she failed to use the imagery strategy spontaneously.

A final test procedure was temporal discrimination (Squire et al., 1981). Elizabeth was shown two lists of sentences separated by a two-minute interval. She was then given a recognition test and, if she identified a sentence, she was asked to say whether it came in the first or second list. Her recognition performance was quite good, but her temporal discrimination was at chance levels. Interestingly, it did not improve when the lists were distinguishable in some obvious way, for example, when all the first list sentences contained an arabic number, and all the second list contained numbers in spelt out form.

As noted earlier, Elizabeth's symptoms lasted only twelve weeks. As her negative symptoms lifted it was notable that her frontal signs also reduced markedly, thus suggesting a direct link between frontal pathology and her impairment.

Discussion

Following a traumatic life history Elizabeth underwent a psychotic episode which had, as its major outcome, a temporary period of negative symptoms. Prominent among these symptoms was a pronounced impairment of memory. In attempting to explain what has happened we can approach the problem at two levels. First, we can ask what kind of mechanism was responsible and, second, we can try to explain why the mechanism was called into action.

A proposed mechanism for Elizabeth's memory deficit

It is instructive to begin with a summary of Elizabeth's memory impairment. The principal features were:

1. a decline in intellectual function;

2. a gross impairment of memory when measured by various forms of recall. However, recognition memory appeared reasonably intact;

3. an impairment in temporal discrimination;

4. evidence of disturbed frontal lobe dysfunction.

The general pattern of impairment shown by Elizabeth is not unique and has been described in a number of patients who have suffered focal lesions of frontal lobes and associated structures (Petrides, 1989). Deficits in temporal discrimination in the presence of relatively normal recognition are, for example, a typical finding in frontal patients (ibid.). The pattern of memory impairment associated with frontal damage has been termed the 'dysexecutive syndrome' (Baddeley and Wilson~ 1988). As the name suggests, this disorder is thought to affect higher-level functions in memory rather than more basic functions such as consolidation.

In a recent study Parkin et al. (1994) have described a man known as CB, who developed memory difficulties following rupture and repair of the anterior communicating artery (ACoA). Almost invariably, this kind of event leads to frontal lobe damage (see Parkin and Leng, 1993) and this was clearly evident in CB. tie had markedly reduced word fluency and showed disorganized copying of the Rey Figure that was similar to that of Elizabeth. His memory abilities were widely investigated, but for our present purpose it is important to note that, like Elizabeth, he experienced peculiar difficulty on tests of recall relative to recognition (for a further similar case, see the case of Hanley et al., 1994).

That recall should be particularly affected by damage to the frontal lobes is explained by proposing that recalling information, particularly in the absence of external cues, presents a considerable problem-solving task in that hypotheses concerning the nature of target information must be constructed. In contrast, recognition presents far less of a problem because the basis of the information to be evaluated is actually presented and does not need to be retrieved. This account of memory has been particularly developed by Shallice (1988), who has argued that executive processes in memory involve a description phase in which a hypothesis is formed about the nature of stored information, and a subsequent matching and verification stage in which retrieved information is evaluated. Within this account deficits can arise either because the description phase fails to produced candidate information or because the matching and verification stages fail to confirm the identity of generated target information. This arrangement can produce two fundamental patterns of breakdown, and both have been observed. A failure in the description phase would be expected to produce poor recall but not necessarily poor recognition the argument being that recall cannot be effective without the ability to generate candidate descriptions of to-be-remembered information. Recognition testing, however, may provide sufficient conditions for an item to be matched and verified without any additional derivation of a description.

A failure in the matching and verification phase predicts a different form of deficit, one in which descriptions are made available but are not adequately evaluated. This pattern of performance was observed in the study of a post ACoA patient, RW (Delbecq-Derousne et al., 1990), in which it was shown that the patient had normal levels of recall but only at the expense of recalling large numbers of incorrect items. This suggested an impaired ability to evaluate the memory-based qualities of retrieved information and this was confirmed by the patient's surprisingly poor performance on the Warrington Recognition Test.

Other aspects of Elizabeth's impairment are also consistent with a frontally based dysexecutive deficit. Elizabeth performed exceptionally badly on the Brown-Peterson test and there is now good evidence that this test loads on frontal function. Stuss et al.(1982) found that patients who had undergone frontal leucotomy performed very badly on the Brown-Peterson task even though they performed similarly to controls on other measures of memory. More recently, Parkin et al. (1993) described poor performance on this task in a young woman exhibiting marked frontal problems following recovery from Wernicke's Encephalopathy. Finally, converging evidence for a frontal locus to impaired Brown-Peterson performance comes from a study of age differences, in which the poorer performance of older adults was predicted by impairments exhibited on test of frontal function (Parkin and Walter, 1991).

The manner in which her paired associate learning improved with instructions to use imagery suggests strongly that she did not have a fundamental memory deficit; rather, an inability to execute an optimal encoding strategy. This finding is again mirrored in the frontal lobe literature. Parkin et al. (1988) described similar findings in their AcoA patient JB (see also Parkin et al., 1993). Also, the fact that Elizabeth could not improve her temporal discrimination when an obvious cue distinguished the two lists also points towards an executive deficit.

Elizabeth's memory impairments thus resemble those encountered in frontal patients; but how strongly can we argue that she has disrupted frontal lobe function? First, we can point to her bad performance on figure copying, fluency and Wisconsin Card Sorting measures as evidence for frontal disturbance. However, doubts have often been ex-pressed about the localizing value of these tests (for example, Bigler, 1988) and, on their own, they are perhaps an insufficient basis for asserting the neuroanatomical locus of Elizabeth's problems.

A stronger case for a frontal basis to Elizabeth's memory loss comes from the observation that Elizabeth's psychiatric presentation can be characterized as that of negative smptoms (Andreasen, 1982). These symptoms are generally associated with, but not exclusive to, the aftermath of schizophrenia. They include blunting of affect, apathy, disorganization, impaired memory and bradykinesia. These symptoms were all shown in Elizabeth, and thus any information available concerning the neural basis of negative symptoms might be germane to a greater understanding of Elizabeth's difficulties.

A pattern of hypofrontality is now well established in schizophrenia (see for example, Buchsbaum et al., 1992) and there are a number of indications that a direct consequence of this may be the appearance of negative symptoms. Neurophysiological studies (Besson et al., 1987; Volkow et al., 1987) have indicated that hypofrontality and negative symptoms are correlated, and Merriam et al. (1990) reported a significant relationship between neurological 'soft' signs of frontal dysfunction and negative symptoms. Liddle (1987; Liddle and Barnes, 1990) has also linked negative symptoms and frontal dysfunction but argues for two subsets of negative symptoms each relating to a different aspect of frontal function. More recently, Wolkin et al. (1992) have confirmed that the pattern of hypofrontality associated with negative symptoms is not an artefact of medication, and Andreasen et al. (1992) have shown that hypofrontality in schizophrenics correlates with both the extent of negative symptoms and the Tower of London Test - a newly developed test of frontal dysfunction.

What initiated memory loss?

We have suggested that the memory loss shown by Elizabeth bears considerable resemblance to that encountered in frontal lobe dysfunction. The next and perhaps more difficult question is to explain why this loss of memory occurred.

The first possibility is that Elizabeth's memory loss is due to the depression that accompanied her negative symptoms. This would seem very unlikely. It is well recognized that the memory impairment in severe depression resembles that of dementia, and the impairment is anything but protective. In particular, the person reveals a memory impairment that makes day-to-day lift more difficult while long-term memory is left relatively intact. As a result the depressed person characteristically focuses on past events that confirm their negative perceptions of themselves and the world (Eyesnck and Mogg, 1991). In contrast, Elizabeth's deficit appeared to achieve the opposite in that it served to insulate her from her disturbing past.

It is reasonable to suppose that our intellectual functions may not only help us to understand the 'world out there'; they may also protect our 'world within' when the world out there proves too much for us to bear. Everyone knows we 'rationalize' at times to give ourselves an excuse and we often advise those in great distress after some emotionally traumatic event to 'forget all about it', 'put it out of your mind'. It is the possible purpose underlying functional memory impairment that makes all comparisons with machine impairment or breakdown inappropriate. A car does not develop a functional timing problem to protect the engine from overheating, but a person evidently can develop amnesia for some emotionally overbearing event. Normal memory is a 'functional' brain attribute, and there is a lot of clinical observation to suggest that protective forgetting is part of normal memory functioning (Parkin, 1987).

Stampfer (1990) has specifically considered the protective role that might be played by the memory loss associated with negative symptoms. He notes the marked similarity between the nature of negative symptoms and the pattern of impairment seen in patients with post-traumatic stress disorder (PTSD). Both disorders have memory loss as a prominent symptom, and Stampfer argues that for the PTSD patient the memory loss serves to keep the distressing predisposing event out of mind. Memory loss arising within the context of negative symptoms is addressed from the standpoint that negative symptoms allow an acute psychotic episode. Here Stampfer argues that the protective mechanism prevents the patient from recalling the disturbing hallucinations and thought disturbance of the acute phase.

Elizabeth's personal history created an ideal background within which protective forgetting could arise. Her life featured a number of extremely negative life events culminating in a psychotic episode characterized by disturbing hallucinations. It is not difficult to envisage how her subsequent negative symptoms, especially loss of memory, could have imparted a degree of functional value to her.

Elizabeth's experience is not unusual in that there are many instances of patients showing dissociative memory disturbances following disturbing life events. Most notable are the circumscribed amnesias surrounding combat and crime - many of which are now likely to come into the category of PTSD. In addition, one can include the more pervasive disruptions of memory underlying fugue, multiple personality and the kind of mechanisms responsible for disruptive memories of incest or similar traumatic experiences which return many years after the event.

The mechanism observed here can be given a number of names with 'motivated forgetting' or 'repression' the most obvious. By combining observations of these phenomena with neuropsychologicai analysis we have shown how an explication of these somewhat vague psychodynamic terms might bc incorporated into the framework provided by the newly emerging cognitive neuropsychology.

[excerpt from Clinical Disorders of Memory]

D. MEMORY IMPAIRMENT IN DISSOCIATIVE DISORDERS

According to the diagnostic criteria of DSM-III, 'dissociative disorders' are characterized by a sudden, temporary alteration in the normally integrated functions of consciousness, identity, or motor behavior so that some part of one or more of these functions is lost. The most important characteristic of dissociative disorders is a partial or complete loss of memory for a specified period of time. The memory, however, is restored with the resolution of the anxiety and conflict. This distinguishes these disorders from organic causes such as postconcussion amnesia, epilepsy, and intoxication with permanent loss of memory.

DSM-III recognizes four primary and several less common types of dissociative disorders: (1)psychogenic amnesia; (2)psychogenic fugue; (3) depersonalization; and (4) multiple personality.

Psychogenic Amnesia

This disorder is characterized by sudden inability to recall important personal information. This inability is too extensive to be explained by ordinary forgetfulness. The amnesia may take several forms such as localized, generalized, systematized, or continuous. Localized amnesia is referred to as a loss of memory for a short period of time (a few hours or less), whereas generalized amnesia covers a much larger segment of time or the whole of the past life.

In systematized amnesia, the patient loses memory for a specific event such as the birth of a baby or the death of a friend. Memories for other events occurring during the same period of time are retained. In continuous amnesia (anterograde), memory is lost for all events as soon as they occur for a specific period of time. The patient at all times is aware of his inability to recall the amnesic event or the period.

Psychogenic Fugue

This disorder is characterized by a sudden, unexpected trip away from home or one's customary place of work with an inability to recall one's past and an assumption of a new identity (partial or complete). During the state of fugue, the person is amnesic about his past. After assuming a new identity, the individual may work on a job and carry on his daily life quietly without experiencing the conflicts that precipitated the fugue. The person does not seem to be experiencing any psychiatric problem. At the termination of the fugue, the memory for the past life before the onset of fugue is restored, but the memory for the events occurring during the state of the fugue is temporarily lost.

Depersonalization Disorder

Depersonalization refers to the feeling that one's own self is unreal or temporarily lost. (This is frequently distinguished from "derealization' in which only the environment appears to be unreal.) There is no loss of memory but a dissociation of feelings about one's self or the environment. The person is fully aware of himself and his environment. However, there is a dreamlike sense of one's self and a feeling of unreality. The person may feel detached from his own body and other mental operations. This disorder occurs in different degrees of intensity and for variable durations. The frequency of recurrence may be quite frequent or only occasional. The periods of depersonalization may last from a few minutes to hours, occurring several times a day. Occasionally, the disorder may continue uninterruptedly for days. The feelings of depersonalization and derealization are common in young adults under conditions of stress, anxiety, and bereavement. However, the disorder is more commonly associated with severe psychiatric conditions such as depression and schizophrenia.

Multiple Personality

This disorder is characterized by the existence of two or more personalities within an individual. At any given time, the person is dominated by one of the personalities that determines his behavior. Each individual personality is complex and integrated with its own unique behavior patterns and social relationships. Transition from one personality to another is sudden.

There is generally an amnesic barrier between the personalities The initial psychiatric examination rarely reveals any abnormal findings with the exception of possible amnesia for varying periods of time. Without the history and the collateral information, it may be very difficult to assess if the patient leads other lives at other times.