Psychological Medicinehttp://journals.cambridge.org/PSM

Additional services for Psychological Medicine:

Email alerts: Click hereSubscriptions: Click hereCommercial reprints: Click hereTerms of use : Click here

Remote memory in a patient with amnesia due to hypoxia

William W. Beatty, David P. Salmon, Nancy Bernstein and Nelson Butters

Psychological Medicine / Volume 17 / Issue 03 / August 1987, pp 657 ­ 665DOI: 10.1017/S0033291700025897, Published online: 09 July 2009

Link to this article: http://journals.cambridge.org/abstract_S0033291700025897

How to cite this article:William W. Beatty, David P. Salmon, Nancy Bernstein and Nelson Butters (1987). Remote memory in a patient with amnesia due to hypoxia. Psychological Medicine, 17, pp 657­665 doi:10.1017/S0033291700025897

Request Permissions : Click here

Downloaded from http://journals.cambridge.org/PSM, IP address: 132.203.235.189 on 04 Mar 2013

Psychological Medicine, 1987, 17, 657-665

Printed in Great Britain

Remote memory in a patient with amnesia dueto hypoxia

WILLIAM W. BEATTY1 DAVID P. SALMON, NANCY BERNSTEINAND NELSON BUTTERS

From the Department of Psychology, North Dakota State University, Psychology Service, San Diego andDepartment of Psychiatry, University of California, San Diego, USA

SYNOPSIS It has been suggested that amnesic patients suffer a selective loss of episodic memorywhile semantic memory remains well preserved. To assess the validity of this idea we studied remotememory in an amnesic patient, (M.R.L.), using several different measures that differ in the extentto which they engage episodic or semantic memory. On two different versions of the Albert et al.(1979) remote memory battery M.R.L. displayed severe retrograde amnesia (RA) extendingbackwards in time for about 15 years with excellent preservation of older memories. With standardrecall instructions his overall performance on the Crovitz test of autobiographical memory wasimpaired and all of M.R.L.'s specific, temporally dated memories were given from the first half ofhis life. When asked to reconstruct his past residential history in detail, M.R.L. provided specificand generally accurate information for residences occupied from his boyhood until 1970, butthereafter his memory became quite unreliable. On a test of knowledge of terms commonly employedin the surveying profession, in which he worked for the past 20 years, M.R.L.'s performance wasalso impaired. The consistent pattern of RA displayed by this patient on all of the tests of remotememory indicates that both episodic and semantic memory are impaired in amnesia.

INTRODUCTION preserved learning without awareness has beenoffered by Kinsbourne and colleagues (Kins-

It is now well established that patients with bourne & Wood, 1975; Wood et al. 1982). Theirglobal amnesia can display nearly normal model suggests that in amnesia episodic memoryacquisition and retention on a variety of tasks, is impaired but semantic memory remains intact.These include motor skills such as pursuit rotor Following Tulving's (1972) distinction, episodicand mirror-star tracing (Cermak et al. 1973; memory is conceptualized as consisting ofCermak & O'Connor, 1983; Corkin, 1968; autobiographical information for events thatMilner et al. 1968), perceptual skills such as remain tightly linked to the spatial and temporalidentifying Gollin figures (Milner et al. 1968; context in which they were originally acquired,Warrington & Weiskrantz, 1968) or learning to whereas semantic memory is concerned withread transformed or mirror-reversed text (Cohen general knowledge, rules and procedures which& Squire, 1980; Martone et al. 1984; Mosco- are highly overlearned and essentially context-vitch, 1984), classical eyelid conditioning (Weis- free. The normal performance of most amnesicskrantz & Warrington, 1979) and memory for on intelligence tests as well as their ability to usesimple paired associates (Winocur & Weiskran- language and adhere to social conventions aretz, 1976). Despite clear evidence of progressive taken as examples of the preservation ofimprovement in performance from one session semantic memory. In support of the idea thatto another, amnesics typically seem unable to episodic memory is selectively impaired inremember the occurrence of the earlier testing amnesia, Kinsbourne & Wood (1975) reportedsessions on the task. that alcoholic Korsakoff patients had great

One explanation of this phenomenon of difficulty recalling specific personal memories• Address for correspondence: Dr W. W. BeaUy, Department of ('•«• episodes) in response to the Cue words (e.g.

Psychology, North Dakota state University, Fargo, ND 58105, USA. flag, bird) employed in the technique devised by

657

658 W. W. Beatty and others

Crovitz & Schiffman (1974) for studying auto-biographical memory. Many of their patientscould generate only vague responses which weredevoid of either spatial or temporal context (e.g.'flags are for waving in parades', p. 278). Fromtheir perspective such responses indicated thatepisodic memory were compromised but sem-antic memory was preserved. A similar dissocia-tion was reported by Weingartner et at. (1983).

By contrast, Zola-Morgan et al. (1983), whoalso studied remote memory in alcoholicKorsakoff patients using the Crovitz technique,found only modest deficits in their patient'sability to produce episodic memories. Since thesame patients had demonstrated retrogradeamnesia (RA) on other tests of remote memory(e.g. identifying famous persons, recalling pastpublic events) which are thought to tap semanticmemory, these workers questioned the view thatepisodic memory is selectively impaired inamnesia.

Tests of general information which requireidentification of famous people, public events, ortelevision programmes from the past provide apotentially more rigorous test of the integrity ofsemantic memory in amnesia. However, studiesof remote memory in amnesic patients with suchtests have yielded widely varying estimates of theextent of their RA (Markowitsch & Pritzel,1985; Parkin, 1984 for reviews).

Further, Cermak (1984) has argued that newlyacquired knowledge of the sort measured by testsof general information is principally episodic innature, and only with time and continuedrehearsal do memories become independent ofspecific spatial and temporal contexts. From thisviewpoint both the temporally limited RAexhibited by patients H.M. and N.A. (Cohen &Squire, 1981; Marslen-Wilson & Teuber, 1975)as well as the more extensive, but temporallygraded RA displayed by alcoholic KorsakofFspatients (e.g. Albert et al. 1979) can be attributedto the failure of episodic memory.

It seems unlikely that a selective loss ofepisodic memory can explain the pattern of RAobserved in the postencephalitic patients, S.S.(Cermak & O'Connor, 1983) and D.R.B.(Damasio et al. 1985). These patients showmarked loss of remote memory which isapproximately equivalent for all past decadesregardless of whether remote memory is studiedby techniques intended to probe episodic orsemantic memory.

Nevertheless, if Cermak's argument thatsuccessful performance on remote memory testsof general information may reflect retrieval fromeither episodic or semantic memory is valid, thenone must attempt to devise a test of remotesemantic memory which taps facts which are sohighly overlearned that they are likely to becontext-free. A test of factual information abouta person's profession might satisfy this require-ment, but to our knowledge only the alcoholicKorsakoff patient, P.Z., has received such a test.Butters & Cermak (1986) required P.Z. toidentify famous scientists and important theor-etical and empirical results from his own field.They reported that P.Z. exhibited temporallygraded RA on tests of professional knowledgewhich closely paralleled his performance on testsof his autobiographical memory and ability toidentify famous people.

These observations, based on a single case,argue against the view that amnesics suffer aselective loss of episodic memory, but unfortu-nately data from another well studied amnesicpatient, S.S., suggest another conclusion. PatientS.S. exhibited a marked loss of remote memoryon tests which tap general knowledge of pastevents and famous people and was unable togenerate specific episodes from his past. Appar-ently he retained an excellent command ofknowledge of his professional speciality (opticalphysics) which he acquired prior to his enceph-alitic episode (Cermak & O'Connor, 1983), butthe results of his performance on a formal testof professional knowledge have not beenpublished.

In an effort to clarify the question of whetheror not episodic and semantic memory aresusceptible to RA we studied remote memory inpatient M.R.L., who developed an enduringanterograde amnesia for both verbal andnonverbal information after a hypoxic episode.To probe his episodic memory we utilized theCrovitz technique and a structured interview inwhich he was asked to provide a detailedreconstruction of all of the residences he hadoccupied throughout his life. To assess hissemantic memory we utilized standard tests ofhis ability to identify famous persons and publicevents from the past as well as a test of hisknowledge of his profession (surveyor's assist-ant) which we devised.

Retrograde amnesia 659

CASE HISTORYA detailed description of the case history and resultsof the general neuropsychological testing for patientM.R.L. has been published elsewhere (Beatty et al.1987), so only a brief summary will be provided here.M.R.L. is a 54 year old Caucasian male who was bornand raised in Lynn, MA. After graduation from highschool he served 20 years in the Marine Corps. Fromage 20 to 54 he resided in Southern California exceptfor 3 year-long tours of duty overseas and a 5month-long residence in Alaska in 1975-76. M.R.L.learned surveying in the Marine Corps and worked asa surveyor's assistant for 17 years after leaving theservice. During this time he completed collegecoursework equivalent to about 3 years.

M.R.L. has a long history of alcohol abuseincluding alcohol-related blackouts, convulsions,hallucinations and convictions for driving whileintoxicated and public drunkenness, but he wastreated for alcoholism only once (in 1980).

Since 1964, M.R.L. has suffered periodically fromgrand mal seizures (about 2-3/yr). In September 1984he suffered a series of generalized seizures as apassenger in a car (no accident or head injury wasinvolved). Postictally he was comatose for about 2 h.At this time an EEG revealed severe sharp spikes anddelta waves; one week later an EEG showed mild tomoderate generalized nonspecific abnormalities, butno focal abnormalities or epileptiform discharges wereapparent. A CT scan revealed no evidence of sulcalwidening or ventricular enlargement and a possiblelucency in the left temporal lobe.

During this 5 week-long hospitalization M.R.L.received extensive neuropsychological testing whichsuggested a global amnesic syndrome affecting bothanterograde and remote memory. His IQ was aboveaverage (VIQ = 112, PIQ = 111 on the WAIS-R) buthis performance on the Wechsler Memory Scale wasimpaired (MQ = 92) especially on the LogicalMemory and Associate Learning scales (scaledscores = 6 on both measures). He scored 135 out of144 on the Dementia Rating Scale (Coblentz et al.1973) losing all 9 points on the memory items. Verballearning deficits were also evident on the ReyAuditory Verbal Learning Task (Lezak, 1983). Hiscopy of the Rey-Osterrieth complex figure (Lezak,1983) was normal (32/36) but when asked toreproduce the drawing 3 min later, his performancewas badly defective (4/36). Parietal lobe drawings tocommand (Goodglass & Kaplan, 1983) were normal,but his score on the Boston Naming Test (49/60without cues) suggested mild dysnomia. His perform-ance on the Famous Faces from the Albert et al.(1979) Remote Memory Battery is described below.

Because of his long history of alcohol abuseM.R.L.'s performances on tests which are known tobe sensitive to frontal lobe dysfunction in general and

to alcoholic Korsakoff s syndrome (AK) in particularwas of considerable interest. M.R.L. performednormally on the FAS test (a measure of verbal fluency)administered as described in Beatty et al. (1987), theWisconsin Card Sorting Task (6 categories, 22 errors,12 perseverative errors), and on the release fromproactive interference (PI) test administered asdescribed by Beatty et al. (1987). On all of these taskspatients with AK display consistent deficits (Butters,1985). On the release from PI test short-term memoryfor lists of 3 semantically related words is studied overa series of trials. Normal subjects or alcoholic controlswithout KorsakofTs syndrome perform well on thefirst trial but remember fewer words on succeedingtrials as PI builds up. If the semantic category of thestimulus words is changed without warning, perform-ance by controls recovers demonstrating release fromPI. Although patients with frontal lobe injury or AKexhibit normal buildup of PI when given a series oftrials with lists of words from the same category, theydo not exhibit release from PI after an unannouncedcategorical shift (Cermak et al. 1974; Moscovitch,1982; Squire, 1982a). M.R.L. performed normally onall of these measures, therefore his amnesic syndromeis not likely to arise from AK. The absence of signsof Wernicke's encephalopathy, nutritional deficien-cies, confabulation as well as his clear insight into andconcern about his amnesic syndrome support thisconclusion.

In the eight months after his discharge from thehospital to the care of a friend M.R.L. was retestedperiodically as described below and elsewhere (Beattyet al. 1987). His IQ and MQ (111 and 90 respectively)were essentially unchanged on retest 3 months laterand his performance on verbal learning tasksremained impaired. His mild dysnomia improved asindicated by his score (53 without cues) on readmin-istration of the Boston Naming Test 3 months afterdischarge from hospital. His seizure disorder was wellcontrolled with medication during this period, andexcept for amnesia and anosmia he exhibited no otherneurological signs.

MATERIALS AND PROCEDURES

Revised remote memory battery (RMB)The remote memory test was a revised andupdated version of the battery developed byAlbert et al. (1979). It consisted of twocomponents: Famous Faces and Public EventsRecall Questionnaire. The Famous Faces por-tion consisted of photographs of persons whowere best known during the 1940s, 50s, 60s, 70sand 80s. There were 15 items from each decade.The Public Events Recall Questionnaire con-sisted of questions about events that occurred in

660 W. W. Beany and others

each of the above mentioned decades. Therewere 15 questions from each decade. All itemsfrom 1940s, 50s and 60s and about one-half ofthe items from the 1970s were taken from theoriginal Albert et al. (1979) battery. Theremaining items from the 1970s and all of theitems from the 1980s were created for this study.

The subjects were first shown the 75 photo-graphs from the Famous Faces test, one at atime, and asked to name the person pictured. Ifthe subject answered incorrectly or could notname the person, a standard set of semantic cues(e.g. actress, World War II pinup girl, starred in'The Outlaw' for Jane Russell) was read and thesubject was allowed to respond again. The75 questions from the Public Events RecallQuestionnaire were presented in a similarmanner. On both tests items were presented in anordered manner (i.e. a question from the 40s wasfollowed by one from the 50s, 60s, 70s, 80s, etc.).Control subjects for this test were 14 neuro-logically normal individuals aged 52-75 years(Mean = 62-5 years) whose average educationwas 13-5 years.

Surveyor's testIn order to assess M.R.L.'s memory for hisoccupation, his ability to define surveying termswas tested. Twenty terms commonly employedin the surveying profession (e.g. bench mark,transit) were selected from an elementarysurveying text (Breed & Hosmer, 1977). Theselection criteria were that each term wascommon enough to be repeatedly encounteredby someone at M.R.L.'s level in the profession(surveyor's assistant), but not so common thatthey could be easily defined by a non-surveyingperson. A short, one or two sentence definitionof each term was elicited, and scored 0, 1, or 2points depending upon its quality. Two controlgroups were employed. One group (N = 4),consisting of surveyor's assistants, averaged 30-3years of age (range, 27-36), 130 years ofeducation (range, 12-14) and 100 years ofprofessional experience (range, 7-12, comparedto 20 years for M.R.L.). Thus these subjects weresignificantly younger, less well educated and lessexperienced professionally than M.R.L. Thesecond control group (N = 6), consisting of menwith no surveying experience, was not signifi-cantly different from M.R.L. in terms of age(mean = 45-5 years, range, 24-75) and education(mean = 160 years, range, 12-20).

Episodic memoryThis test of remote memory was based on a testof autobiographical memory described byCrovitz & Schiffman (1974). Subjects were reada list of 10 words (e.g. tree, baby, book) one ata time and instructed to recall a specific event,from any time in the past, which involved thestimulus word. Subjects were asked to describethe experience as completely and specifically aspossible and then to estimate, as precisely aspossible, when the event occurred. If the subjectsfailed to describe a memory that was clearlyspecific with respect to time and place, theexperimenter probed in an attempt to elicit themost detailed recollection possible. Sessionswere tape recorded and later transcribed forscoring.

Responses to each of the 10 words were scoredon a 0-3 point scale based on the first memoryoffered by the subject for each word: 3 = anepisodic memory (i.e. a description of a personalexperience that occurred at a particular time andplace), 2 = a personal memory that was notspecific to both time and place, 1 = a vaguememory that did not involve instantiation of thestimulus word (e.g. 'Trees are nice') and 0 = noresponse.

On the first administration, all of the specificmemories M.R.L. produced that could beaccurately dated were from the first half of hislife. Although he was instructed to recallexperiences from any time in his life, he seemedto be trying to recall old memories. Accordingly,he was given the episodic memory test a secondtime (about one week after the first administra-tion), with the explicit instructions to 'try torecall the most recent specific experience thatincorporates the stimulus word'.

Sixteen neurologically normal controls, aged40-76 years were tested once with the standardinstructions which did not specify the age of thememories to be produced.

Former residences

M.R.L. was asked to draw floor plans andprovide other details (e.g. addresses) of homesthat he had lived in throughout his life. M.R.L.'sbrother, a 66 year old man with 12 years ofeducation, verified the accuracy of M.R.L.'sdrawings of homes from his boyhood. M.R.L.'sformer wife and son verified the floor plans ofother homes in which he had lived later in his life.

Retrograde amnesia 661

RESULTS

Semantic memory (RMB)

Fig. 1 summarizes M.R.L.'s performance on theFamous Faces test from the Remote MemoryBattery developed by Albert et al. (1979) incomparison to that of the standardizationsample of normal men at least 50 years of age.M.R.L.'s performance clearly exceeded that ofcontrols for items from the 30s and 40s, wasabout average for items from the 50s and 60s,but fell below normal for items from the 70s.Since this test was administered while M.R.L.was still dysnomic, one cannot be certain that histemporally graded performance reflects onlyamnesia for events from the recent past.

The results of tests with the revised andupdated form of the RMB (Fig. 2) clarify thisissue, since this test was administered afterM.R.L.'s dysnomia had cleared. Again M.R.L.displayed better than average performance onitems from the 40s approximately averageperformance for items from the 50s and 60s, butmarkedly deficient memory for items from the70s and 80s. For the latter two decades M.R.L.'sperformance fell below the 95% confidencelimits for controls (both zs > 200). Providingsemantic cues improved performance by controlsand patient M.R.L., but M.R.L.'s performance

Controls M.R.L.

o Controls M.R.L.

90

70

t3E 50

30

10

O—

30s 40s 50s 60s 70s

Decades

FIG. 1. Performance of M.R.L. and controls on the Famous Facestest from the Albert el al. (1979) remote memory battery. Solid linesindicate performance without cues; broken lines depict performancewith semantic cues.

90

70

50

30

10

o ^ \

-

-

-

-

• i

V\

i i i

40s 50s 60s

Decades

70s 80s

FIG. 2. Performance of M.R.L. and controls on the revised remotememory battery. The data shown are the overall performance forFamous Faces and Public Events questions Solid lines indicateperformance without cues; broken lines depict performance withsemantic cues.

on items from the 70s and 80s remained impaired(both zs > 2-30).

Surveyor's testOn the surveyor's test M.R.L.'s performance(65% correct) was significantly better than thatof the nonsurveyor controls (mean = 15-4%correct, s.D. = 17-3), but significantly belowthat of the surveyor's assistant controls(mean = 82-5% correct,s.D. = 61). Within bothcontrol groups there was no correlation betweenperformance on the surveyor's test and eitherage or education.

Episodic memoryFig. 3 summarizes M.R.L.'s performance on theCrovitz test in comparison to that of controls.Because the subjects' ages varied over a range ofalmost 40 years, the relative age of specificmemories was quantified by determining theproportion of specific memories reported whichoccurred in each quartile of each subject's life.Under the standard conditions of administra-tion, in which subjects were asked simply toproduce specific memories from any time periodin their lives to the target probe words, controlsproduced a somewhat greater number of specificmemories than M.R.L. (mean score = 25-8 v19), but this effect fell just short of statisticalsignificance (z=l-8). The specific memories

662 W. W. Beany and others

Controls

M.R.L.

M.R.L.

\ Standard instructions'Most recent memory'

|

80

60

40

20

0

\\

\

\

- ^ x ^ \ /'v \ /

\ ' '

X -X/_i i T

Ai

ii

ti

ii

ii

ii

i*~~—_„

o

1st 2nd 3rd

Quartile of life

4th

FIG. 3. Temporal distribution of specific personal memories recalledby M.R.L. and controls under standard instructions (no instructionsconcerning the period of time from which memories were to berecalled), and for M.R.L. when he was asked to recall his most recentmemory.

recalled by controls were about equally likely toarise from each of the 4 quartiles of their lives(F(3,45) for Quartiles < 1). Relative to controls,M.R.L. recalled a greater percentage of specificmemories from the first quartile of his life(z = 2-93).

When explicitly asked to recall his most recentmemories, M.R.L.'s overall memory score (18)was significantly lower than that of controls(z = 200), but three of the four temporallyspecified memories he recalled involved experi-ences that occurred after his hypoxic episode.Since his family members were able to verify thetime and occurrence of these events, M.R.L.'sresponses appear to be genuine episodic mem-ories rather than confabulations.

One of these recent memories (elicited by theword 'street') is particularly notable. To thisprobe M.R.L. described an occasion shortlyafter his release from the V.A. hospital when hebecame lost while attempting to return home atnight after a social event. He was unable to findhis way home and had to be rescued by thepolice. M.R.L. described his fear so vividly thathe almost seemed to be re-experiencing the event.He was also able to date his experience within 1month of the actual occurrence even though theepisode occurred 4 months prior to testing.

Memory of past residencesM.R.L. lived in 5 different homes in Lynn, MA.He occupied three of these residences for lessthan 2 years each. He lived in the remaining twohomes for about 8 years (age 3-11) and 5 years(age 12-17). From 1947-1968 while M.R.L. wasin the Marine Corps he lived mainly in basehousing for short periods in each residence. Thenotable exception to this was a period of about5 years (1958-63, age 28-33) when M.R.L. andhis first wife lived in a single family house inTwenty-nine Palms, CA. From 1968—84 (age38-54) M.R.L. lived in 13 different homes (11 inSan Diego, CA, 2 in Alaska).

M.R.L.'s memory for his boyhood homes inMA appeared to be quite intact. He was able toprovide specific addresses for all of the 5 placeswhere he had lived. His drawing of the home inwhich he lived from age 3-11 was identical to thedrawing provided by his brother in terms of thenumber and kinds of rooms in the house andtheir spatial arrangement. M.R.L.'s drawing ofthe apartment he lived in from age 12-17 differedfrom the drawing provided by his brother onlyin the labelling of one room. However, it shouldbe noted that his brother was not confident ofthe accuracy of his own drawing of thisparticular home.

No attempt was made to have M.R.L. drawdiagrams of any of the barracks or other militaryhousing in which he had resided. M.R.L.provided an accurate address for the house inTwenty-nine Palms, CA in which he and his firstwife had lived in from 1958-63. His drawing ofthe floor plan of that house was identical to theone produced by his son, except that M.R.L.remembered the house as having 3 bedroomswhile his son remembered it as having only 2bedrooms.

From 1968-70 M.R.L. resided at his brother'shome in San Diego. His drawing of the floor planof this 3-bedroom home was identical to thatprovided by his brother with respect to number,kind and spatial location of all rooms. He alsoreported the address accurately.

By contrast, M.R.L.'s ability to recall theremaining 12 homes in which he lived between1968-84 was severely impaired. He failed toreport having lived in 6 of these homes at all. Forthe remaining 6 homes he was able to provideonly vague and sometimes inaccurate informa-

Retrograde amnesia 663

tion about the general area of the city in whichthese homes were located. He confused onehome which he had lived in during the early1970s with another that he had lived in from1978-82. He could not immediately recall thename of his third wife with whom he lived in1980-1, although he eventually produced hername with cueing. He did not remember theregion of San Diego in which they lived.

M.R.L. did remember the general location ofthe apartment that he had lived in for the twoyears preceding his hypoxic episode, but at firsthe incorrectly recalled the street on which it waslocated. He also stated (incorrectly) that theapartment building had two storeys. Eventuallyhe corrected these mistakes and his floor plan ofthis one-bedroom apartment was essentiallyidentical to one drawn by his former wife.

DISCUSSION

Our studies of remote memory in the amnesicpatient M.R.L. suggest a consistent pattern ofRA which involves both episodic and semanticmemory. On those tests on which the age ofmemories could be estimated, the data suggest atemporally graded RA extending backwards intime for at least 15 years with excellentpreservation of older memories. His perform-ance on both versions of the Albert et al. (1979)battery yielded essentially identical results. Theaccuracy of his recall of places that he lived inthroughout his life also seemed well preserveduntil about 15 years before his hypoxic episodewhen his memory for this aspect of his pastbegan to fail. Finally, when M.R.L. was giventhe Crovitz test of autobiographical memorywithout instructions concerning the age of thememories that he was to recall, he recalled fewerspecific memories than controls and thoseautobiographical memories that he did recallwere from the distant past.

The highly consistent estimates of the extentof M.R.L.'s remote memory loss we obtainedusing tests of personal, context-dependent infor-mation as well as tests of more general andpublic knowledge, and are consistent with resultsobtained in studies of several other amnesicpatients. For example, alcoholic Korsakoffpatients suffer temporally graded RA regardlessof whether remote memory is studied using testsof general information (e.g. Albert et al. 1979) or

the Crovitz & Schiffman (1974) technique forstudying autobiographical memory. Likewise,the postencephalitic patients, S.S. (Cermak &O'Connor, 1983) and D.R.B. (Damasio et al.1985), exhibit severe losses of both personalmemories and knowledge about world eventswith no indication that older memories are betterpreserved. One interpretation of this highlysimilar pattern of loss across different measuresof remote memory is that both episodic andsemantic memory are affected in amnesia.However, if tasks such as the Albert et al. (1979)battery actually measure episodic as well assemantic memory, as Cermak (1984) suggests,then M.R.L.'s RA on both versions of theremote memory battery could be attributed todefects in episodic memory alone.

M.R.L.'s impaired performance on our test ofhis knowledge of surveying is in good agreementwith the deficits in professional knowledgedisplayed by patient P.Z. (Butters & Cermak,1986). If memory for definitions is an appro-priate measure of semantic memory, as Tulving(1983) suggested, then the data from bothpatients question the view (Kinsbourne & Wood,1975; Weingartner et al. 1983) that semanticmemory is selectively spared in amnesia. Sinceboth M.R.L. and P.Z. exhibit defects in theirknowledge of their professions, it is surprisingthat patient S.S., whose RA on the Albertet al. (1979) Famous Faces test is considerablymore severe than that of M.R.L., apparentlyretains knowledge of his profession (Cermak &O'Connor, 1983). Perhaps a more rigorous andsystematic test of S.S.'s knowledge of theoriesand empirical findings in physics would revealdeficits in his professional knowledge similar tothose observed for M.R.L. and P.Z.

Recently McKoon et al. (1986) argued thatthe episodic-semantic distinction is too poorlydefined to be empirically testable and that muchof the empirical evidence cited by Tulving (1983)to support the episodic-semantic dichotomy is,in fact, inconclusive. Since the supposedlynormal performance of amnesics on semanticmemory tests coupled with their obvious defectsin episodic memory has been advanced tosupport the episodic-semantic distinction, thepresent findings with patient M.R.L. (and P.Z.as well) may raise additional doubts about theutility and validity of the episodic-semanticdichotomy.

664 W. W. Beatty and others

An alternative model to account for thepattern of memory preservation and loss inamnesia has attracted much contemporaryinterest (Baddeley, 1984; Squire, 19826). Itsuggests that amnesics can learn and remembernormally if the information is skill-based (i.e.procedural knowledge), whereas they are im-paired when required to learn and rememberspecific data-based information (i.e. declarativeknowledge). Tulving himself (1984) has con-ceded the distinction between episodic andprocedural memory.

According to the procedural-declarative dis-tinction, impairments in remote memory includ-ing the deficits in professional knowledgeexhibited by M.R.L. and P.Z. are to be expectedsince the information that must be rememberedis factual or data-based. Thus, the presentfindings are not inconsistent with the pro-cedural-declarative dichotomy. However, inother studies with M.R.L. (Beatty et al. 1987) wehave shown that he does not perform normallyon all tasks which are supposed to be measuresof procedural learning. Although M.R.L.showed normal improvement in learning to readtransformed script and normal within-sessionimprovement on a mirror-reading task, hisretention of the mirror-reading skill from oneday to the next was impaired, and he learned theTower of Hanoi puzzle more slowly than normal.These latter observations suggest that thedistinction between declarative and procedurallearning is presently no more operational thanthe difference between episodic and semanticmemory.

This study was supported in part by funds from theMedical Research Service of the Veterans Adminis-tration and by NIAAA grant AA-00187.

REFERENCESAlbert, M. A., Butters, N., & Levin, J. A. (1979). Temporal gradients

in the retrograde amnesia of patients with alcoholic KorsakofTsdisease. Archives of Neurology 36, 211-216.

Baddeley, A (1984). The fractionation of human memory.Psychological Medicine 14, 259-264.

Beatty, W. W., Solmon, D. P., Bernstein, N., Martone, M., Lyon, L.,& Butters, N. (1987). Procedural learning in a patient with amnesiadue to hypoxia. Brain and Cognition (in the press).

Breed, C. S. & Hosmer, G. L. (1977) Elementary Surveying, 1 lth edn(revised by W. Faig and B. A. Barry). Wiley: New York.

Butters, N. (1985). Alcoholic KorsakofTs syndrome: some unresolvedissues concerning etiology, neuropathology, and cognitive deficits.Journal of Clinical and Experimental Neuropsychology 7, 181-210.

Butters, N., & Cermak, L. S. (1986). A case study of the forgettingof autobiographical knowledge: implications for the study ofretrograde amnesia. In Autobiographical Memory (ed D. Rubin),pp. 253-272. Cambridge University Press: New York.

Cermak, L. S. (1984). The episodic-semantic distinction in amnesia.In Neuropsychology of Memory (ed. L. R. Squire and N. Butters),pp. 55-62. Guilford Press. New York.

Cermak, L. S., Butters, N., & Moreines, J. (1974). Some analyses ofthe verbal encoding deficit in alcoholic KorsakofTs patients. Brainand Language 1, 141-150.

Cermak, L. S., Lewis, R., Butters, N., & Goodglass, H. (1973). Roleof verbal mediation in performance of motor tasks by Korsakoffpatients. Perceptual and Motor Skills 37, 259-262.

Cermak, L. S., & O'Connor, M. (1983). The anterograde andretrograde retrieval ability of a patient with amnesia due toencephalitis. Neuropsychologia 21, 213-224.

Coblentz, J. M., Mattis, S., Zingesser, L. H., Kassof, S. S.,Wisniewski, H. M. & Katzman, R. (1973). Presenile dementia:clinical aspects and evaluation of cerebrospinal fluid dynamics.Archives of Neurology 29, 299-308.

Cohen, N. J., & Squire, L. R. (1980). Preserved learning and retentionof pattern analyzing skills: dissociations of knowing how andknowing that. Science 210, 207-210.

Cohen, N. J., & Squire, L. R. (1981). Retrograde amnesia and remotememory impairment. Neuropsychologia 19, 337-356.

Corkin, S. (1968). Acquisition of motor skill after bilateral medialtemporal-lobe excision. Neuropsychologia 6, 255-264.

Crovitz, H. F., & Schiffman, H. (1974). Frequency of episodicmemories as a function of their age. Bulletin of the PsychonomicSociety 4, 517-518.

Damasio, A. R., Eslinger, P. J., Damasio, H., Van Hoesen, G. W., &Cornell, S. (1985). Multimodal amnesic syndrome followingbilateral temporal and basal forebrain damage. Archives ofNeurology 42, 252-259.

Goodglass, H., & Kaplan, E. (1983). The Assessment of Aphasia andRelated Disorders Lea & Fibiger: Philadelphia.

Kinsbourne, M., & Wood, F. (1975). Short-term memory processesand the amnesic syndrome. In Short-term Memory (ed. D. Deutschand J. A. Deutsch), pp. 258-291. Academic Press: New York.

Lezak, M. (1983). Neuropsychological Assessment 2nd edn. OxfordUniversity Press: New York.

Markowitsch, H. J. & Pritzel, M. (1985). The neuropathology ofamnesia. Progress in Neurobiology 25, 189-287.

Marslen-Wilson, W. D. & Teuber, H. L. (1975). Memory for remoteevents in anterograde amnesia: recognition of public figures fromnewsphotographs. Neuropsychologia 13, 347-352.

Martone, M., Butters, N., Payne, M., Becker, J. & Sax, D. S. (1984).Dissociation between skill learning and verbal recognition inamnesia and dementia. Archives of Neurology 41, 965-970.

McKoon, G., Ratcliff, R. & Dell, G. S. (1986). A critical evaluationof the semantic-episodic distinction. Journal of ExperimentalPsychology: Learning, Memory and Cognition 12, 295-306.

Milner, B., Corkin, S. & Teuber, H.-L. (1968). Further analysis of thehippocampal amnesic syndrome: 14-year follow-up study of H.M.Neuropsychologia 6, 215-234.

Moscovitch, M. (1982). Multiple dissociations of function in amnesia.In Human Memory and Amnesia (ed. L. S. Cermak), pp. 337-370.Erlbaum: Hillsdale, NJ.

Moscovitch, M. (1984). The sufficient conditions for demonstratingpreserved memory in amnesia: a task analysis. In Neuropsychologyof Memory (ed. L. R. Squire & N. Butters), pp. 104-113. GuilfordPress: New York.

Parkin, A. J. (1984). Amnesic syndrome: a lesion specific disorder.Cortex 20, 479-508.

Squire, L. R. (1982a). Comparisons between forms of amnesia: somedeficits are unique to KorsakofTs syndrome. Journal of Experi-mental Psychology: Learning, Memory and Cognition 8, 560-571.

Squire, L. R. (19826). The neuropsychology of memory. AnnualReview of Neuroscience 5, 247-273.

Tulving, E. (1972). Episodic and semantic memory. In Organizationof Memory (ed. E. Tulving & W. Donaldson), pp. 382-403.Academic Press: New York

Retrograde amnesia 665

Tulving, E. (1983). Elements of Episodic Memory Oxford UniversityPress: New York.

Tulving, E. (1984). Precis of elements of episodic memory. TheBehavioral and Brain Sciences 1, 223-268.

Warrington, E. K. & Weiskrantz, L. (1968). A new method of testinglong-term retention with special reference to amnesic patients.Nature 217, 972-974.

Weingartner, H., Grafman, J., Boutelle, W., Kaye, W. & Martin,P. R. (1983). Forms of memory failure. Science 221, 380-382.

Weiskrantz, L., & Warrington, E. K. (1979). Conditioning in amnesicpatients. Neuropsychologia 17, 187-194.

Winocur, G. & Weiskrantz, L. (1976). An investigation ofpaired-associate learning in amnesic patients. Neuropsychologia 14,97-110.

Wood, F., Ebert, V. & Kinsbourne, M. (1982). The episodic-semanticmemory distinction in memory and amnesia. In Human Memoryand Amnesia (ed. L. Cermak), pp. 167-193. Erlbaum: Hillsdale,NJ.

Zola-Morgan, S., Cohen, N. J. & Squire, L. R. (1983) Recall ofremote episodic memory in amnesia. Neuropsychologia 21,487-500.