This article was downloaded by: [University of Edinburgh]On: 16 September 2013, At: 14:43Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office:Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Cognitive NeuropsychologyPublication details, including instructions for authors and subscriptioninformation:http://www.tandfonline.com/loi/pcgn20

Restoring primacy in amnesic free recall:Evidence for the recency theory of primacyMichaela Dewar a , Gordon D. A. Brown b & Sergio Della Sala aa Human Cognitive Neuroscience and Centre for Cognitive Ageing andCognitive Epidemiology, Psychology, University of Edinburgh, Edinburgh, UKb Department of Psychology, University of Warwick, Coventry, UKPublished online: 23 Feb 2012.

To cite this article: Michaela Dewar , Gordon D. A. Brown & Sergio Della Sala (2011) Restoring primacy inamnesic free recall: Evidence for the recency theory of primacy, Cognitive Neuropsychology, 28:6, 386-396,DOI: 10.1080/02643294.2012.665802

To link to this article: http://dx.doi.org/10.1080/02643294.2012.665802

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)contained in the publications on our platform. However, Taylor & Francis, our agents, and ourlicensors make no representations or warranties whatsoever as to the accuracy, completeness, orsuitability for any purpose of the Content. Any opinions and views expressed in this publicationare the opinions and views of the authors, and are not the views of or endorsed by Taylor &Francis. The accuracy of the Content should not be relied upon and should be independentlyverified with primary sources of information. Taylor and Francis shall not be liable for anylosses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilitieswhatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantialor systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, ordistribution in any form to anyone is expressly forbidden. Terms & Conditions of access and usecan be found at http://www.tandfonline.com/page/terms-and-conditions

Restoring primacy in amnesic free recall: Evidence for therecency theory of primacy

Michaela Dewar1, Gordon D. A. Brown2, and Sergio Della Sala1

1Human Cognitive Neuroscience and Centre for Cognitive Ageing and Cognitive Epidemiology, Psychology, University of

Edinburgh, Edinburgh, UK2Department of Psychology, University of Warwick, Coventry, UK

Primacy and recency effects at immediate recall are thought to reflect the independent functioning of along-term memory store (primacy) and a short-term memory store (recency). Key evidence for thistheory comes from amnesic patients who show severe long-term memory storage deficits, coupledwith profoundly attenuated primacy. Here we challenge this dominant dual-store theory of immediaterecall by demonstrating that attenuated primacy in amnesic patients can reflect abnormal workingmemory rehearsal processes. D.A., a patient with severe amnesia, presented with profoundly attenu-ated primacy when using her preferred atypical noncumulative rehearsal strategy. In contrast, despiteher severe amnesia, she showed normal primacy when her rehearsal was matched with that of controlsvia an externalized cumulative rehearsal schedule. Our data are in keeping with the “recency theory ofprimacy” and suggest that primacy at immediate recall is dependent upon medial temporal lobe invol-vement in cumulative rehearsal rather than long-term memory storage.

Keywords: Amnesia; Primacy effect; Free recall; Short-term memory; Rehearsal.

When asked to recall an immediately precedingword list, neurologically intact people typicallyrecall more words from the beginning of thelist—a primacy effect—and the end of the list—arecency effect—than from the middle of the list,resulting in a U-shaped serial position curve.Patients with medial temporal lobe (MTL)amnesia show recency, but no or grossly attenuated

primacy (Baddeley & Warrington, 1970; Capitani,Della Sala, Logie, & Spinnler, 1992; Carlesimo,Fadda, Sabbadini, & Caltagirone, 1996). Sincesuch patients perform normally on tests of short-term memory (STM) but severely abnormally ontests of long-term memory (LTM), their patternof intact recency and abnormal primacy has pro-vided key evidence for the traditional dual-store

Correspondence should be addressed to Michaela Dewar, Human Cognitive Neuroscience, Psychology, University of

Edinburgh, 7 George Square, Edinburgh, EH8 9JZ, UK. (E-mail: m.dewar@ed.ac.uk).

This research was supported by postdoctoral research fellowships to M. Dewar from Alzheimer’s Research UK and the Royal

Society of Edinburgh/Lloyds TSB Foundation for Scotland, and by an Economic and Social Research Council (UK) grant to

G. Brown (RES-062-23-2462). We thank D.A. and all controls who volunteered to take part in this study and Beata Michalska

who helped with the collection of control data. We are grateful to the editor and three anonymous reviewers for their very

helpful comments on this manuscript. All clinical data listed in Table 1 (apart from the Rivermead Behavioural Memory Test)

were collected by Sharon Abrahams during formal clinical neuropsychology assessment.

386 # 2011 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business

http://www.psypress.com/cogneuropsychology http://dx.doi.org/10.1080/02643294.2012.665802

COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6), 386– 396

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

theory of immediate recall in neurologically intactpeople—namely, that early list words benefit fromsuperior LTM storage, while late list items stillreside within STM at the time of recall(Atkinson & Shiffrin, 1968; Glanzer, 1972).

This traditional dual-store theory of primacyhas recently been called into question by work onneurologically intact people (Tan & Ward,2000). In short, it has been shown that primacyin immediate recall can be accounted for ratherwell by a recency theory of primacy, without assum-ing superior LTM storage by early list items. Thiswork indicates that primacy depends heavily uponthe temporal proximity of online rehearsal of earlylist items to the time of recall. Tan and Ward(2000) used an overt rehearsal paradigm, inwhich participants are instructed to rehearsealoud any list words that come to mind duringunfilled interitem gaps (Rundus, 1971). This pro-cedure revealed that participants used the interi-tem gaps to rehearse multiple list words—that is,they adopted a cumulative rehearsal strategy.Importantly, most early list words, but only fewmiddle list words, were rehearsed until the finalportion of the list. When recall performance wasplotted as a function of position of last rehearsal(i.e., the interitem gap in which a word was lastrehearsed), only a “recency” effect was shown. Inother words, recall performance was highest forrecently rehearsed words, independently of theirserial position in the original list. Crucially, Tanand Ward (2000) showed that the heightenedrecall performance of recently rehearsed itemswas not an irrelevant by-product of the quantityof rehearsals. They controlled the number ofrehearsals and recency of rehearsals via exper-imenter-generated rehearsal schedules, which donot alter the primacy and recency effects observedwith self-generated cumulative rehearsal(Murdock & Metcalfe, 1978; Tan & Ward,2000). Tan and Ward (2000) found that recencyof last rehearsal was more important to primacythan was the mere number of rehearsals. Forexample, performance was higher for a word thatwas only presented twice, but whose last presen-tation occurred close to recall, than for a wordthat was presented 8 times, but whose last

presentation occurred in the early portion of listpresentation. Tan and Ward’s data indicate thatprimacy arises because early list items are carriedforward through the list via rehearsal, thus effec-tively shortening their retention interval atimmediate recall. In essence, the above data ofneurologically intact people suggest that primacyin immediate recall is in fact recency in disguise,rather than the result of superior LTM storageof early list items. This interpretation of primacyis consistent with unitary accounts of memory,such as temporal distinctiveness models (Brown,Chater, & Neath, 2008; Neath & Brown, 2006),which assume that no useful distinction can bemade between STM and LTM at the behaviourallevel and that primacy and recency can both beunderstood in terms of the temporal discriminabil-ity of items and their rehearsals at the time of recall(Brown, Vousden, & McCormack, 2009).

If primacy is dependent upon recency of rehearsalof early list items at the time of recall, then whyshould patients with MTL amnesia present withattenuated primacy given that their recency andworking memory (WM) rehearsal are typicallynormal? Indeed, even the densely amnesic patientH.M. was able to retain the number “584” inmind for 15 minutes via continuous rehearsal ifhis attention was not diverted (Milner, Squire, &Kandel, 1998; Scoville & Milner, 1957).Cumulative rehearsal is more complex, however,since rehearsals of earlier list items have to be inter-spersed with the encoding of new items. Perhapssuch noncontinuous rehearsal is dependent uponthe integrity of the MTL and is therefore impairedin patients with MTL amnesia. While the tra-ditional model of memory has assumed an exclusiverole of the MTL in LTM (Alvarez, Zola-Morgan,& Squire, 1994; Cave & Squire, 1992), more recentneuroimaging and patient work implicates theMTL in somes aspects of WM as well (e.g.,Axmacher et al., 2007; Olson, Moore, Stark, &Chatterjee, 2006; Ranganath & Esposito, 2001).Elucidation of a link between the MTL and cumu-lative rehearsal would further our understanding ofnoncontinuous WM rehearsal processes andprovide a cognitive and neural basis to therecency theory of primacy.

COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6) 387

RESTORING PRIMACY IN AMNESIA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

While some deficits in cumulative rehearsalhave been reported in amnesic Korsakoff patients(Cermak, Naus, & Reale, 1976; Meudell &Mayes, 1980), very little is known about cumulat-ive rehearsal ability in patients with isolated MTLamnesia. We recently reported one such case pre-senting with starkly reduced primacy coupledwith an apparent inability to rehearse cumulatively(Brown, Della Sala, Foster, & Vousden, 2007):Patient S.J. suffered severe amnesia as a result ofisolated bilateral MTL damage followinghypoxia. S.J. adopted a fixed-rehearsal pattern,almost only ever rehearsing the most recently pre-sented item, despite encouragement to engage incumulative rehearsal. This finding resonates withearlier work on neurologically intact people,showing that primacy is substantially attenuatedwhen they are instructed to engage in fixed rehear-sal (Fischler, Rundus, & Atkinson, 1970; Glanzer& Meinzer, 1967).

It is as yet unknown whether the attenuation ofprimacy in amnesia can be explained by defectivecumulative rehearsal alone. This question can beaddressed by comparing primacy performance ina patient with impaired LTM storage withprimacy performance in a group of controls whencumulative rehearsal is matched for the patientand controls via the use of external cumulativerehearsal schedules. Doing so bypasses the needto generate cumulative rehearsal and ensures thatthe patient and controls differ only on one vari-able—LTM storage ability. If the cognitive basisof primacy is largely MTL-dependent cumulativerehearsal of early words rather than superiorLTM storage of early words, then a patient withimpaired LTM storage should show normalprimacy under such externalized cumulativerehearsal condition. If, on the other hand, superiorLTM storage of early words does contribute toprimacy in neurologically intact people, thenprimacy in an amnesic patient should still be atte-nuated relative to controls, even when the patientand controls are matched for cumulative rehearsal.

We ran two experiments to examine this ques-tion: First (Experiment 1a), we assessed the serialposition curve and rehearsal pattern in a patientwith severe and isolated impairment of LTM

storage and in a group of controls, as in Brown,Della Sala, et al. (2007). We then examinedwhether primacy could be restored to a normallevel in this amnesic patient when her rehearsalwas matched with that of controls via an externalizedcumulative rehearsal schedule (Experiment 1b).

EXPERIMENT 1A: FREE REHEARSAL

Method

ParticipantsA severely amnesic patient—patient D.A.—as wellas 18 young neurologically intact controls (stu-dents and research staff), aged 18–34 years, volun-teered to take part in this experiment (most of thecontrol data were taken from the previously pub-lished study by Brown, Della Sala, et al., 2007).All participants were completely naıve and blindwith respect to the general subject matter andgoals of this experiment.

Patient historyAt the time of testing, D.A. was a 31-year-oldwoman with 11 years of education. D.A. has ahistory of dense anterograde amnesia andcomplex partial seizures, which are assumed to beof MTL origin. It is strongly suspected that bothher amnesia and seizures are the result of birthasphyxia. At school, she showed profoundmemory impairment, as evinced by a schoolreport, in which it is stated that D.A., then aged13, “cannot remember something that she hasread, for example 5 minutes ago”. Psychologicalassessment a year later confirmed that D.A. had“severe memory difficulties for verbal information”.

D.A.’s complex partial seizures responded wellto medication (carbamazepine), and her electroen-cephalography (EEG) and computed tomography(CT) scan were normal.

Following school, D.A. worked as a nurseryassistant but resigned due to her memory impair-ment, which, according to a family member, haddeteriorated substantially. A subsequent neurologi-cal assessment led to the diagnosis of a focal memorydeficit. She presented with dense amnesia not only

388 COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6)

DEWAR, BROWN, DELLA SALA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

for recently presented material such as words butalso for autobiographical information, includingtypically highly memorable episodes such as herown wedding or the birth of her children.

Neuropsychological profile of patientD.A.’s memory impairment was classified as “verysevere” according to the Rivermead BehaviouralMemory Test (screening score ¼ 2; Wilson,Cockburn, & Baddeley, 1985).

During detailed neuropsychological testing,D.A. presented with a profound LTM impairment.Her 30-minute delayed recall of the WechslerMemory Scale–Third Edition (WMS–III;Wechsler, 1998) stories was at floor (0), and shedid not benefit from a reminder of the generaltheme of the story material. Moreover, her per-formance at word list learning (WMS–III) wasbelow the average range, and D.A. was unable torecall any of the presented words following a dis-tractor list or a 30-minute delay.

D.A.’s LTM impairment was also apparent inthe visual domain as evinced by her inability toreproduce a previously well-copied complexfigure after a delay (Caffarra, Vezzadini, Dieci,Zonato, & Venneri, 2002).

In contrast to this severe LTM impairment,D.A. showed normal STM functioning, asevinced by her low average score on the WMS–III immediate story recall and a normal digitspan (WMS–III; 6 forwards and 5 backwards).Her executive function was also found to be pre-served, as demonstrated by normal performanceon Part B of the Trail Making Test (Giovagnoliet al., 1996) as well as on phonological and seman-tic verbal fluency (Tombaugh, Kozak, & Rees,1999). D.A. also presented with normal languagefunction, as tested via the Graded Naming Test(McKenna & Warrington, 1983), and herNational Adult Reading Test (NART)-estimatedIQ (Nelson, 1982) was in the average range (97).

D.A.’s scores on these tests are provided inTable 1.

The study was approved by the local EthicalCommittee, and informed consent was obtainedfrom each participant according to the Declarationof Helsinki.

Materials and procedureThe procedure of this experiment replicated thatof Brown, Della Sala, et al. (2007). All participantsundertook 16 free rehearsal trials, each of whichcontained 16 words, and 4 practice trials (2 con-taining 4 words, and 2 containing 8 words).

The words were presented to the participantsvisually for 1 s each, with interstimulus intervalsof 3 s.

In each trial, participants were asked to readaloud each word as it appeared and to try toremember as many of them as possible.Participants were informed that there would belong gaps between each word and that theyshould use these gaps to try to remember andrehearse aloud any words from the current listthat came to mind. Such overt free rehearsal doesnot alter the serial position curve observed underconditions of covert free rehearsal (Horton,1976; Murdock & Metcalfe, 1978; Wixted &McDowell, 1989). Following the last item, partici-pants were prompted to try and recall as many ofthe list words that they could, in any order. Theentire session was recorded to allow transcribingand scoring of rehearsal and recall.

Results

Serial positionWhen performance was plotted by serial position,the controls showed the typical serial positioncurve, displaying clear primacy and recency (seeFigure 1A). In line with previous amnesia findings,D.A. showed recency but no primacy (see Figure1C). Modified t tests for use with single-casedata (Crawford & Garthwaite, 2002) revealedthat the controls recalled significantly more earlylist words than did D.A. (Serial Positions 1–4,p , .05). However, no statistical differenceemerged between D.A.’s and the controls’ latelist words (Serial Positions 13–16, p . .1).

Position of last rehearsal and rehearsal patternWhen performance was plotted by position of lastrehearsal (Tan & Ward, 2000; see introduction),primacy, but not recency, substantially decreasedin the controls (see Figure 1B). As can be seen in

COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6) 389

RESTORING PRIMACY IN AMNESIA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

Table 1. Neuropsychological profile of D.A.

Clinical test Score Reference

Rivermead Behavioural Memory Test Wilson et al. (1985)

Screening 2a

Classification very severea

Wechsler Memory Scale III Wechsler (1998)

Prose Recall–Immediate 25

Prose Recall–Delayed 0a

Word List Learning–Total Immediate 20a

Word List–Delayed Recall 0a

Digit Span 6

Rey–Osterrieth Complex Figure Caffarra et al. (2002)

Copy 35

Delayed Recall 0a

Trail Making B 99 Giovagnoli et al. (1996)

FAS 34 Tombaugh et al. (1999)

Animals 21 Tombaugh et al. (1999)

Graded Naming Test 15 McKenna & Warrington (1983)

NART–estimated IQ 97 Nelson (1982)

Note: Scores are raw scores. NART ¼ National Adult Reading Test.aPerformance in the abnormal range (≤5th percentile).

Figure 1. Free recall serial position curves in the free rehearsal condition. Left column: Free recall plotted by serial position for (A) the controls

and (C) D.A. Right column: Free recall plotted by position of last rehearsal for (B) the controls and (D) D.A. Position of last rehearsal is

defined as the interitem gap (1–16), in which a list word was last rehearsed overtly.

390 COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6)

DEWAR, BROWN, DELLA SALA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

Figure 1D, D.A.’s performance remained virtuallythe same as that in the serial position graph.

Analysis of the rehearsal protocols revealed thatthe controls used the interitem gaps to rehearseseveral of the words, and that they rehearsed theearly list words throughout list presentation (seeFigure 2) in such a way as to carry them forwardthrough the list.

In contrast, D.A. only ever rehearsed the mostrecently presented word in the interitem gaps(see Figure 2), despite much encouragement torehearse cumulatively.

EXPERIMENT 1B: IMPOSEDCUMULATIVE REHEARSAL

In Experiment 1b, we examined whether primacycould be restored to a normal level in D.A. whenher rehearsal was matched with that of controls viaan externalized cumulative rehearsal schedule.These trials were designed to mimic typical cumulat-ive rehearsal without the need for self-initiatedrehearsal and were based on the rehearsal schedulesof the control participants of Experiment 1a.

Method

ParticipantsPatient D.A. as well as 12 new young neurologicallyintact controls, aged 25–35 years, volunteered to

take part in this experiment. One of the controlswas selected to be matched very closely with D.A.for age (34 years), sex, education (11 years), andoccupational background (nursery nurse). All par-ticipants were completely naıve and blind withrespect to the general subject matter and goals ofthis experiment.

Materials and procedureAs in Experiment 1a, there were 16 trials, each ofwhich contained 16 words, and 4 practice trials (2containing 4 words, and 2 containing 8 words).The words were presented visually for 1 s each,and with interstimulus intervals of 3 s.

In contrast to the free rehearsal trials inExperiment 1a, the 3-s gaps were filled with 3visually presented words that had been previouslypresented in that trial. The words selected ineach of the 16 gaps were chosen to reflect typicalcumulative free rehearsal patterns as evidenced bythe controls in Experiment 1a, and in particularto mirror the tendency for early list items to be“carried forward” through the lists, while retainingan “even” rate of rehearsal of three words in eachinteritem gap. (The latter requirement made itundesirable to copy controls’ schedules directly,and there is in any case no straightforward wayto average many different protocols arithmeti-cally.) Two different rehearsal protocols wereused; one was assigned to odd-numbered listsand one to even-numbered lists (exact protocolsare available from the authors).

In each trial, participants were asked to readaloud each word as it appeared and to try toremember them for later recall. They were toldthat if the same words were presented more thanonce, they should simply read it out and rememberit each time it appeared. Following the last item,participants were prompted to try and recall asmany of the list words that they could remember,in any order.

Results

As shown in Figure 3, D.A. showed clear primacyas well as recency under these conditions ofimposed cumulative rehearsal. Indeed, her

Figure 2. Average movement due to rehearsal between position of

presentation and position of last rehearsal for the controls and

D.A. in the free rehearsal condition.

COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6) 391

RESTORING PRIMACY IN AMNESIA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

performance did not differ significantly from thatof the controls at any of the serial positions (allps . .1, modified t tests for use with single-casedata; Crawford & Garthwaite, 2002). These datathus reveal that primacy could be restored in theamnesic patient D.A. when a cumulative rehearsalstrategy was imposed.

GENERAL DISCUSSION

The present study shows that D.A., a denselyamnesic patient who presents with impairedprimacy (see Figure 1C), utilizes an atypical “fixedrehearsal” strategy when asked to rehearse previouslist words during interitem gaps (Experiment 1a).Thus, rather than using provided interitem gaps torehearse several presented words throughout listpresentation, as is typical for neurologically intactpeople (Tan & Ward, 2000; see also Figure 2),D.A. almost only ever rehearsed the most recentlypresented word (see Figure 2)—for example,“TASK, task, task–FLIP, flip, flip–GOLF, golf,golf . . . ” (capitals ¼ presented words, italics ¼rehearsed words). These results replicate the find-ings of impairment in primacy and cumulativerehearsal in patient S.J. (Brown, Della Sala, et al.,2007), whose aetiology overlapped in part withthat of D.A.: S.J.’s isolated amnesia was also associ-ated with hypoxia and consequent malfunction ofthe MTL, but was of acquired rather than congeni-tal origin. The present data thus bolster the evidencefor the hypothesis that defective cumulative

rehearsal contributes to the attenuated primacy inamnesic patients, at least in those with MTLmalfunction.

Importantly, the results of the novelExperiment 1b suggest that such deficit in cumu-lative rehearsal can in fact account for all of theattenuation of primacy in amnesia. D.A.’sprimacy performance was very similar to that ofthe controls when cumulative rehearsal wasmatched for D.A. and the controls by means ofpresentation of previous list words during theinteritem gaps (see Figure 3). If primacy dependedin part on superior LTM storage of early list items,then D.A.’s primacy performance should havebeen attenuated relative to that of the controls,even after the matching of cumulative rehearsalin D.A. and the controls. While D.A.’s perform-ance on the first serial position was slightly inferiorto that of the average control, this difference wasnonsignificant, and no difference emergedbetween D.A. and her (age-, education-, and occu-pation background-) matched control whose recallprobability for the first four list words was .75, .75,.75, and .38, respectively (D.A.: .75, .81, .81, .25).Moreover, Figure 3 shows that the subtle differ-ence between D.A. and the controls was consistentacross the curve, thus indicating that the differenceon the first word was not specific to this item andprobably reflected differences in general memoryor cognitive ability.

Data from amnesic patients have provided keyevidence for the traditional dual-store theory ofimmediate recall, claiming that primacy is theproduct of superior LTM storage of early listitems (Atkinson & Shiffrin, 1968; Glanzer,1972). The present patient data challenge this tra-ditional theory and instead provide strong cre-dence to the theory that primacy is the productof recency of early list items, brought about bycumulative rehearsal (Tan & Ward, 2000). Ourdata suggest that such cumulative rehearsal isdependent upon the integrity of the MTL, thusmaking impaired cumulative rehearsal the cruciallink between amnesia and attenuated primacy—that is, amnesia leads to impaired cumulativerehearsal, which consequentially leads to impairedprimacy. We are confident that it is the amnesia

Figure 3. Free recall plotted by serial position in the cumulative

rehearsal condition for D.A. and the control group.

392 COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6)

DEWAR, BROWN, DELLA SALA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

that caused the impaired cumulative rehearsal inD.A. and patient S.J. (Brown, Della Sala, et al.,2007) since both patients presented with isolatedanterograde amnesia. Indeed, both D.A. and S.J.failed to engage in cumulative rehearsal despitemuch encouragement from the experimenter,thereby arguing against the possibility thatamnesic patients do not use cumulative rehearsalspontaneously. It is also unlikely that amnesicpatients fail to use cumulative rehearsal becauseof a lack of understanding of the instructions.Both patients understood the instructions andattempted such rehearsal. However, crucially,they both stated that doing so was too difficultfor them and thus resorted to their preferredfixed rehearsal strategy.

An executive deficit hypothesis is also unlikelygiven that D.A. performed normally on executivetests and never showed dysexecutive symptoms.Such was also true for patient S.J. D.A.’s normalTrail Making B performance in particular istelling since this test assesses task-switchingability, which is likely to be implicated whenmoving repeatedly between rehearsals and theencoding of novel list words. An executive “task-switching” deficit hypothesis of D.A.’s rehearsaldeficit therefore appears unlikely.

How might amnesia impair cumulative rehear-sal? This is an important question to answer if wewish to gain better insight into how the MTLsystem is implicated in primacy and cumulativerehearsal in neurologically intact people.

Cumulative rehearsal essentially involves theretrieval of prior list words following presentationand encoding of novel words and as such has beenhypothesized to overlap substantially with theprocess of free immediate recall (Laming, 2006;Tan & Ward, 2000). Any impairment of thememory retrieval process could thus impede suchrehearsal substantially. However, D.A. had anormal digit span (6 digits), and she was able toretrieve several recent list words at immediaterecall, both when using fixed rehearsal (mean ¼3.4 words, maximum recall ¼ 5) and when usingexternalized cumulative rehearsal (mean ¼ 4.25words, maximum recall ¼ 6), thus indicating thather retrieval was intact. The same was true for

patient S.J. We note that D.A.’s immediate recallwas lower in the free rehearsal (“fixed” rehearsal)condition than in the imposed cumulative rehear-sal condition and digit span. This is likely to reflectthe fact that the recency portion in the free (fixed)rehearsal condition was occupied by only a fewwords and their repetitions (14, 14, 15, 15, 15,16), while the same temporal space was occupiedby a larger number of different words in theimposed cumulative rehearsal condition (e.g., 1,15, 14, 3, 2, 16) and digit span (1, 2, 3, 4, 5, 6).

In contrast to the retrieval at the end of the list,retrieval during the interitem gaps was timelimited (3 s). This time limitation might haverevealed subtle retrieval deficits such as slowedset searching, which could go unnoticed in theabsence of time restraints. However, like neurolo-gically intact people, D.A. typically recalled wordswithin fast succession at immediate recall, indicat-ing that she was able to access these words rapidly.

While a spared rehearsal system such as thephonological loop (Baddeley & Hitch, 1974) orfocus of attention (Cowan, 2001) is likely to be suf-ficient for continuous verbal rehearsal or immedi-ate verbal recall at span level, the retrieval ofpreviously rehearsed words following new distract-ing encoding might require different or additionalprocesses subsumed by the MTL. Evidence forsuch a view can be gleaned from the early workon patient H.M. who could rehearse new infor-mation for several minutes but forgot this infor-mation the moment his rehearsal was interrupted(Scoville & Milner, 1957). H.M. could not main-tain this information offline during distraction,and so it became inaccessible thereafter. Inkeeping with this hypothesis, it has been demon-strated via functional imaging that MTL structuresare required for the reactivation—that is, retrie-val—of recently rehearsed material if such rehearsalhas been interrupted by a brief period of distraction(Sakai, Rowe, & Passingham, 2002). This is notthe case for the retrieval of material that has beenmaintained in mind continuously via uninterruptedrehearsal right until the time of retrieval. In thiscase, executive (i.e., frontal) processes appear tosuffice for the purpose of retrieval (Sakai et al.,2002). Presumably, during cumulative rehearsal,

COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6) 393

RESTORING PRIMACY IN AMNESIA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

the MTL acts as a temporary back-up for recentlyrehearsed material during distraction and sobecomes necessary for postdistraction retrieval.This hypothesis meshes well with recent functionalmagnetic resonance imaging (fMRI) findings byFaraco et al. (2011). They examined MTL activityin a complex WM task requiring neurologicallyintact participants to retain a sequence of letterswhose encoding was interspersed with short arith-metic problems. While both the complex WM taskand the arithmetic task alone activated the hippo-campus, this hippocampal activity was significantlylarger in the complex WM task, suggesting that themaintenance of new verbal material during concur-rent cognitive activity requires the MTL, possiblyin the form of a temporary back-up store (Faracoet al., 2011). Further evidence for this hypothesiscomes from a recent study on an MTL amnesiapatient who showed intact performance on asimple verbal WM span task but impaired per-formance on a complex verbal WM span task, inwhich continuous rehearsal was disrupted by sec-ondary task operations (Rose, Olsen, Craik, &Rosenbaum, 2012). Together with several otherstudies (e.g., Axmacher et al., 2007; Faraco et al.,2011; Olson et al., 2006; Ranganath & Esposito,2001; Rose et al., 2012; Warren, Duff, Tranel, &Cohen, 2011), our patient findings challenge thetraditional notion that WM function is indepen-dent from LTM (Atkinson & Shiffrin, 1968) andthe MTL (Alvarez et al., 1994; Cave & Squire,1992). In an attempt to bridge this traditionaldivide between WM and LTM, Baddeley (2000)proposed the “episodic buffer”, which (a) ishypothesized to provide a temporary store forinformation too large and complex for the phono-logical loop, and (b) is thought to be impaired inmany patients with amnesia (Baddeley, 2000;Baddeley & Wilson, 2002). Moreover, recentwork suggests MTL involvement in episodicbuffer function (Berlingeri et al., 2008; Rudner &Ronnberg, 2008; though see also Baddeley,Jarrold, & Vargha-Khadem, 2011).

We propose that it is damage to such a tempor-ary back-up store that causes the deficit in cumulat-ive rehearsal and primacy in amnesic patients, withno additional contributions from impaired LTM

storage. It remains to be established whetheramnesic patients with intact episodic buffer func-tion (see Baddeley & Wilson, 2002) also showintact cumulative rehearsal and primacy, as wouldbe predicted from our hypothesis. Future work isalso needed to examine whether amnesic patientslike D.A. show primacy when explicitly instructedto rehearse only the very first word throughout thelist. It is likely that such could be achieved withouta temporary back-up store since capacity demandson WM would be somewhat reduced. Indeed,even though D.A. used fixed rehearsal in themajority of trials, she was able to carry forwardthe first word to the second position in sometrials, as shown in Figure 2.

We posit that a temporary back-up store such asthe episodic buffer allows neurologically intactpeople to (a) back up recently encoded andrehearsed words for a limited time, and consequen-tially (b) to rehearse them in a noncontinuous wayby revisiting and rerehearsing them following theencoding and rehearsal of intervening list words.Our data suggest that such temporary back-upand rehearsal capacity is not specific or superiorfor early list words. Figure 2 shows that evenmiddle list words were typically carried overacross at least a couple of positions, indicatingthat they could be reaccessed for rehearsal follow-ing new encoding and rehearsals. In other words,we propose that primacy arises because we choseto make use of this back-up and rehearsal capacityfor early list words especially. Doing so allows thesewords to be carried forward to the recency portionof the list and therefore to morph into recencyitems, as suggested by Tan and Ward (2000). Wenote that in addition to being rehearsed throughthe list, first list items can also benefit fromhigher distinctiveness by virtue of their specialtemporal status in the list (Brown, Neath, &Chater, 2007; Strange, Otten, Josephs, Rugg, &Dolan, 2002; Tan & Ward, 2000). However, nosuch effects could be detected in the data to hand.

Manuscript received 26 June 2011

Revised manuscript received 7 February 2012

Revised manuscript accepted 7 February 2012

First published online 23 February 2012

394 COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6)

DEWAR, BROWN, DELLA SALA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

REFERENCES

Alvarez, P., Zola-Morgan, S., & Squire, L. R. (1994).The animal model of human amnesia: Long-termmemory impaired and short-term memory intact.Proceedings of the National Academy of Sciences of the

United States of America, 91, 5637–5641.Atkinson, R. C., & Shiffrin, R. M. (1968). Human

memory: A proposed system and its control pro-cesses. In J. T. Spence & S. K. W. Spence (Eds.),The psychology of learning and motivation: Advances

in research and theory (pp. 89–195). New York, NY:Academic Press.

Axmacher, N., Mormann, F., Fernandez, G., Cohen,M. X., Elger, C. E., & Fell, J. (2007). Sustainedneural activity patterns during working memory inthe human medial temporal lobe. The Journal of

Neuroscience, 27, 7807–7816.Baddeley, A. (2000). The episodic buffer: A new com-

ponent of working memory? Trends in Cognitive

Sciences, 4, 417–423.Baddeley, A, & Hitch, G. (1974). Working memory. In

G. Bower (Ed.), The psychology of learning and motiv-

ation (pp. 47–89). New York, NY: Academic Press.Baddeley, A., Jarrold, C., & Vargha-Khadem, F. (2011).

Working memory and the hippocampus. Journal of

Cognitive Neuroscience, 23, 3855–3861.Baddeley, A., & Warrington, E. (1970). Amnesia and

the distinction between long- and short-termmemory. Journal of Verbal Learning and Verbal

Behavior, 9, 176–189.Baddeley, A., & Wilson, B. A. (2002). Prose recall and

amnesia: Implications for the structure of workingmemory. Neuropsychologia, 40, 1737–1743.

Berlingeri, M., Bottini, G., Basilico, S., Silani, G.,Zanardi, G., Sberna, M., et al. (2008). Anatomy ofthe episodic buffer: A voxel-based morphometrystudy in patients with dementia. Behavioural

Neurology, 19, 29–34.Brown, G. D. A., Chater, N., & Neath, I. (2008). Serial

and free recall: Common effects and common mech-anisms? A reply to Murdock (2008). Psychological

Review, 115, 781–785.Brown, G. D. A., Della Sala, S., Foster, J. K., &

Vousden, J. I. (2007). Amnesia, rehearsal, and tem-poral distinctiveness models of recall. Psychonomic

Bulletin & Review, 14, 256–260.Brown, G. D. A., Neath, I., & Chater, N. (2007). A

temporal ratio model of memory. Psychological

Review, 114, 539–576.

Brown, G. D. A., Vousden, J. I., & McCormack, T.(2009). Memory retrieval as temporal discrimination.Journal of Memory and Language, 60, 194–208.

Caffarra, P., Vezzadini, G., Dieci, F., Zonato, F., &Venneri, A. (2002). Rey–Osterrieth complexfigure: Normative values in an Italian populationsample. Neurological Sciences, 22, 443–447.

Capitani, E., Della Sala, S., Logie, R. H., & Spinnler,H. (1992). Recency, primacy, and memory:Reappraising and standardising the serial positioncurve. Cortex, 28, 315–342.

Carlesimo, G. A., Fadda, L., Sabbadini, M., &Caltagirone, C. (1996). Recency effect inAlzheimer’s disease: A reappraisal. The Quarterly

Journal of Experimental Psychology A, 49, 315–325.Cave, C. B., & Squire, L. R. (1992). Intact verbal and

nonverbal short-term memory following damage tothe human hippocampus. Hippocampus, 2, 151–163.

Cermak, L. S., Naus, M. J., & Reale, L. (1976).Rehearsal strategies of alcoholic Korsakoff patients.Brain and Language, 3, 375–385.

Cowan, N. (2001). The magical number 4 in short-termmemory: A reconsideration of mental storagecapacity. Behavioral and Brain Sciences, 24, 87–114.

Crawford, J. R., & Garthwaite, P. H. (2002).Investigation of the single case in neuropsychology:Confidence limits on the abnormality of test scoresand test score differences. Neuropsychologia, 40,

1196–1208.Faraco, C. C., Unsworth, N., Langley, J., Terry, D., Li,

K., Zhang, D., et al. (2011). Complex span tasks andhippocampal recruitment during working memory.NeuroImage, 55, 773–787.

Fischler, I., Rundus, D., & Atkinson, R. C. (1970).Effects of overt rehearsal procedures on free recall.Psychonomic Science, 19, 249–250.

Giovagnoli, A. R., Del Pesce, M., Mascheroni, S.,Simoncelli, M., Laiacona, M., & Capitani, E.(1996). Trail making test: Normative values from287 normal adult controls. Italian Journal of

Neurological Sciences, 17, 305–309.Glanzer, M. (1972). Storage mechanisms in recall. In

G. H. Bower (Ed.), The psychology of learning and

motivation: Advances in research and theory

(pp. 129–193). New York, NY: Academic Press.Glanzer, M., & Meinzer, A. (1967). The effects of

intralist activity on free recall. Journal of Verbal

Learning and Verbal Behavior, 6, 928–935.Horton, K. D. (1976). Phonemic similarity, overt

rehearsal, and short-term store. Journal of

COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6) 395

RESTORING PRIMACY IN AMNESIA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013

Experimental Psychology: Human Learning &

Memory, 2, 244–251.Laming, D. (2006). Predicting free recalls. Journal of

Experimental Psychology. Learning, Memory &

Cognition, 32, 1146–1163.McKenna, P., & Warrington, E. K. (1983). The Graded

Naming Test. Windsor, UK: NFER-Nelson.Meudell, P., & Mayes, A. (1980). Do alcoholic amnesic

patients passively rehearse information? Brain and

Language, 204, 189–204.Milner, B., Squire, L. R., & Kandel, E. R. (1998).

Cognitive neuroscience and the study of memory.Neuron, 20, 445–453.

Murdock, B., & Metcalfe, J. (1978). Controlled rehear-sal in single-trial free recall. Journal of Verbal

Learning and Verbal Behavior, 17, 309–324.Neath, I., & Brown, G. D. A. (2006). Further

applications of a local distinctiveness model ofmemory. Psychology of Learning and Motivation, 46,

201–243.Nelson, H. E. (1982). National Adult Reading Test

(NART): Test manual. Windsor, UK: NFER-Nelson.

Olson, I. R., Moore, K. S., Stark, M., & Chatterjee, A.(2006). Visual working memory is impaired whenthe medial temporal lobe is damaged. Journal of

Cognitive Neuroscience, 18, 1087–1097.Ranganath, C., & Esposito, M. D. (2001). Medial tem-

poral lobe activity associated with active maintenanceof novel information. Neuron, 31, 865–873.

Rose, N. S., Olsen, R. K., Craik, F. I. M., & Rosenbaum,R. S. (2012). Working memory and amnesia:The role of stimulus novelty. Neuropsychologia, 50,

11–18.Rudner, M., & Ronnberg, J. (2008). The role of the epi-

sodic buffer in working memory for language proces-sing. Cognitive Processing, 9, 19–28.

Rundus, D. (1971). Analysis of rehearsal processes in freerecall. Journal of Experimental Psychology, 89, 63–77.

Sakai, K., Rowe, J. B., & Passingham, R. E. (2002).Parahippocampal reactivation signal at retrievalafter interruption of rehearsal. The Journal of

Neuroscience, 22, 6315–6320.Scoville, W. B., & Milner, B. (1957). Loss of recent

memory after bilateral hippocampal lesions. Journal

of Neurology, Neurosurgery, and Psychiatry, 20, 11–21.Strange, B. A, Otten, L. J., Josephs, O., Rugg, M. D., &

Dolan, R. J. (2002). Dissociable human perirhinal,hippocampal, and parahippocampal roles duringverbal encoding. The Journal of Neuroscience, 22,

523–528.Tan, L., & Ward, G. (2000). A recency-based account

of the primacy effect in free recall. Journal of

Experimental Psychology: Learning, Memory &

Cognition, 26, 1589–1625.Tombaugh, T., Kozak, J., & Rees, L. (1999). Normative

data stratified by age and education for two measuresof verbal fluency FAS and animal naming. Archives of

Clinical Neuropsychology, 14, 167–177.Warren, D. E., Duff, M. C., Tranel, D., & Cohen, N. J.

(2011). Observing degradation of visual represen-tations over short intervals when medial temporallobe is damaged. Journal of Cognitive Neuroscience,23, 3862–3873.

Wechsler, D. (1998). Wechsler Memory Scale–Third

Edition (UK). London, UK: The PsychologicalCorporation.

Wilson, B., Cockburn, J., & Baddeley, A. D. (1985).The Rivermead Behavioural Memory Test. Reading,UK: Thames Valley Test Group.

Wixted, J. T., & McDowell, J. J. (1989). Contributionsto the functional analysis of single-trial free recall.Journal of Experimental Psychology: Learning,

Memory & Cognition, 15, 685–697.

396 COGNITIVE NEUROPSYCHOLOGY, 2011, 28 (6)

DEWAR, BROWN, DELLA SALA

Dow

nloa

ded

by [

Uni

vers

ity o

f E

dinb

urgh

] at

14:

43 1

6 Se

ptem

ber

2013