Priming and the Brain. Schacter, 1988

8162019 Priming and the Brain Schacter 1988

httpslidepdfcomreaderfullpriming-and-the-brain-schacter-1988 111

Neuron Vol 20 185ndash195 February 1998 Copyright 9832091998 by Cell Press

Priming and the Brain Review

minutes to several hours subjects are given three-letterDaniel L Schactersect and Randy L BucknerdaggerDagger

Dep artment of Psyc hology word st ems wit h mult ip le p ossib le c omp letions and areasked to complete each stem with the first word thatHarvard University

Cam bridge M as sac huset ts 02138 c om es t o m ind (eg m ot___ f ort het arget w ord ldquom ot elrdquo)Priming occurs when subjects complete the stem withdaggerDepartment of Psychology

Washington University a designated target completion more often for wordsthat had been studied earlier than for words that wereSt Louis Missouri 63130

DaggerMGH-NMR Center not studied previously In a related kind of experimentsubjects study a list of words and are later given a wordCharlestown Massachusetts 02129identification test where words are flashed briefly (eg

Introduction for 35 ms)and subjects attempt to identifythem Primingoccurs when subjects identify more previously studied

Research in cognitive psychology neuropsychology words than nonstudied words Similar kinds of tasksand neuroscience has converged recently on the idea have been developed for examining priming of numer-that memory is composed of dissociable forms and sys- ous different types of stimuli including pseudowordstems (Squire 1992 Roediger and McDermott 1993 (Keane et al 1995 Bowers 1996) familiar and unfa-

Schacter and Tulving1994 Willingham 1997) This con- miliar objects (Biederman and Cooper 1991 Srinivasclusion has been based on experimental and theoretical 1993 Schacter et al 1993b) visual patterns (Musenanalyses of a variety of different phenomena of learning and Squire 1992) and environmental sounds (Chiu andand memory One of the most intensively studied of Schacter 1995 for extensive reviews see Roedigersuch phenomena is known as pr iming a change in the and McDermott 1993 Schacter et al 1993a)ability to identify or produce an item as a result of a Cognitive studies of healthy volunteers have providedspecific prior encounter with the item (Tulving and a great deal of information about the behavioral andSchacter 1990) psychological properties of priming (for discussion of

Priming represents an example of what is often re- recent research see Wiggs and Martin1998) Forexam-ferred to as implicit memory (Graf and Schacter 1985 ple it is known from numerous cognitive studies thatSchacter1987 Roediger1990)or nondeclarative mem- explicit recall and recognition of recently studied wordsory (Squire 1992)mdash a nonconscious influence of past objects or other materials is greatly affected by howexperienceon current performance or behavior Priming those materials are initially encoded during the studyand related types of implicit memory had been noted phase of an experiment (eg Craik and Tulving 1975)occasionally since the nineteenth c entury in clinical and More specificallywhen target itemsare initially encodedexperimental reports but have only been studied sys- in a semantic or elaborative manner (eg generatingtematically during the p ast two d ecades (for a historical meanings of a word or rating its pleasantness) subse-review see Schacter 1987) Modern interest in priming quent explicit recall or recognition is much more accu-was stimulated by several developments studies con- rate than when target items are initially encoded in acerning lexical access and representation that provided nonsemantic or shallow manner (eg counting the num-methods for assessing priming (eg Scarborough et al ber of vowels and consonants in a word) In an early1977) early theoretical attempts to distinguish among study of priming Jacoby and Dallas (1981) reporteddifferent forms of memory (eg Tulving 1972) and re- that p riming effects on a word identification test are noports of preserved learning in patients with amnesic greater after semantic encoding than after nonsemanticsyndromes (eg Milner et al 1968 Warrington and

encodingmdash even though recognition accuracy was muchWeiskrantz 1974 see below for discussion) Based on

higher after the semantic than the nonsemantic encod-these developments a number of experiments reported

ing task (although some subsequent studies have foundduring the early 1980rsquos helped to characterize the phe-

small effects of semantic versus nonsemantic encod-nomenon of priming and distinguish it from explicit

ing manipulations on word identification and similarmemory (cf Jacoby and Dallas 1981 Tulving et al

tasks they are much smaller than the effects of the1982 Graf et al 1984)same encoding manipulations on explicit memory for

Priming is assessed with experimental tasks that doa review and discussion see Brown and Mitchell 1994)

not require conscious recollection of particular previousOther early data established that healthy volunteers can

experiences Such t asks differ from traditional tests ofshow p riming even when they fail to c onsciously recol-

recall and recognition which tap an explicit or declara-lect that they have previously studied an item and that

tive form of memory and require conscious intentionalpriming can persist across days (Tulving et al 1982)

recollection of recently studied information Considermore recent data indicate that priming effects can per-

for example a prototypical experiment in which p rimingsist for as long as a year (eg Cave 1997)

is assessed with a word-stem completion test In theWhereas the foregoing studies involved college stu-

first phase of the experiment subjects study a seriesdents research on priming quickly broadened to include

of words After a delay that usually ranges from severala variety of subject populations For example studiescomparing elderly adults with young adults have shownthat age-related decreases in priming tend to be smallersect To whom correspondence should be addressed



Neuron186

than age-related decreases in explicit recall and recog- and Weiskrantz had uncovered a form of memory that isfully preserved in amnesic patients or whether fragmentnition (see Mitchell 1993 La Voie and Light 1994)Like-

wise studies of priming in young children have shown cueing performance like other kinds of memory is im-paired by amnesiaa relative invariance of priming in younger and older

children (eg from 3ndash10 years of age) even though Studies conducted during the early 1980rsquos clarified

these findings by demonstrating that performance onexplicit memory shows strong d evelopmental t rendsacross the same time period (see Parkin 1993 Naito the word-stem completion task depends critically on

the instructions given to subjects (these instructionsand Komatsu 1993) Studies of patients with variouspsychopathological conditions including depression were not systematically examined in the Warrington and

Weiskrantz studies) When instructed to complete aand anxiety have also revealed robust p riming effects(see M ineka and Nugent 1995) st em w it h t he first word t hat comes t o mind amnesic

patients showed a normal priming effect compared toCognitive studies have been essential in delineatingthe basic properties of priming but to obtain a fuller control subjects but when instructed to produce the

correct item from the study list amnesic patients showedunderstanding of the phenomenon it is necessary tounderstand its neural underpinnings Here we review less recall than did controls (Graf et al 1984) Similar

patterns were observed when amnesic patients wereevidence concerning thebrain systems that are involvedin primingfocusing on studiesof patients with neurolog- given related kindsof priming and recall tasks (Schacter

1985) Normal priming in amnesic patients has sinceical damage and on experiments that have used brainimaging techniquesmdash positron emissiontomography(PET) been observed across a variety of different tasks and

conditions (for reviews see Schacter et al 1993aand functional magnetic resonance imaging (fMRI)mdash toexplore the neural correlatesof priming in healthy volun- Squire 1992)

Overall evidence gathered during the p ast 15 yearsteers We believe that characterizing priming at the levelof brain systems provides an important foundation that supports a distinction between explicit memory which

is impaired in amnesic patients and depends on thecan inform and guide further explorations of priming atc ellular and molec ular levels med ial t emp oral diencep halic regions t hat are d amaged

in amnesia and priming which is preserved in amnesiaand does not depend on medial temporaldiencephalic

Neuropsychology of Priming Insights from structures Although the possibility has been raised thatStudies of Brain-Injured Patients word priming effects on word completion and identifica-P r i m i n g a n d t h e A m n e s i c S y n d r o m e tion tests are not preserved in amnesic patients ap-The amnesic syndrome occurs as a consequence of pearing to be normal because of methodological prob-damage to structures in the medial temporal lobe and lems and limitations of relevant studies (Ostergaard anddiencephalon and is characterized by a severe impair- Jernigan 1993) experiments that have addressed pos-ment in the explicit recollection of everyday facts and sible methodolog ical problems still reveal normal prim-

events together with preservation of general intelli- ing in amnesia (Hamann et al 1995) Indeed recentgence (Squire 1992) The syndrome w as first b rought evidence indicates that priming of word completion andto the widespread attention of neuroscientists during word identification performance can be fully preservedthe 1950rsquos when Milner and her colleagues described even when a patient is unable to attain above-chancethe famous patient H M who developed profound am- scores on forced-choice tests of explicit recognitionnesia for recent events following bilateral medial tempo- memory (Hamann and Squire 1997a) This kind of find-ral lobe resection for the relief of intractable epilepsy ing helps to ruleout the possibility that priming in amne-(eg Milner et al 1968) Although it initially appeared sic patients is mediated by residual explicit memorythat H M had no ability to retain new information of abilities and instead supports the conclusion that prim-any kind Milner et al (1968) reported that H M could ing is mediated by processes or systems that are notlearn new motor skills in a relatively normal manner essential for explicit memory and are spared in amnesia

Early interpretations attributed the findings of Milner V ari et i es of P ri mi ng Dat a and Theoret i cal I mpl i cat i ons and c olleagues to the uniqueness of motor memory In considering the generality of the conclusiont hat prim-However evidence obtained from other amnesic pa- ing is preserved in amnesic patients it is important totients soon raised the possibilit y that preserved memory note that priming can be fractionated into several differ-

function in amnesics might extend beyond the domain ent subtypes One basic distinct ion is between p e r c e p - of motor skill acquisitionIn aseriesof important studies tu a lp r im in g and conceptual pr iming Perceptual primingWarrington and Weiskrantz (1974)reported that amnesic is modality specific and does not depend on semanticpatients sometimes show relatively preserved memory or elaborative encoding of an item at the time of studyperformance when they are given word beginnings as whereas conceptualp riming is not modality specific andcues for previously studied items This was a striking benefits from semantic encoding (Blaxton 1989) Prim-finding because patients demonstrated little explicit ing on a word identification test (where subjects tryrecollection of the same items when they were shown to identify briefly flashed items) represents a relativelyan entire word and were required to make an oldnew ldquopurerdquo form of perceptual priming the priming effect is

judg ment about it However in Warringt on and Weisk- eliminated by modality change between the study andrantzrsquos work the amnesic patientsrsquo performance on the test phases of the experiment (visual-to-aud itory or au-fragment cueing test varied across experiments some- ditory-to-visual) and does not benefit or benefits onlytimes attaining normal levels and sometimes showing slightly fromprior semantic encoding of target materials

during study compared to nonsemantic encoding (evenimpairments Thus it was unclear whether Warrington



Review187

though explicit recall and recognition benefit greatly Haist et al 1991) or orthographically illegal nonwords(eg ldquoVXRBrdquo Keane et al 1995 Hamann and Squirefrom prior semantic encoding) Priming on word comple-

tion tests is largely but not entirely perceptually based 1997b for a review see Bowers and Schacter 1993)Related studies have sought to determine whether(ie the effect is reduced but not eliminated by modality

changes and is sometimes influenced significantly by amnesic patients can exhibit priming for new associa-

tions between entirely unrelated words For examplesemantic study processing) Conceptual priming incontrast is most clearly observed on the category in- after studying an unrelated word pair (eg windowndash

reason) control subjects are more likely to completestance production task where subjects are given cate-gory cues (eg ldquofruitrdquo) and produce relevant exemplars the word stem ldquorea___rdquo with ldquoreasonrdquo when the stem is

paired wit h ldquowindow rdquo than when it is repaired with an-(eg ldquoorangerdquo ldquopearrdquo) prim ing occ urs when subjec tsare biased to produce previously studied category ex- other unrelated word from the study list (Graf and

Schacter 1985) Severely amnesic patients typicallyemplars Other tasks for investigating conceptual prim-ing include answering general knowledge questions and show impaired priming of new associationson the word

completion test (eg Schacter and Graf 1986 Shima-generating word associations As we will argue later itis likely that most priming tasks have some perceptual mura and Squire 1989) However priming of new asso-

ciations on the word completion test often dependsand conceptual componentsConclusions about preservation of priming in amnesia on the encoding and retrieval of both perceptual and

conceptual information (Graf and Schacter 1989 seeare usually based on studies of perceptual priming Inaddition to previously mentioned findings of normal also Goshen-Gottstein and Moscovitch 1995) Recent

evidence indicates that amnesic patients can showword completion priming in amnesics it has also beenfound that amnesic patients show normal priming on priming of new associations on a word identification

test (which depends largely or entirely on the encodingboth visual and auditory word identification tests (egCermak et al 1985 Schacter et al 1994) as well as and retrieval of perceptual information) when previously

studied word pairs or repaired words are flashed simul-on a picture naming task (Cave and Squire 1992 Ver-faellie et al 1996) Several studies have examined con- taneously and subjects try to identify both words (Gabri-

eli et al 1997) This finding raises the possibility thatceptual priming in amnesic patients and they haveyielded mixed results amnesic patientshaveshown nor- amnesic patients show intact priming of new perceptual

associations but evidence from related t asks is mixedmal priming on the category instance production taskand word association tasks (Shimamura and Squire (cf Moscovitch et al 1986 Musen and Squire 1993

Paller and Mayes 1994) It remains unclear as to whether1984 Schacter 1985 Graf et al 1985 Carlesimo 1994)but showed impaired priming on a conceptual task that impaired priming of new associations in amnesic pa-

tients indicates that amnesics are unable to form theinvolves answering general knowledge questions (Blax-t on 1992 Vaid ya et al 1996) kind s o f no vel rep resent at io ns t hat sup po rt so me t yp es

of implicit retrieval of new associations (Musen andOne problem in evaluating conceptual priming is that

explicit retrieval processes maysometimes playa signif- Squire 1993) or that such priming effects in controlsubjects sometimes depend on explicit retrieval (Bow-icant role in conceptual priming tasks Whereas studies

of perceptual priming with healthy volunteers have pro- ersand Schacter1990Reingold and Goshen-Gottstein1996)duced experimental dissociations between priming and

explicit memorymdash dissociations that effectively rule out Similar issues have arisen in studies examining whetheramnesic patients exhibit what have come to be knownthe possibility that subjects use explicit retrieval on the

priming task (cf Schacter et al 1989 Jacoby 1991)mdash it as form-specific priming effects Studies of healthy vol-unteers have shown that priming effects can be largerhas proven more difficult to obtain such dissocations

with conceptual priming tasks (see Mulligan 1997) when target items are studied and tested in the identicalphysical format than when a perceptual feature of theThus while preserved conceptual priming in amnesic

patients shows that explicit memory need not be in- target item is changed between study and test For ex-ample priming effects on a visual word identificationvolved in such priming effects impaired conceptual

priming in amnesic patients may reflect the fact that test or a visual word completion test are sometimesgreater when words are studied and tested in the sameconceptual priming in control subjects sometimes de-

pends on explicit retrieval processes at least when the typefont or case (upper versus lower) than when type-

font or case is changed between study and test (egtest involved answering general knowledge questions(see Vaidya et al 1996) Roediger and Blaxton 1987 Graf and Ryan 1990 M ar-

solek et al 1992 1996 Wiggs and Martin 1994 CurranSimilar issues have arisen in studies examiningwhether amnesic patients show priming for novel infor- et al 1996 but see Carr et al 1989 and Rajaram and

Roediger 1993for failures to obt ain form-specific prim-mation or new associations In most studies that havereported intact perceptual priming in amnesia target ing) Similarly priming effects on auditory word comple-

tion and identification tests can be greater when wordsinformation has a preexisting memory representation(eg familiar words are represented in memory prior are studied and tested in the same voice than in a differ-

ent voice or even when the fundamental frequency ofto the experiment) Several studies have documentedhowever that amnesics can exhibit normal percep- a single speakerrsquos voice is changed between study and

test (Church and Schacter 1994) Form-speci fic primingtual priming for stimuli that do not have a preexistingmemory representationmdash novel visual patterns (Gabrieli effects have also been observed on tasks involving

identification of or decisions about pictures of visualet al 1990 Musen and Squire 1992) unfamiliar objects(Schacter et al 1993b) pseudowords (eg ldquoBOAFrdquo objects For example in picture identification tasks



Neuron188

greater priming effects are observed when the same B e y o n d A m n e s i a I n s ig h t s f r o m D e m e n t i a s

exemplar (egthe identical picture of a car)is presented a n d F o c a l C o r t i c a l L e s io n s

at study and test compared to two different exemplars Studies of amnesic patients have provided crucial in-(eg pictures of two different carsBiederman and Coo- sights into the relation between priming and the brainper 1991 Srinivas 1993) form-specific visual priming Studiesof other kindsof patient populations have begun

has also been observed when objects are rotated be- to yield a number of new insights that extend and com-tween study and test (see Cooper and Schacter 1992 plement the data from amnesic patients For exampleSrinivas 1995) amnesic patients exhibit both intact perceptual priming

Form-specific priming is potentially important be- and intact motor skill learning so it is unclear whethercause it indicates that priming can be influenced by these two phenomenadepend on a common neuralsub-novel highly specific information about the particular strate or whether they depend on different brain sys-features of items presented during the study episode tems To address this question a number of studiesand because it may alsoreveal important characteristics have examined priming in patients with Huntingtonrsquosof the representations that support priming (for discus- disease who are characterized by damage to the basalsion see Marsolek et al 1992 1996 Tenpenny 1995 gangliaand exhibit severe deficits in motor skill learningBowers 1996 Curran et al 1996) Relevant evidence However patients with Huntingtonrsquos disease show in-from amnesic patients is mixed some studies have tact perceptual priming on tasks such as word comple-failed to find form-specific visual priming (Kinoshita and tion and identification (for a review see Salmon andWayland 1993) or auditory priming (Schacter et al Butters1995) thereby suggesting that perceptual prim-1995a) whereas other studies have delineated c ondi- ing and motor skill learning depend on different braintions in which amnesic patients can show intact form- systemsspecific priming in both visual (Vaidya et al 1998) and The relation between perceptual priming and seman-auditory (Schacter and Church 1995) domains Al- tic processing has been illuminated by evidence fromthough the exact reasons for the different outcomes is patients with focal cortical lesions Schacter et alnot well understood Schacter and Church (1995) and

(1993c) examined perceptual priming in a patient J PVaidya et al (1998) provide discussions of alternative

who had suffered a stroke in the left medialinferior tem-possibilities

poral gyrus that produced a rare condition known asIn summary the evidence indicates that perceptual

word meaning deafness J P had great difficulty com-priming of both familiar and unfamiliar words objects

prehending spoken words even though he could repeatand patterns is generally preserved in amnesic patients

thosewords (indicatingrelatively intact auditorypercep-whereas conceptual priming of familiar information

tion) and could also perceive and understand writtenpriming of new associations and form-specific priming

words Schacter et al found that J P showed normalare sometimes preserved and sometimes imp aired The

perceptual priming on an auditory word identificationfindings on preserved perceptual priming in amnesia

test (in which previously studied and nonstudied wordsprovide support for the view that this type of priming

were masked by white noise) even though he exhibiteddepends on a brain system that is distinct from theseverely impaired auditory comprehension of the targetmedial temporaldiencephalic system that supports ex-words Geva et al (1997) examined priming in a patientplicit memory For example Tulving and Schacter (1990)(N G) with a focal cortical (left temporal) lesion whoproposed that p riming depends on a perceptual repre-is characterized by another relatively rare syndromesentation system (PRS) involved in processing informa-known as prosopanomia mdash a spec ific impairment in thetion about the form and structure but not the meaningretrieval of proper names Though N G had great diffi-and associative properties of words and objects Itculty spontaneously retrieving the names of personshas been proposed that the PRS contains a number ofandcities afterexposure to personand city namesN Gsubsystems dedicated to the representation of visualshowed a normal perceptual priming effect on a word-word formsauditory word formsand shapesof objectsfragment completion test Geva et al suggested thatand that such subsystems depend on posterior corticalpriming in patient NG is supported by intact perceptualregions (Schacter 1994) Because such regions are usu-representations whereas the patientrsquos name retrievalally not damaged in amnesic patients it seems likelydeficit results from a problem in retrieving the phonologythat perceptual priming depends on these subsystemsof the name from a semantic representationWhile proposals of this kind focus on the brain systems

Studies of patientswith focal lesionst o posteriorc orti-that support priming and explicit memory they havecal areas together with studies of patients with Alzhei-been complemented b y cognitive models that focus onmerrsquos disease are helping to clarify the relationship be-the processing activities that are engaged by differenttween perceptual and conceptualp rimingLike amnesictypes of tasks Such models attempt to explain differ-patients Alzheimerrsquos patients are characterized by se-ences between priming and explicit memory in termsvere deficits in explicit recall and recognition whichof the different forms of processing that are requiredare produced by pathology (ie neuritic plaques andby implicit and explicit memory tasks (Blaxton 1989neurofibrillary tangles) in the hippocampal formationRoediger and McDermott 1993) By combining the sys-(for a recent review see Boller and Duyckaerts 1997)tems-based view developed on the b asis of neuropsy-In addit ion however Alzheimerrsquos patients are also char-chological observations with the process-oriented viewacterized by pathology to frontal and temporal neocorti-developed from cognitive analyses it is possible to pro-cal regions together with relative sparing of occipitalvide a fuller theoretical analysis than either view cancortex (Arnold et al 1991) Several studies have shownsupply alone (cf Roediger 1990 Schacter 1992 Gabri-

el i 1995 Rat cl if f and M cKoon 1997) t hat Alzheimerrsquos pat ient s ex hibi t normal perceptual



Review189

priming on the word identification test but show im- that havefocused on conceptual componentsof primingand also consider evidence bearing on the interrelationspaired conceptual priming on the categoryinstance pro-

duction test (Keane et al 1991 1994 Monti et al 1996 between the two components of primingN e u r o i m a g i n g o f P e r c e p t u a l Pr i m i n g cf Maki and Knopman 1996) Alzheimerrsquos patients have

also shown intact perceptual priming of novel visual The first neuroimaging studies of perceptual priming

were reported by Squire and colleagues (Squire et alpatterns on a task that requires completion of partialpatterns (Postle et al 1996) In contrast priming on 1992 Buckneret al1995) Subjects studied words prior

to PET scans during a series of scans several minutesthe word-stem completion test is oft en but not alwaysimpaired in patients with Alzheimerrsquos disease (seeM onti later they were asked to complete the stems to form the

first words that came to mind Behavioral data revealedet al 1996) perhaps reflecting a conceptual componentto stem-completion priming The conceptual priming evidence of priming stems of previously studied words

were completed more often with target items than wereimpairment is likely attributable to pathology in anteriorneocortical regions that are spared in amnesia novel stems and subjectstook less time to produce the

studied words in comparison to completion of novelIn contrast the opposite pattern of findings has beenreported in patients with focal damage to occipital cor- stems PET blood flow data indicated that bilateral pos-

terior perceptual processing areas in extrastriateoccipi-tex For example Gabrieli et al (1995) described a pa-tient M S who had most of his right occipital lobe tal cortex (which were activated during completion of

novel stems) showed reduced activation w hen wordremoved for relief of intractable epilepsy M S fails toshow perceptual priming on a word identification test stems were primed Such an effect may reflect a neural

correlate of perceptual primingafterexposure to a stim-but shows normal conceptual priming on a categoryinstance production test (see alsoCarlesimo et al 1994 ulus subsequent processing is faster and requires less

neural activity (for converging evidence from researchfor similar observations of a related patient) Combinedwith the results from studies of amnesic patients and using event-related potentials [ERPs] see Badgaiyan

and Posner 1997)patients with Alzheimerrsquos disease these findings pro-vide evidence for a double dissociation between per- Interpretation of these early findings wascomplicated

by the additional finding that in the study of Squireceptual and conceptual priming These kinds of resultsalso illustrate how studies of different types of patient et al (1992) priming was associated with a small but

significant blood flow increase in the vicinity of the hip-populations can provide complementary evidence con-cerning the brain regions that are involved in priming pocampal formation (right parahippocampal gyrus) The

hippocampal activation may have reflected the fact thatStudies of amnesic patients provide strong evidencethat medial temporaldiencephalic structures are no t subjects had studied the target words twice under se-

mantic encoding conditions that promote high levelsnecessary for perceptual priming but do not provideevidence concerning which structures ar e involved in of explicit memory Consistent with the idea that the

hippocampal activation was not directly related to per-such priming effects Studies of patients with focal oc-

cipital lesions help to complete the puzzle by showing ceptual priming subsequent research has shown thatwhen explicit memory is greatly reduced by the use of athat regions of occipital cortex do play a role in percep-

tual (but not conceptual)priming Additional converging nonsemantic encoding task that produces poor explicitmemory the priming-related hippocampal activationre-evidence concerning the brain regions that are relevant

to priming has been provided by recent neuroimaging ported by Squire et al is eliminated whereas the prim-ing-related blood flow reduction in extrastriate cortexstudiesis still observed (Schacter et al 1996)Aft er consideringthese and related findings at lengthSchacter and Buck-

Priming and Neuroimaging ner (1998) concluded that the hippocampal activationDuring the past several years there has been a sharp observed by Squire et al during perceptual priming mayincrease in studies using two major neuroimaging tech- reflect an involuntary ldquocontaminationrdquo of priming perfor-niques to investigate memory encoding and retrieval mance by explicit retrieval (for further relevant evidenceprocesses positron emission tomography (PET) which see Rugg et al 1997)measures changes in regional cerebral blood flow and Importantly PET findings of blood flow reductions infunctional magnetic resonance imaging (fMRI) which posterior cortical regions during perceptual priming

measures changes in blood oxygenation level associ- have been extended to severalother priming tasks usingated with changes in blood flow and volume(for reviews fMRI Studies in which priming was assessed with suchsee Ungerleider 1995 Cabeza and Nyberg 1997 Buck- tasks as visual word-stem completion visual word-frag-ner and Koutstaal 1998) Because changes in net neural ment completion generation of verbs to visually pre-activity are associated with changes in these blood sented nouns and classification of objects all show re-properties (Raichle 1987) functional neuroimaging pro- ductions in posterior cortical regions when repeatedvides a unique window through which to view normal items are compared to novel items (Martin et al 1995human brain function With respect to priming PET and Soc Neurosci abstract Blaxton et al 1996 BucknerfMRI allow examination of brain areas that are activated et al 1997 for a review and discussion see Bucknerduring performance of tasks for which priming effects and Koutstaal 1998 Schacter and Buckner 1998)are observed and they also permit investigation of how In a recent study (Buckner et al 1998 [this issue ofactivation levels change when items are primed We will Neuron ]) we examined priming using newly developedfirst consider studies that have focused on perceptual procedures for analyzing event-related fMRI Onelimita-

tion of previous PET and fMRI studies is that items fromcomponents of priming We will then summarize studies



Neuron190

different experimental conditions had to be presented exposure Wagner et al(1997b)and Buckner et al(1998)in separate ldquoblocksrdquo in contrast to the usual practice recently reported similar prefrontal reductions for re-in cognitive studies of intermixing items from different peatedsemantic classificationof object picturesImpor-conditions With respect to priming presenting old and tantly the left prefrontal regions that showed activationnew items in separate blocks could allow subjects to reductions during conceptual priming have also shown

change strategies or allocate attention differently to old activation increases duringsemantic encodingtasks rel-and new items By contrast when using event-related ative to perceptual encoding tasks (eg Kapur et alfMRI procedures different types of stimuli can be inter- 1994 Demb et al 1995 Gabrieli et al 1996 Bucknermixed randomly and responses to the novel and re- and Koutstaal 1998 for relevant evidence from patientspeated items can be separated and analyzed in an with lesions of left inferior frontal cortex see Swick andevent-locked manner similar to how ERPs and single- Knight 1996) Considered along with the previously dis-unit physiologic data are handled (Dale and Buckner cussed findings concerning perceptual priming the1997 Josephs et al 1997 for a review of event-related overall pattern of data raises the possibilit y that activa-fMRI procedures see Rosen et al 1998) tion reductions in posterior (ie extrastriate) cortical

In our event-related fMRI study of priming subjects regionsare specifically associated with perceptual prim-were instructed to classify colored pictures of objects ing whereas activation reductions in left prefrontal re-as tending either to move (eg a car) or to stay still gions are specifically associated with concept ual prim-(eg a tree) Priming was revealed by the finding that ing (see Figure 1) Though such an interpretation issubjects were faster to make decisions about repeated appealing in its simplicity this view is no doubt oversim-objects as compared t o novel objects By rapidly pres-

plistic because it does not take into account importantenting the trials in randomly intermixed format subjects differencesin task parametersbetween studies Toillus-could not have anticipated or adjusted strategies for trate the point consider studies that used word-stemnovel or repeated items to p roduce such an effect the completion to investigate perceptual priming and deci-behavioral facilitation was produced entirelyb y the prior sions about whether words are abstract or concrete tohistory(ienovel versus repeated)of the individual item investigate conceptual priming In prior studies of word-Analysis of fMRI data indicated robust activation reduc- stem completion (Squire et al 1992 Buckner et altions in posterior cortical areas when comparing the 1995 Schacter et al 1996) subjects engaged in anrepeated and novel objects Moreover examination of orienting task during study (eg they made semanticthe anatomic locations of the reductions indicated that or perceptual judgments about the target words) theyhigher-level visual areas were most affected whereas did not engage in a word-stem completion task Thusearly visual areas falling at or near striate cortex were the same items were repeated (at least partially) fromleast affected thereby suggesting that the priming- the study list to the word-stem test but the task itselfrelated decreases show anatomic specificity This finding differed at study and test By contrast in the studies byis consistent with a study that examined the anatomic Raichle et al (1994) and Demb et al (1995) that showed

specificity of visual word priming-related reductions in priming-relatedp refrontal reductions the identical tasksrelation to visual area boundaries defined by retinotopic were performed across item repetitions This observa-organization (Halgren et al 1997) In the study of Hal- tion raises two questions (1) would prefrontal reduc-gren et al robust activation reductions were noted in tions be diminished or eliminated when itemsare initiallylate visual areas with little or no reductions noted in

presented under different task conditions than at testearly visual areas as defined by retinotopy

evenfor the same tasks thathave shown robust prefron-N e u r o i m a g i n g o f C o n c e p t u a l P r im i n g

tal reduction when repeated exactly (eg Demb et alR e la t i o n s t o P e r c e p t u a l P r im i n g

1995) and (2) would tasks such as w ord-stem comple-The foregoing findings are generally consistent with the

tion whicht ypically show priming effects in across-taskidea that priming can be attributable at least in part

conditions show priming-related prefrontal reductionsto facilitation of perceptual processes Several studies

when the exact task is repeated In a series of fMRIhave also examined conceptual priming effects on tasks

studies (Buckner et al 1997 Koutstaal et al 1997that require semantic processing (for direct compari-

Wagner et al 1997a for a discussion see Buckner andsons between conceptual and perceptual priming and

Koutstaal 1998 and Schacter and Buckner 1998) ourrecall tasks see Blaxton et al 1996) More specifically

group has provided evidence bearing on each of t hesesuch studies have documented that robust and consis-questionstent reductions can be detected in higher-order prefron-

In one experiment subjects performed word-stemtal brain regions when words are repeated across verbalcompletion with a repeating set of stem cues duringprocessing tasks requiring semantic elaboration Forparticular fMRI task blocks and with novel word-stemexample Raichle et al (1994) observed that generatingcues during other blocks Contrasting the novel with theverbs that are meaningfully related to presented nounsrepeated blocks fMRI data showed reduced activationcould be facilitated by repeated exposure to the nounsfor repeated word-stems in left dorsal and inferior p re-After many repetitions robust b lood flow reductionsfrontal regions (Buckner et al 1997 Koutstaal et alwere noted in left prefrontal cortex Demb et al (1995)1997) A parallel design was used for a verb generationpresented subjects with words and asked them to clas-task that has been linked previously with prefrontal re-sify each word as abstract or concrete Similar to theductions (Raichle et al 1994) and similar results wereRaichle et al study left prefrontal regions showed re-obtained Thus contrary to the possibility that word-duced activation during performance with repeated

items compared to novel items even after a single item stem completion is associated only with blood flow



Review191

Figure 1 Activation Reductions Associated with Perceptual and Conceptual Priming

Activation reductions have been observed frequently across a number of studies in which item repetition yields priming (eg Squire et al1992 Raichle et al 1994 Demb et al 1995 Buckner et al 1995 1997 1998 Schacter et al 1996 see text for discussion) This phenomenon

is illustrated above in two sets of fMRI images from our laboratory Each set shows one activation image (upper left) where novel items arecompared to a low level reference task revealing all areas activated by the task The second activation image (bottom right) shows the directcomparison between the novel and repeated items revealing those areas in which activation levels are reduced as a consequence of primingLeft prefrontal areas that activate for novel items (A) have in many instances shown reductions in activation (B) perhaps as a reflection ofconceptual priming An area in extrastriate visual cortex that is activated by presentation of novel visual words (C) reduces activation to

repeated visual words (D) Reductions in visual cortex activity may underlie perceptual priming Not all areas activated by a task show priming-related reductions as can be observed in the comparison between (C) and (D) certain areas activated in (C) do not show reductions in (D)

As noted in the text the areas showing reductions are anatomically selective The color scale in the images represents the significance ofthe activation with brighter colors indicating increased significance The schematic of the brain on the left shows the approximate level fromwhich the section is taken

reductions in brain regions that perform perceptual anal- left prefrontal cortex initially so it is difficult to interpretthe lack of activation reduction across trials Wagneryses the overall patterns of blood flow reductions indi-

cate that word-stem completion (like verb generation or et al (1997a) used a paradigm in which words werepresented initially during the perceptual decision taskword categorization) is neither an entirely perceptually

driven task nor an entirely conceptually driven task and were tested again on the conceptual (abstractcon-crete) decision t ask This manipulation revealed that (1)Rather item and task overlap determines which kinds

of processes will be engaged at study and test and left dorsal prefrontal cortex was activated by the ab-stractconcrete decision task and (2) the activation wastherefore which will manifest priming-related modula-

tions Note that findings of left prefrontal reductions not reduced as a consequence of prior exposure to theitems in the perceptual decision task By c ontrast in aduring word- stem completion performance converge

with recent neuropsychological evidence from studies second condition using the abstractconcrete decisiontask where the exact items and t ask demands wereof elderlyadultsshowing a correlationbetween the mag-

nitude of stem-completion priming and performance on repeated over several trials across-trial reductions inleft prefrontal activity were noted as initially reportedneuropsychological t asks sensitive to frontal lobe dys-

function (Winocur et al 1996) by Demb et al (1995)Considered together all of the foregoing findings con-Evidence bearing on this latter point has been pro-

vided in studies by Wagner et al (1997a) based on verge to suggest one neural correlate of priming re-peating items during performance of t he same task orprevious work by Demb et al(1995) in which items were

repeated across two tasks one involved a perceptual even during performance of different tasks can lead todecreases in the amount of activation present in specificdecision and the other involved a conceptual decision

Demb et al (1995) reported that when items were re- brain areas The reductions appear to be selective de-pending on item and task overlap across repetitionspeated across the perceptual decision task there were

no activation reductions in left prefrontal cortex How- they are also characterized by anatomic specificity withhigher-level visual areas showing greater reductionsever the perceptual decision task likely did not activate



Neuron192

than earlier visual areas (Figure 1) Overall the neuro- absence of a well-developed animal model of primingHowever recent studies using single-unit recordingimaging data converge with the previously reported re-

sults of studies with patient populations the regions techniques in nonhuman primates have begun to fill thegap and to illuminate possible mechanisms of primingthat consistently show priming-related blood flow re-

ductions are not usually damaged in amnesic patients For example in studies by Desimone and Miller and

their colleagues (summarized by Desimone et al 1995and one region that has consistently shown blood flowreductions during perceptual primingmdash extrastriate vi- see also Ringo 1996) animals fixated on a series of

complex visual objects such as patterns or faces whilesual cortexmdash is selectively damaged in a patient whoshows impaired perceptual priming (Gabrieli et al the activityof populations of cells in inferior temporal (IT)

cortex was recorded Repeated exposure to the same1995) To link the results from neuropsychological popu-lations and neuroimaging more tightly it will be neces- stimulus yielded reducedresponses in a substantial pro-

portion of IT cells the reduced responding persistedsaryt o useneuroimaging techniquesto examineprimingin relevant patient populations Although little data are across as many as 150 intervening items The effect

could not be attributed to generalized habituation be-yet available Buckner and Koutstaal (1998) provide pre-liminary fMRI data indicating that conceptual priming in cause introduction of novel stimuli produced a return

to initial levels of activity These findings bear a strik-amnesia is accompanied by left prefrontal activationreductions as previously observed in normal subjects ing resemblance to the previously discussed priming-

related activation reductions observed in neuroimaging(Demb et al 1995)One open question is how functional changes re- studies with humans (see Desimone 1996) Wiggs and

Martin (1998) have pointed out a number of additionalvealed after a singlerepetitionof anitemmdash the procedureused most often in studies of primingmdash will relate to similarities between phenomena of ldquorepetition suppres-

sionrdquo (Desimone et al 1995) observed in animals andparadigms in which items are repeated numerous timesOn the one hand both kinds of procedures have re- priming phenomena Although we cannot be certain that

repetition suppression effects are based on the samevealed similar activation reductions in specific brainareas For example the previously mentioned studies processes that underlie perceptual priming in humans

the similarities between the two types of phenomenaby Raichle et al (1994) and Demb et al (1995) revealqualitatively similar reductions in prefrontal areas even are suggestive enough to merit careful study in future

investigations The development of an animal model ofthough Raichle et al examined task performance afternumerous exposures whereas Demb et al character- priming could help to laythe foundations for an eventual

understanding of priming at cellular and even molecularized changesafter a singlerepetitionOn the other handextensive repetition has been associated with changes levels of analysis Such a development would greatly

extend the systems-level analyses that have proven sobeyond those described here in relation to priming Forexample Raichle et al (1994) noted increases in activa- useful in the cognitive neuroscience investigations that

have helped to uncover and illuminate the phenomenontion in bilateral insular cortex following extensive repeti-

tion They hypothesized that with extensive exposure of primingto items alternative functional anatomy is utilizedmdash

Acknowledgmentsperhaps reflecting the transition from a brain pathwayunderlying a more effortful mode of task performance

This work was supported by National Institute on Aging AG08441with novel items to one driven by automatic response

and the Human Frontiers Science Foundation We thank Alex Martinselection for items that have been repeated extensively and Eric Kandel for comments and discussion and Carolyn BrennerAlthough priming-related increases in activation have for help with preparation of the manuscriptoccasionally been observed in single repetition experi-

Referencesments (eg Schacter et al 1995b) priming-related de-creases are far more common in such studies Thus

Arnold SE Hyman BT Flory J Damasio AR and Van Hoesenfurther research is needed to clarify how functional

GW (1991) The topographical and neuroanatomical distribution ofchanges observed during the early stages of item repeti- neurofibrillary tangles and neuritic plaques in the cerebral cortex oftion which we tend to describe in terms of priming patients with Alzheimerrsquos disease Cerebral Cortex 1 103ndash116are related to later stages such as those observed by Badgaiyan RD and Posner MI (1997) Time course of corticalRaichle et al which may be more akin to what has activations in implicit and explicit recall J Neurosci17 4904ndash4913

traditionally b een termed skill learning (see Ungerleider Biederman I and Cooper EE (1991) Priming contour deletedimages evidence for intermediate representations in visual object1995 for initial discussion of this issue)recognition Cogn Psychol 23 393ndash419

Blaxton TA (1989) Investigating dissoc iations among memorymeasures supportfor a transfer-appropriate processingframeworkConcluding CommentsJ Exp Psychol [Learn Mem Cogn] 1 5 657ndash668Progress in understanding the relation between primingBlaxton TA (1992) Dissociations among memory measures inand the brainhas been facilitated greatly by the interplaymemory-impaired subjects evidence for a processing account ofbetween cognitive studies of individuals with intactmemory Mem Cogn 20 549ndash562

memory functionsneuropsychological investigations ofBlaxton TA Bookheimer SY Zeffiro TA Figlozzi CMGaillard

patients with memory disorders and functional neuro-WD and Theodore WH (1996) Functional mapping of human

imaging research It seems likely that the interplay memory using PET comparisons of conceptual and perceptualamong these different approaches will continue to tasks Can J Exp Psychol 5 0 42ndash56

deepen our understanding of priming Boller F and Duyckaerts C (1997) Alzheimer diseaseclinical and

anatomic aspects In Behavioral Neurology and NeuropsychologyOne notable gap in current approaches concerns the



Review193

TE Feinberg and MJ Farah eds (New York McGraw-Hill) pp Desimone R Miller EK Chelazzi L and Lueschow A (1995)Multiple memory systems in the visual cortex In The Cognitive521ndash544Neurosciences MS Gazzaniga ed (Cambridge MA MIT Press)Bowers JS (1996) Different perceptual codes support priming forpp 475ndash486words and pseudowords was Morton right all along J Exp Psy-

chol [Learn Mem Cogn] 2 2 1336ndash1353 Gab rieli JDE (1995) A syst em at ic view o f hum an mem ory p ro-cesses J Int Neuropsychol Soc 1 115ndash118Bowers JS and Schacter DL (1990) Implicit memory and test

awareness J Exp Psychol [Learn Mem Cogn] 1 6 404ndash416 Gabrieli JDE Milberg W Keane MM and Corkin S (1990)Intact priming of patterns despite impaired memory Neuropsycho-Bowers JS and Schacter DL (1993) Priming of novel informationlogia 2 8 417ndash428in amnesic patients issues and dataIn Implicit Memory New Direc-

tions in Cognition Development and Neuropsychology P Graf and Gabrieli JDE Fleischman DAKeane MM Reminger SL andMorrell F (1995) Double dissociation between memory systemsMEJ Masson eds (New York Academic Press) p p 303ndash326underlying explicit and implicit memory int he human brain PsycholBrown AS and Mitchell DB (1994) A reevaluation of semanticSci 6 76ndash82versus nonsemantic processing in implicit memory Mem Cogn22

533ndash541 Gabrieli JDE Desmond JE Demb JB Wagner AD StoneMV Vaidya CJ and Glover GH (1996) Functional magneticBuckner RL and Koutstaal W (1998) Functional neuroimagingresonance imaging of semantic memory processes in the frontalstudies of encoding priming and explicit memory retrieval Proclobes Psychol Sci 7 278ndash283Natl Acad Sci USA in pressGabrieli JDE Keane MM Zarella MM and Poldrack RABuckner RL Petersen SE Ojemann JG Miezin FM Squire(1997) Preservation of implicit memory for new associations inLR and Raichle ME (1995) Functional anatomical studies ofglobal amnesia Psychol Sci 8 326ndash329explicit and implicit memory retrieval tasks J Neurosci 15 12ndash29Geva A Moscovitch M and Leach L (1997) Perceptual primingBuckner RL Koutstaal WSchacter DL PetersenSE Raichleof proper names in young and older normal adults and a patientME and Rosen BR (1997) FMRI studies of item repetition during

with prosopanomia Neuropsychology 1 1 232ndash242word generation Cogn Neurosci Annu Meet 4 67Goshen-Gottstein Y and Moscovitc h M (1995) Repetition primingBuckner R Goodman J Burock M Rotte M Koutstaal Wfor newly-formed associations perceptual and conceptual influ-Schacter DL Rosen B and Dale A (1998) Functional-anatomicences J Exp Psychol [Learn Mem Cogn] 2 1 1229ndash1248correlates of object priming in humans revealed by rapid presenta-

tion event-related fMRI Neuron 2 0 this issue 285ndash296 Graf P and Schacter DL (1985) Implicit and explicit memory fornew associations in normal subjects and amnesic patients J ExpCabeza R and Nyberg L (1997) Imaging cognition an empiricalPsychol [Learn Mem Cogn] 1 1 501ndash518review of PET studies with normal subjects J Cogn Neurosci 9

1ndash26 Graf P and Schacter DL (1989) Unitization and grouping mediatedissociations in memory for new associations J Exp PsycholCarlesimo GA(1994) Perceptual and concept ual priming in amne-[Learn Mem Cogn] 1 5 930ndash940sic and alcoholic patients Neuropsychologia 3 2 903ndash921Graf P and Ryan L (1990) Transfer-appropriate processing forCarlesimoGA FaddaL Sabbadini M andCaltagirone C(1994)implicit and explicit memory J Exp Psychol [Learn Mem Cogn]Visual repetition priming for words relies on access to the visual16 978ndash992input lexicon evidence from a dyslexic patient Neuropsychologia

32 1089ndash1100 Graf P Squire LR and Mandler G (1984) The information thatamnesic patients do not forget J ExpPsychol [Learn Mem Cogn]CarrTH Brown JSand Charalambous A (1989) Repetition and10 164ndash178reading perceptual encoding mechanisms are very abstract but not

very interactive J Exp Psychol [Learn Mem Cogn] 15 763ndash778 Graf P Shimamura AP and Squire LR (1985) Priming acrossmodalities and priming across category levels extending the do-Cave CB (1997) Very long-lasting priming in pic ture naming Psy-main of preserved functioning in amnesia J Exp Psychol [Learnchol Sci 8 322ndash325Mem Cogn] 1 1 385ndash395

Cave CB and Squire LR (1992)Intact and long-lasting repetitionHaist F Musen G and Squire LR (1991) Intact priming of wordspriming in amnesia J Exp Psychol [Learn Mem Cogn] 18 and nonwords in amnesia Psychobiology 1 9 275ndash285509ndash520Halgren E Buckner RL Marinkovic K Rosen BR and DaleCermak LS Talbot N Chandler K and Wolbarst LR (1985)AM (1997) Cortical localization of word repetition effects CognThe perceptual priming phenomenon in amnesia NeuropsychologiaNeurosci Soc Annu Meet 4 3423 615ndash622Hamann SBand Squire LR (1997a)Intact perceptual memory inChiu CYP and Schacter DL (1995)Auditory p riming for nonver-the absence of conscious memory Behav Neurosci 11 1 850ndash854bal information implicit and explicit memory for environmental

sounds Conscious Cogn 4 440ndash458 Hamann SB and Squire LR (1997b) Intact priming for novelperceptual representations in amnesia J Cogn Neurosci 9 Church BA and Schacter DL (1994) Perceptual specificity of699ndash713auditory priming implicit memory for voice intonation and funda-

mental frequency J ExpPsychol [Learn MemCogn] 20 521ndash533 Hamann SB Squire LR and Schacter DL (1995) Perceptualthresholds and priming in amnesia Neuropsychology 9 1ndash13Cooper LA and Schacter DL (1992) Dissociations between

structural and episodic representations of visual objects Curr Di- Jacoby LL (1991) A process dissociation framework separatingrect Psychol Sci 1 141ndash146 automatic from intentional uses of memory J Mem Lang 30 513ndash541Craik FIM and Tulving E (1975) Depth of processing and the

retention of words in episodic memory J Exp Psychol [Gen] 1 04 Jacoby LL and Dallas M (1981) On the relationship between268ndash294 autobiographical memory and perceptual learning J Exp Psychol

11 0 306ndash340CurranT Schacter DLand Bessenoff G (1996) Visualspecif icityeffects on word stem completion Beyond transfer appropriate pro- Josephs O Turner R and Friston K (1997) Event-related fMRIcessing Can J Exp Psychol 5 0 22ndash33 Hum Brain Map 5 243ndash248

Dale AM and Buckner RL (1997) Selective averaging of rapidly Kapur S Craik FIM Tulving E Wilson AA Houle S andpresented individual trials using fM RI Hum Brain Map 5 329ndash340 Brown GM (1994) Neuroanatomical correlates of encoding in epi-

sodic memory levels of processing effect Proc Natl Acad SciDemb JB Desmond JE Wagner ADVaidyaCJ GlouerGHUSA 91 2008ndash2011and Gabrieli JDE (1995) Semantic encoding and retrieval in the

left inferior prefrontal cortex a functional MRI study of task difficulty Keane MM Gabrieli JDE Fennema AC Growdon JH andand process specificity J Neurosci 15 5870ndash5878 Corkin S (1991) Evidence for a dissociation between perceptual

and conceptual priming in Alzheimerrsquos disease Behav NeurosciDesimone R (1996) Neural mechanisms for visual memory andtheir role in attention Proc Natl Acad Sci USA 9 3 13494ndash13499 105 326ndash342



Neuron194

Keane MM Gabrieli JDE Growdon JH and Corkin S (1994) memory for novel patterns in Alzheimerrsquos disease Learn Mem 3 305ndash312Priming in perceptual identification of pseudowords is normal in

Alzheimerrsquos disease Neuropsychologia 3 2 343ndash356 Raichle ME (1987) Circulatory and metabolic correlates of brainfunction in normal humans In The Handbook of Physiology theKeane MM Gabrieli JDE Noland JS and M cNealySI (1995)

Normal perceptual priming of orthographically illegal nonwords in Nervous System F Plum and V Mountcastle eds (Bethesda MDAmerican Physiological Association) pp 643ndash674amnesia J Int Neuropsychol Soc 5 425ndash433

Kinoshita S and Wayland SV (1993) Effects of surface features Raichle ME Fiez JA Videen TO MacLeod AM Pardo JVFox PT and Petersen SE (1994) Pract ice-related c hanges inon word-fragment completion in amnesic subjects Am J Psychol

106 67ndash80 human brain functional anatomy during nonmotor learning CerebralCortex 4 8ndash26Koutstaal W Buckner RL Schacter DL and Rosen BR (1997)

An fMRIstudy ofitem repetition during auditorily cued word genera- Rajaram S and Roediger HL (1993) Direct c omparison of four

implicit memory tests J Exp Psychol [Learn Mem Cogn] 19 tion Cogn Neurosci Soc Annu Meet 4 68765ndash776La Voie D and Light LL (1994) Adult age diff erences in repetition

priming a meta-analysis Psychol Aging 9 539ndash553 Rat cli ff R and McKoon G (1997) A count er model f or i mpli ci t

priming in perceptual word identification Psychol Rev 104 Maki PM and Knopman DS (1996) Limitations of the dist inction319ndash343between conceptual and perceptual implicit memory a study of

Alzheimerrsquos disease Neuropsychology 1 0 464ndash474 Rei ngol d E M and Goshen-Got tstei n Y (1996) Separat ing con-sciously controlled and automatic influences in memory for newMarsolek CJ Kosslyn SM and Squire LR (1992)Form specificassociations J Exp Psychol [Learn Mem Cogn] 2 2 397ndash406visual priming in the right cerebral hemisphere J Exp Psychol

[Learn Mem Cogn] 1 8 492ndash508 Ringo JL (1996) St imulus sp ec ific ad ap tat ion in inferior t emporaland medial temporal cortex of the monkey Behav Brain Res 4 0 Marsolek CJ Schacter DL and Nicholas CD (1996) Form-191ndash197specific visual priming for new associations in the right cerebral

hemisphere Mem Cogn 2 4 539ndash556 Roed ig er H L I II (1990) Im plic it m emo ry r et ent ion w it hout r emem-bering Am Psychol 45 1043ndash1056Milner B Corkin S and Teuber HL (1968) Further analysis of

the hippocampal amnesic syndrome fourteen year follow-up study Roediger HL III and McDermott KB (1993) Implicit memoryin normal human subjects In Handbook of Neuropsychology Hof HM Neuropsychologia 6 215ndash234

Spinnler and F Boller eds (Amsterdam Elsevier) pp 63ndash131Mineka S and Nugent K (1995) Mood -congruent memory biasesin anxiety and depressionIn MemoryDistortion How MindsBrains RoedigerHL IIIand Blaxton TA(1987) Retrieval modesproduce

dissociations in memory for surface information In The Ebbinghausand Societies Reconstruct the Past DL Schacter JT Coyle GD

Fischbach M-M Mesulamand LESullivaneds (CambridgeMA Centennial Conference DS Gorfein and RR Hoffman eds (Hills-dale NJ Lawrence Erlbaum Associates) pp 349ndash379Harvard University Press) pp 173ndash196

Mitchell DB (1993)Implicit and explicit memoryfor picturesmulti- Rosen BR Buckner RL and Dale AM (1998) Event related

fMRI Past present and future Proc Natl Acad Sci USA in pressple views across the lifespan In Implicit Memory New Directions inCognition Development and Neuropsychology P Graf and MEJ Rugg MD Fletcher PC Frith CD Frackowiak RSJ and Do-Masson eds (Hillsdale NJ Lawrence Erlbaum Associates) pp lan RJ (1997) Brain regions supporting intentional and incid ental171ndash190 memory a PET study Neuroreport 8 1283ndash1287Monti LA Gabrieli JDE Reminger SL Rinaldi JA Wilson Salmon DP and Butters N (1995) Neurobiology of skill and habitRS and Fleischman DA (1996) Differential effects of aging and learning Curr Opin Neurobiol 5 184ndash190

Alzheimerrsquos disease upon conceptual implicit and explicit memory Scarborough DL Cortese C and Scarborough HS (1977) Fre-Neuropsychology 1 0 101ndash112

quency and repetition effects in lexical memory J Exp PsycholMoscovitc h M Wonocur G and McLachlin D (1986) Memory as [Hum Percep Perf] 3 1ndash17assessed by recognition and reading time in normal and memory

Schacter DL (1985) Priming of old and new knowledge in amnesicimpaired people with Alzheimerrsquos disease and other neurological

patients and normal subjects Ann NY Acad Sci 4 44 44ndash53disorders J Exp Psychol 11 5 331ndash346

Schacter DL (1987) Implicit memory history and current statusMulligan NW (1997) Attention and implicit memory tests the ef-

J Exp Psychol [Learn Mem Cogn] 1 3 501ndash518fects of varying attentional load on conceptual priming Mem Cogn

Schacter DL (1992) Understanding implicit memory a cognitive25 11ndash17neuroscience approach Am Psychol 47 559ndash569

Musen G and Squire LR (1992) Nonverbal priming in amnesiaSchacter DL (1994) Priming and multiple memory systems per-Mem Cogn 20 441ndash448ceptual mechanisms of implicit memory In Memory Systems 1994Musen G and Squire LR (1993) On the implicit learning of newDL Schacter and E Tulving eds (Cambridge MA MIT Press) pp associations by amnesic patients and normal subjects Neuropsy-244ndash256chology 7 119ndash135Schacter DL and Graf P (1986) Preserved learning in amnesicNaito M and Komatsu S-I (1993) Processes involved in child-patients perspectives on research from direct priming J Clin Exphood development of implicit memory In Implicit Memory NewNeuropsychol 8 727ndash743

Directions in Cognition Development and Neuropsychology P Schact er DL and Tulving E (1994) Memor y Systems 1994 (Cam-Graf and MEJ Masson eds (Hillsdale NJ Lawrence Erlbaumbridge MA MIT Press)Associates) pp 231ndash260

Schacter DL and Church B (1995) Implicit memory in amnesicOstergaard AL and Jernigan TL (1993) Are word priming andpatients when is auditory priming spared J Int Neuropsycholexplicit memory mediated by different brain structures In ImplicitSoc 1 434ndash442Memory New Directions inCognition Development and Neuropsy-

chology P Graf and MEJ Masson eds (Hillsdale NJ Lawrence Schacter DL and Buckner RL (1998) On the relations amongErlbaum Associates) pp 327ndash349 priming conscious recollection and intentional retrievalNeurobiol

Learn Mem in pressPaller KA and Mayes AM (1994) New-association priming ofword identification in normal and amnesic subjects Cortex 30 Schacter DL Bowers J and Booker J (1989) Intention aware-53ndash73 ness and implicit memory the retrieval intentionality criterion In

Implicit Memory Theoretical Issues S Lewandowsky JC DunnParkin A (1993) Implicit memory across the lifespan In Implicitand K Kirsner eds (Hillsdale NJ Lawrence Erlbaum Associates)Memory New Directions inCognition Development and Neuropsy-

chology P Graf and MEJ Masson eds (Hillsdale NJ Lawrence pp 47ndash69Erlbaum Associates) pp 191ndash206 Schacter DL Chiu CYP and Ochsner KN (1993a) Imp licit

memory a selective review Annu Rev Neurosci 16 159ndash182Postle BR Corkin S and Growdon JH (1996) Intact implicit



Review195

Schacter DL Cooper LA and Treadwell J (1993b) Preserved Warrington EK and Weiskrantz L (1974)The effect of prior learn-ing on subsequent retention in amnesic patients Neuropsychologiapriming of novel objects across size transformation in amnesic pa-

tients Psychol Sci 4 331ndash335 12 419ndash428

Wiggs C and Mart in A (1994) Aging and feature specific primingSchacter DL McGlynn SM Milberg WP and Church BA(1993c) Spared priming despite impaired comprehension implicit of familiar and novel stimuli Psychol Aging 9 578ndash588

memory in a case of word meaning deafness Neuropsychology 7 Wiggs CL and Martin A (1998) Properties and mechanisms of107ndash118 visual priming Curr Opin Neurobiol in pressSchacter DL Church B and Treadwell J (1994) Implicit memory Willingham DB (1997) Systems of memory in the human brainin amnesic patients evidence for spared auditory priming Psychol Neuron 1 8 5ndash8Sci 5 20ndash25

Winocur G Moscovitch M and Stuss DT (1996) Explicit andSchacter DL Church B and Bolton E (1995a) Implicit memory implicit memory in the elderly evidence for double dissociationin amnesic patients impairment of voice-specific priming Psychol involving medial temporal- and frontal-lobe functions Neuropsy-Sci 6 20ndash25 chology 1 0 57ndash65

Schacter DL Reiman E Uecker A Polster MR Yun LSand Cooper LA (1995b) Brain regions associated wit h retrieval ofstructurally coherent visual information Nature 37 6 587ndash590

Schacter DL Alpert NM Savage CR Rauch SL and Albert

MS (1996) Conscious recollection and the human hippocampalformation evidence from positron emission tomography Proc Natl

Acad Sci USA 9 3 321ndash325

Shimamura AP and Squire LR (1984) Paired-associate learningand priming effects in amnesia a neuropsychological approach J

Exp Psychol 1 13 556ndash570

Shimamura AP and Squire LR (1989) Impaired priming of newassociations in amnesia J Exp Psychol [Learn Mem Cogn] 1 5 721ndash728

SquireLR (1992) Memory and the hippoc ampus a synthesis fromfindingswith ratsmonkeysand humansPsychol Rev99 195ndash231

Squire LR Ojemann JG Miezin FM Petersen SE VideenTO and Raichle ME (1992) Activation of the hippocampus innormal humans a functional anatomical study of memory ProcNatl Acad Sci USA 8 9 1837ndash1841

Srinivas K (1993) Perceptual specificity in nonverbal priming JExp Psychol [Learn Mem Cogn] 1 9 582ndash602

Srinivas K (1995) Representation of rotated objects in explicit and

implicit memory J Exp Psychol 21 1019ndash1036

Swick D and Knight RT (1996) Is prefrontal cortex involved incued recall A neuropsychological test of PET findings Neuropsy-

chologia 34 1019ndash1028

Tenpenny PL (1995) Summaryionist versus episodic theories ofrepetition priming and word identification Psychol Bull Rev 2 339ndash363

Tulving E (1972) Episodic and semantic memory In Organizationof Memory ETulving and W Donaldson eds (New York AcademicPress) pp 381ndash403

Tulving E and Schacter DL (1990) Priming and human memorysystems Science 2 47 301ndash306

Tulving E Schacter DL and Stark H (1982) Priming effects inword-f ragment completion are independent of recognition memoryJ Exp Psychol [Learn Mem Cogn] 8 336ndash342

Ungerleider LG (1995)Functional brain imaging studies of cortical

mechanisms for memory Science 27 0 760ndash775

Vaidya CJ Gabrieli JDE Demb JB Keane M M and WetzelLC (1996) Impaired priming on the g eneral knowledge task in am-

nesia Neuropsychology 1 0 529ndash537

Vaidya CJ Gabrieli JDE Verfaellie M Fleischman DA andAskari N (1998) Font-specific priming following global amnesia

and occipital-lobe damage Neuropsychology in press

Verfaellie M Gabrieli JDE Vaidya CJ Croce P and RemingerSL (1996) Implicit memory for pict ures in amnesia role of etiology

and priming task Neuropsychology 1 0 517ndash528

Wagner AD Buckner RL Koutstaal W Schacter DL GabrieliJDE and Rosen BR (1997a) An fMRIstudy of within- and across-task item repetition during semantic classification Cog NeurosciSoc Annu Meet 4 68

WagnerAD Desmond JE Demb JB GloverGH and GabrieliJDE (1997b) Semantic repetition priming for verbal and pictorialknowledge a functional MRI study of left inferior prefrontal cortexJ Cogn Neurosci 9 714ndash726



Neuron186

than age-related decreases in explicit recall and recog- and Weiskrantz had uncovered a form of memory that isfully preserved in amnesic patients or whether fragmentnition (see Mitchell 1993 La Voie and Light 1994)Like-

wise studies of priming in young children have shown cueing performance like other kinds of memory is im-paired by amnesiaa relative invariance of priming in younger and older

children (eg from 3ndash10 years of age) even though Studies conducted during the early 1980rsquos clarified

these findings by demonstrating that performance onexplicit memory shows strong d evelopmental t rendsacross the same time period (see Parkin 1993 Naito the word-stem completion task depends critically on

the instructions given to subjects (these instructionsand Komatsu 1993) Studies of patients with variouspsychopathological conditions including depression were not systematically examined in the Warrington and

Weiskrantz studies) When instructed to complete aand anxiety have also revealed robust p riming effects(see M ineka and Nugent 1995) st em w it h t he first word t hat comes t o mind amnesic

patients showed a normal priming effect compared toCognitive studies have been essential in delineatingthe basic properties of priming but to obtain a fuller control subjects but when instructed to produce the

correct item from the study list amnesic patients showedunderstanding of the phenomenon it is necessary tounderstand its neural underpinnings Here we review less recall than did controls (Graf et al 1984) Similar

patterns were observed when amnesic patients wereevidence concerning thebrain systems that are involvedin primingfocusing on studiesof patients with neurolog- given related kindsof priming and recall tasks (Schacter

1985) Normal priming in amnesic patients has sinceical damage and on experiments that have used brainimaging techniquesmdash positron emissiontomography(PET) been observed across a variety of different tasks and

conditions (for reviews see Schacter et al 1993aand functional magnetic resonance imaging (fMRI)mdash toexplore the neural correlatesof priming in healthy volun- Squire 1992)

Overall evidence gathered during the p ast 15 yearsteers We believe that characterizing priming at the levelof brain systems provides an important foundation that supports a distinction between explicit memory which

is impaired in amnesic patients and depends on thecan inform and guide further explorations of priming atc ellular and molec ular levels med ial t emp oral diencep halic regions t hat are d amaged

in amnesia and priming which is preserved in amnesiaand does not depend on medial temporaldiencephalic

Neuropsychology of Priming Insights from structures Although the possibility has been raised thatStudies of Brain-Injured Patients word priming effects on word completion and identifica-P r i m i n g a n d t h e A m n e s i c S y n d r o m e tion tests are not preserved in amnesic patients ap-The amnesic syndrome occurs as a consequence of pearing to be normal because of methodological prob-damage to structures in the medial temporal lobe and lems and limitations of relevant studies (Ostergaard anddiencephalon and is characterized by a severe impair- Jernigan 1993) experiments that have addressed pos-ment in the explicit recollection of everyday facts and sible methodolog ical problems still reveal normal prim-

events together with preservation of general intelli- ing in amnesia (Hamann et al 1995) Indeed recentgence (Squire 1992) The syndrome w as first b rought evidence indicates that priming of word completion andto the widespread attention of neuroscientists during word identification performance can be fully preservedthe 1950rsquos when Milner and her colleagues described even when a patient is unable to attain above-chancethe famous patient H M who developed profound am- scores on forced-choice tests of explicit recognitionnesia for recent events following bilateral medial tempo- memory (Hamann and Squire 1997a) This kind of find-ral lobe resection for the relief of intractable epilepsy ing helps to ruleout the possibility that priming in amne-(eg Milner et al 1968) Although it initially appeared sic patients is mediated by residual explicit memorythat H M had no ability to retain new information of abilities and instead supports the conclusion that prim-any kind Milner et al (1968) reported that H M could ing is mediated by processes or systems that are notlearn new motor skills in a relatively normal manner essential for explicit memory and are spared in amnesia

Early interpretations attributed the findings of Milner V ari et i es of P ri mi ng Dat a and Theoret i cal I mpl i cat i ons and c olleagues to the uniqueness of motor memory In considering the generality of the conclusiont hat prim-However evidence obtained from other amnesic pa- ing is preserved in amnesic patients it is important totients soon raised the possibilit y that preserved memory note that priming can be fractionated into several differ-

function in amnesics might extend beyond the domain ent subtypes One basic distinct ion is between p e r c e p - of motor skill acquisitionIn aseriesof important studies tu a lp r im in g and conceptual pr iming Perceptual primingWarrington and Weiskrantz (1974)reported that amnesic is modality specific and does not depend on semanticpatients sometimes show relatively preserved memory or elaborative encoding of an item at the time of studyperformance when they are given word beginnings as whereas conceptualp riming is not modality specific andcues for previously studied items This was a striking benefits from semantic encoding (Blaxton 1989) Prim-finding because patients demonstrated little explicit ing on a word identification test (where subjects tryrecollection of the same items when they were shown to identify briefly flashed items) represents a relativelyan entire word and were required to make an oldnew ldquopurerdquo form of perceptual priming the priming effect is

judg ment about it However in Warringt on and Weisk- eliminated by modality change between the study andrantzrsquos work the amnesic patientsrsquo performance on the test phases of the experiment (visual-to-aud itory or au-fragment cueing test varied across experiments some- ditory-to-visual) and does not benefit or benefits onlytimes attaining normal levels and sometimes showing slightly fromprior semantic encoding of target materials

during study compared to nonsemantic encoding (evenimpairments Thus it was unclear whether Warrington



Review187




































Neuron188





















Review189



























Neuron190





















Review191




















Neuron192



































Review193























Neuron194




































Review195




































Review187




































Neuron188





















Review189



























Neuron190





















Review191




















Neuron192



































Review193























Neuron194




































Review195




































Neuron188





















Review189



























Neuron190





















Review191




















Neuron192



































Review193























Neuron194




































Review195




































Review189



























Neuron190





















Review191




















Neuron192



































Review193























Neuron194




































Review195




































Neuron190





















Review191




















Neuron192



































Review193























Neuron194




































Review195




































Review191




















Neuron192



































Review193























Neuron194




































Review195




































Neuron192



































Review193























Neuron194




































Review195




































Review193























Neuron194




































Review195




































Neuron194




































Review195




































Review195


































Priming and the Brain. Schacter, 1988

Documents

Transcript of Priming and the Brain. Schacter, 1988