copyright 1993 by the american psychological...

Neuropsychology1993, Vol. 7, No. 2, 107-118

Copyright 1993 by the American Psychological Association, Inc.0894-4105/93/S3.00

Spared Priming Despite Impaired Comprehension:Implicit Memory in a Case of Word-Meaning Deafness

Daniel L. Schacter, Susan M. McGlynn, William P. Milberg, and Barbara A. Church

Previous research has established that direct priming effects on implicit memory testscan be dissociated from explicit remembering. Evidence from studies of collegestudents suggests that priming on various implicit tests can be characterized as apresemantic phenomenon: Priming can occur at full strength following nonsemanticencoding tasks that typically produce low levels of explicit memory. To provide astrong test of this idea, we examined auditory priming in a case of word-meaningdeafness. Despite the patient's impaired auditory comprehension abilities, he showednormal priming on auditory identification tests after study-list exposure to spokenwords. Results support the idea that priming depends on perceptual systems thatoperate at a presemantic level.

Explicit memory refers to conscious or intentionalrecollection of previous experiences, whereas implicitmemory refers to changes in task performance that areattributable to previous experiences but do not involveconscious recollection of them (Graf & Schacter,1985; Schacter, 1987). Considerable research hasshown that implicit memory, as indexed by direct orrepetition priming effects on such tests as word iden-tification, word stem and word fragment completion,and lexical and object decision, can be dissociatedexperimentally from explicit memory in normal sub-jects and amnesic patients (for reviews, see Richard-son-Klavehn & Bjork, 1988; Roediger, 1990;Schacter, 1987; Schacter, Chiu, & Ochsner, in press;Shimamura, 1986).

A striking feature of priming on various implicit

Daniel L. Schacter and Barbara A. Church, Department ofPsychology, Harvard University; Susan M. McGlynn (nowat Maclean Hospital, Belmont, Massachusetts) and WilliamP. Milberg, Geriatric Research Education and Clinical Cen-ter, Department of Veterans Affairs Medical Center, WestRoxbury, Massachusetts.

This research was supported by National Institute on Ag-ing Grant RO1 AGO 8441 to Daniel L. Schacter and by theDepartment of Veterans Affairs Merit Review 097^4-3765-001 to William P. Milberg at the West Roxbury De-partment of Veterans Affairs Medical Center.

We thank Alexander Aminoff, Lynn Leseur, and DanaOsowiecki for help with various aspects of the work.

Correspondence concerning this article should be ad-dressed to Daniel L. Schacter, Department of Psychology,Harvard University, 33 Kirkland Street, Cambridge, Massa-chusetts 02138.

tests is that it is little affected or entirely unaffected bymanipulations of semantic versus nonsemantic studyprocessing, which have large effects on explicit mem-ory. Several studies have shown that whereas explicitmemory is much higher after study tasks that requiresemantic elaboration than after study tasks that re-quire low-level structural or perceptual analysis, prim-ing effects on identification, completion, and similartasks are of comparable magnitude following the twotypes of study tasks (cf. Bowers & Schacter, 1990;Graf & Mandler, 1984; Jacoby & Dallas, 1981;Roediger, Weldon, Stadler, & Riegler, 1992; Schacter,Cooper, & Delaney, 1990; Schacter & McGlynn,1989). These and related observations have led to theproposal that priming is based to a large extent on apresemantic perceptual representation system (PRS)that is composed of modality-specific subsystems thatrepresent information about the form and structure,but not the meaning and associative properties, ofwords and objects (Schacter, 1990, 1992a, 1992b;Tulving & Schacter, 1990; cf. Gabrieli, Milberg,Keane, & Corkin, 1990; Keane, Gabrieli, Fennema,Growdon, & Corkin, 1991).

Although the foregoing ideas have been directedlargely toward visual PRSs, and are supported byobservations from visual implicit memory tests, recentresearch has revealed a similar pattern of effects in theauditory domain. Schacter and Church (1992) exam-ined priming effects from study-list exposure to spo-ken words on two auditory tasks: identification ofwords in white noise and completion of initial syllablestems with the first word that comes to mind. Previousresearch had shown that priming occurs on these tasksand that it is to a large extent modality specific(Bassili, Smith, & MacLeod, 1989; Jackson &

107

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

108 SCHACTER, McGLYNN, MILBERG, AND CHURCH

Morton, 1984). In a series of experiments, Schacterand Church manipulated type of study processing:Subjects encoded spoken words either by making se-mantic judgments (e.g., by indicating the category towhich a word belongs) or by making nonsemanticjudgments (e.g., by rating the pitch of the speaker'svoice). Performance on explicit memory tests (recog-nition or cued recall) was much higher after semanticthan nonsemantic encoding, whereas the magnitude ofpriming on the identification and completion taskswas either less affected or unaffected by the samemanipulations.

Schacter and Church (1992) argued that the primingthey observed was based to a large extent on a pre-semantic auditory PRS subsystem that represents thespoken form of words independently of their meaningand noted that converging evidence for such a sub-system is provided by observations of patients whoexhibit the phenomenon of word-meaning deafness.These patients show a severe impairment in compre-hending spoken words, but their ability to repeatwords aloud and write them to dictation is relativelypreserved (cf. Ellis, 1984; Kohn & Friedman, 1986).Although such patients are quite rare, their pattern ofpreserved and impaired abilities is theoretically im-portant because it suggests that processing and repre-sentation of spoken word forms is handled by a sub-system that is distinct from semantic memory (Allport& Funnell, 1981; Ellis, 1984; Ellis & Young, 1988).One strong prediction of the PRS view is that patientswho exhibit word-meaning deafness should show nor-mal priming effects despite their semantic impairment(Schacter & Church, 1992).

In this article, we describe a case of word-meaningdeafness and report the results of two experiments inwhich we examined whether this patient showedpriming on auditory word identification tests. Thekey outcome of the study is that the patient showedintact priming relative to a group of matched controlsubjects.

Experiment 1

In the first experiment, we assessed priming with anauditory identification task used previously by Jack-son and Morton (1984) and Schacter and Church(1992): After auditory study of a list of commonwords, both studied and nonstudied words weremasked by white noise, and subjects attempted toidentify them.

Method

Case report. JP is a 65-year-old right-handed manwho was admitted to the hospital in October 1991 for ageriatric evaluation related to his safety at home. He had ahistory of hypertension, non-insulin dependent diabetesmellitus, and multi-infarct dementia. JP sustained a leftmiddle cerebral artery infarct in 1977, which reportedly re-sulted in Wernicke's aphasia and a right visual field cut.Several additional infarcts have been documented sincethat time.

JP's magnetic resonance image (MRI) in May 1991 re-vealed a large infarction in the territory of the left middlecerebral artery (see Figures 1 and 2). The resulting denselesion included the posterior inferior temporal gyrus, mid-dle temporal gyrus, planum temporale (including Heschl'sgyrus), and the white matter deep to these structures. Thelesion becomes patchy as it extends to the supramarginalgyrus and angular gyrus. There also appears to be an addi-tional small infarction affecting the inferior occipital cor-tex, including part of the calcarine cortex. In the right ce-rebral hemisphere, JP also appears to have suffered from anumber of small (less than 1 cm in diameter) lacunes andinfarctions involving the VPL nucleus of the thalamus, thelenticulostraiate area (including small areas of the globuspallidus, lenticular nucleus, lateral putamen, and the headof caudate), as well as the internal and external capsule.

JP had a lOth-grade education. He worked in construc-tion and owned two liquor stores prior to his retirement. Atthe time of his hospitalization, he still owned property andcollected rent from tenants living there. He was marriedfor about 25 years and had three children, but his relation-ship with his wife had been quite distant for many years.His family was concerned about his ability to function in-dependently at home. He was often left unsupervised athome and had forgotten about things on the stove. His wifereported that JP spent most of his time in front of the tele-vision, eating food that was contraindicated by his diabe-tes, and driving to the local restaurant despite two recentcar accidents.

During his most recent neuropsychological evaluationas an inpatient, JP was frequently observed reading thenewspaper. His spontaneous speech was fluent and charac-terized by appropriate grammar and prosody. However, hemade a number of paraphasic errors (both phonemic andsemantic) and exhibited a restricted vocabulary and vagueresponses to questions. He generally exhibited an appropri-ate range of affect, though he occasionally appeared disin-hibited in his behavior. JP was cooperative, attentive, andmotivated during several testing sessions, though he exhib-ited considerable frustration on tasks that he had difficultyunderstanding.

Assessment of JP's language functioning is summarizedin Tables 1 and 2. The evaluation included the BostonDiagnostic Aphasia Examination (BDAE; Goodglass &Kaplan, 1972), the Controlled Oral Word Association Test(Benton & Hamsher, 1976), the Boston Naming Test

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

IMPLICIT MEMORY AND COMPREHENSION 109

Figure 1. Tl-weighted axial view of JP's lesions. (Image extends from the level of the midbrainto the trigone of the lateral ventricles. Left and right are reversed.)

Figure 2. Tl-weighted parasagittal view of JP's principal lesion in the left cerebral hemisphere.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.


Table 1Boston Diagnostic Aphasia ExaminationSubscores

Scale/subscale Score

Auditory ComprehensionWord DiscriminationBody-Part IdentificationCommandComplex Material

NamingResponsive NamingConfrontation NamingAnimal Naming

RepetitionWordsHigh-Probability PhrasesLow-Probability Phrases

Oral ReadingWordsSentencesWord Picture MatchWord Recognition

Reading ComprehensionSentences and ParagraphsOral Spelling

WritingPrimer DictationConfrontation NamingSpelling to DictationSentences to DictationNarrative Writing

44/726/201/153/12

15/30108/114

6/23

9/104/81/8

30/308/10

10/107/8

8/107/8

14/1510/107/106/123/5

(Kaplan, Goodglass, & Weintraub, 1978), the Sound Rec-ognition Test—Multiple Choice of Pictures (Spreen &Benton, 1969), and a modified version of the PeabodyPicture Vocabulary Test (PPVT; Dunn, 1965). The PPVTwas modified by dividing it into two parts such that odd-numbered items were presented auditorily and even-numbered items were presented visually in typewrittentext.

Test results revealed severely impaired auditory com-prehension on all BDAE subtests. Overall, JP's mean rankon the BDAE tests of auditory comprehension, relative toother aphasic patients, was at about the 20th percentile. Incontrast, JP exhibited much milder problems when re-quired to repeat single words or complex phrases or towrite words to dictation, ranking at about the 70th percen-tile relative to other aphasics. JP's auditory comprehensiondeficit is also highlighted by the fact that he exhibited ex-tremely poor performance on the PPVT when words werepresented auditorily and exhibited relatively intact perfor-mance when words were presented visually (Table 2). JP'sreading comprehension was slow but relatively intact, aswas written confrontation naming, placing him at about the80th percentile. The patient reported that he frequently hasto write down words that he hears in order to understandtheir meaning. Interestingly, he had no difficulty identify-ing numbers or individual letters from auditory presenta-tion, and he was able to recognize environmental sounds.

For the present purposes, then, the critical feature of JP'sperformance on auditory tasks is a severe deficit in com-prehending spoken words and a relatively intact ability torepeat them and write them to dictation.

An estimate of JP's level of premorbid intellectual func-tioning, based on his occupation and his score on themultiple-choice version of the Information subtest (19/29)of the Wechsler Adult Intelligence Scale—Revised(WAIS-R; Wechsler, 1981), places him in the averagerange. Other verbal subtests of the WAIS-R were not ad-ministered because of JP's language difficulties. Heachieved a Performance IQ of 87 on the WAIS-R, a scorein the low average range for his age. His Mini MentalState Exam (Folstein, Folstein, & McHugh, 1975) score(21/30) revealed cognitive decline from premorbid levels.A number of subtests from the Wechsler Memory Scale—Revised (WMS-R; Wechsler, 1987) were administered andrevealed impaired delayed visual memory but intact verbalmemory when information was presented in written form(Figural Memory = 6, Logical Memory = 18, VisualPaired Associates I = 8, Visual Reproduction I = 32, DigitSpan = 9, Logical Memory II = 12, Visual Paired Associ-ates II = 2, and Visual Reproduction II = 5). Additionalneuropsychological testing revealed impaired facial recog-nition on Milner's (1968) Test of Unfamiliar Faces (6/12);impaired mental assembly of object puzzles (HooperVisual Organization Test score = 20.5/30 [Hooper, 1958]);intact copying of the Rey-Osterrieth Complex Figure (29/36; Rey, 1964) but poor incidental recall on an immediate(7.5/36) and delayed trial (6/36); impaired word list gener-ation to both phonemic cues (F = 6, A = 5, and 5 = 5)and a semantic cue (grocery store = 6; Controlled OralWord Association test, Benton & Hamsher, 1976); im-paired confrontation naming (on the Boston Naming Test,Kaplan, Goodglass, & Weintraub, 1978; spontaneous cor-rect responses = 24, correct responses with stimulus cue= 5); and impaired ability to establish, maintain, and shiftcognitive set on the modified Wisconsin Card Sorting Test

Table 2Neuropsychological Assessment of JP'sLanguage Abilities

Scale Score

Boston Naming TestSpontaneous responsesStimulus-cued responses

Total correctControlled Oral Word Association

FASGrocery store

Sound Recognition Test—Multiple Choice Pictures

Series ASeries B

Peabody Picture Vocabulary TestAuditoryVisual

24/605/60

29/60

56

11/1312/13

44/8772/87

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.


(Hart, Kwentas, Wade, & Taylor, 1988; Nelson, 1976; cat-egories = 4, perseverative errors = 21).

Control subjects. The controls were 4 men withoutany known neurological deficits who were matched to JPfor age (M = 63 years, range = 62-68 years) and educa-tion level (M = 10. 3 years, range = 8-12 years) and whoalso had similar work backgrounds.

Design, materials, and procedure. Given the smallnumber of observations concerning the performance of asingle patient in a single paradigm, we attempted to in-crease the reliability of our observations by replicating theexperiment on two different occasions, using different setsof materials on each occasion. The basic design and proce-dure was patterned after Schacter and Church's (1992) Ex-periments 1 and 2. Subjects initially heard a series of 24familiar words, presented over headphones at normal con-versation levels, that were spoken by either a man or awoman. The encoding task required subjects to repeat eachword aloud and to indicate whether it was spoken by aman or a woman. Instructions were presented in writtenform for the encoding task and subsequent tests to ensurethat JP understood task requirements. After presentation ofthe study list, there was a delay of several minutes duringwhich subjects completed a brief filler task that involvedgenerating names of states. Instructions for the word iden-tification test were then given. Subjects were informed thatthey would be hearing a series of words presented in noise,that we were interested in their subjective perception ofwhat the masked words sounded like, and that their taskwas to respond with the first word that came to mind foreach stimulus presentation. The identification test con-sisted of 48 critical words, 24 old words from the studylist, and 24 new, nonstudied words that provided an esti-mate of baseline identification performance. For the oldwords, half were spoken by the same voice as during thestudy list and the other half were spoken by a differentvoice; voice change always involved a change in thespeaker's gender. For the new words, half of the wordswere spoken by a man and half by a woman. The noiselevel was the same as that used by Schacter and Church(1992), and it produced extremely degraded stimuli: Col-lege students were able to identify correctly approximately25% of nonstudied words.

Immediately following the identification task, a yes/norecognition test was given for the same 48 words, exceptthat no white noise was used. Subjects were instructed thatwe were interested in what they remembered from the be-ginning of the experiment, when they decided whetherwords were spoken in a male or female voice. They wereinstructed to answer yes when they remembered that aword had been spoken during the initial phase of the ex-periment and to answer no when they did not rememberthat a word had been spoken during the initial phase of theexperiment. No mention was made of the voice manipula-tion. Several weeks after the conclusion of this session, allsubjects were given the same sequence of study task, wordidentification test, and yes/no recognition test with a new

set of 48 words; 24 of them appeared on the study list, and24 served as nonstudied or baseline items.

The materials for the first study-test session were 48medium-to-low frequency words used by Schacter andChurch (1992). The two sets of 24 studied and nonstudiedwords were matched as closely as possible for wordlength, frequency, and concreteness. We also devised asecond set of 48 words that was similar to the first, and inwhich the sets of 24 studied and nonstudied words wereagain matched as closely as possible for word length, fre-quency, and concreteness. For both sets of materials, the24 study-list words were spoken by three men and threewomen, with four words randomly assigned to each speak-er. Similarly, the 24 nonstudied words in each set werealso spoken by three men and three women. Half of thewords were tested in the same voice in which they werespoken during the study task, and half were tested in a dif-ferent (opposite-gender) voice.

To examine JP's comprehension of words that were pre-sented on the study lists, we administered two differenttests after each of the priming sessions. The first test re-quired JP and the control subjects to provide a definitionof each word from the study list. Words were spoken indi-vidually by the experimenter, and subjects were given asmuch time as they needed to provide a definition. The def-initions were scored by two independent raters—Susan M.McGlynn and a staff member who was unaware of the na-ture of the experiment and the identity of the subjects. Theraters followed a scoring system similar to that used on theWAIS-R Vocabulary subtest: They assigned a score of 0,1, or 2 to each definition, where 0 indicates an incorrectdefinition, 1 indicates a partially correct definition, and 2indicates a complete definition. The second test involvedmaking judgments of semantic relatedness: For each targetword, subjects were given two alternatives and were askedto indicate which of the two was more related in meaningto the target. The correct choices were strongly related tothe targets (e.g., nail for target word hammer), whereas theincorrect choices were weakly or only moderately relatedto the targets (e.g., hook for hammer). Informal pilot workindicated that normal subjects perform at a 100% level ofaccuracy on this test.

Results

Word identification. Table 3 presents the mainword identification data for JP and the matched con-trol subjects for the two test sessions and for studiedand nonstudied words. Data were scored according toa strict criterion (only correct spellings allowed) and amore lenient criterion (misspellings counted as cor-rect). The overall pattern of results was the sameaccording to both criteria, and because misspellingswere relatively infrequent, the more lenient scoringcriterion was adopted.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.


Table 3Proportion of Studied and Nonstudied Words Identified Correctly by JP and Control Subjectson the Auditory Identification Test in Experiment 1

Subject

JPControls

Studied

.208

.281

List 1

Non-studied

.042

.167

Primingscore3

.166

.114

Studied

.083

.167

List 2

Non-studied

.042

.104

Primingscorea

.041

.063

Studied

.146

.244

Overall

Non-studied

.042

.136

Primingscore"

.104

.088' Computed by subtracting proportion of nonstudied words identified from proportion of studied words identified.

The data in Table 3 are reasonably clear-cut. Onthe one hand, JP had difficulty with the word identi-fication task, as indicated by his low level of base-line performance (.042) relative to that of the controlsubjects (.136). Nevertheless, the critical finding isthat JP showed a normal priming effect on the iden-tification task: When identification of nonstudieditems is subtracted from that of studied items, heachieved a priming score of .104, whereas the con-trol subjects achieved a priming score of .088. Theperformance of JP and the controls was similar inas-much as both exhibited stronger evidence of primingon the first list than on the second, which may reflectdifferences in the materials that were used on thetwo lists. JP showed somewhat more priming thantwo controls, whose overall priming scores wereeach .063; he showed somewhat less priming thanone control, whose priming score was .125; and heshowed equivalent levels of priming to the othercontrol, whose priming score was .104. Thus, itseems safe to conclude that the priming effect exhib-ited by JP falls well within the range of the matchedcontrol subjects.

It was also possible to examine the effect ofchanging the speaker's voice between study and teston priming, although the small number of observa-tions per cell in same- and different-voice conditions(n = 12) renders such comparisons tentative. JPshowed similar levels of performance in the same-and different-voice conditions, identifying one moreword in the former condition than in the latter. Con-

trol subjects identified equal numbers of items in thesame- and different-voice conditions. The lack ofvoice-change effects is consistent with previous in-vestigations of college students that examined prim-ing of words masked by white noise (Jackson &Morton, 1984; Schacter & Church, 1992).

Recognition memory. The recognition data, dis-played in Table 4, are partitioned into hits (yes re-sponses to words that had appeared on the study list)and false alarms (yes responses to words that had notappeared on the study list). These data contrastmarkedly with the priming results: JP exhibited es-sentially no recognition memory and was severelyimpaired relative to the control subjects. In fact, JPresponded no to all items—both those that appearedon the study list and those that did not—and claimedthat he simply did not remember having heard any ofthe test items previously. In contrast, when recogni-tion performance was collapsed across the two lists,all of the control subjects had more hits than falsealarms, and their overall hit rate (.547) was muchhigher than their false-alarm rate (.229). The controlsubjects' recognition performance was considerablylower on the first list than on the second, but this re-sult is largely attributable to the unusually poor per-formance of one subject, whose false-alarm rate onthe first list (.417) actually exceeded his hit rate(.375). However, this subject showed high levels ofrecognition on the second list (hit rate = .583, false-alarm rate = .125), which were comparable to thescores attained by the other control subjects, thus

Table 4Proportion of Yes Responses to Studied and Nonstudied Words by JP and Control Subjectsin the Explicit Recognition Test in Experiment 1

List 1 List 2 Overall

Non- Recognition Non- Recognition Non- RecognitionSubject Studied studied score" Studied studied score" Studied studied score"

JPControls

.000

.479.000.302

.000

.177.000.614

.000

.156.000.454

.000

.547.000.229

.000

.318

Note. For studied words, proportion indicates hit rate; for nonstudied words, proportion indicates false-alarm rate.a Computed by subtracting false-alarm rate from hit rate.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.


suggesting that his poor performance on the first listwas an aberration.

Comprehension of studied words. JP's compre-hension of words from the study list, as assessed byhis performance on the definition-production taskand the forced-choice relatedness task, was markedlyimpaired relative to that of the control subjects.

For the definition-production task, interrateragreement was quite high (.896 agreement), so datawere collapsed across the two raters. For each sub-ject, a proportional score was computed by dividingthe total number of points that the two ratersawarded by the total number of possible points. Forcontrol subjects, the mean score was .942 (range =.8S5-.969). In contrast, JP achieved a proportionalscore of .115, thus exhibiting a severe impairment. Itwas of further interest to compare JP's definitions forwords that he did and did not identify correctly, al-though the small number of words identified cor-rectly renders such an analysis tentative. JP's propor-tional score was .143 for words identified correctlyand . 103 for words not identified.

On the two-choice semantic-relatedness test, allcontrol subjects performed at a 100% level of accu-racy for all items on both lists. Thus, the test wasquite easy for normal subjects. JP scored consider-ably lower than the controls, although well above thechance level of .500, achieving .750 correct overall(.667 correct for the first list and .833 correct for thesecond list). There were no systematic differencesfor items that he identified correctly or incorrectly onthe identification task: JP was .714 (5/7) correct foridentified items and .756 (31/41) correct for non-identified items. Thus, there does not appear to be astrong relation between whether or not JP showedpriming on a word and his ability to gain access tosemantic knowledge of that word.

Discussion

The main result of Experiment 1 was that JPshowed a normal priming effect on the auditory iden-tification test despite poor comprehension of targetwords and a lack of explicit memory for them. Al-though these results appear to provide strong supportfor the PRS view of priming, two cautionary pointsneed to be made. First, the number of observations issmall, and only List 1 yielded clear evidence of robustpriming; both JP and the control subjects exhibitedweak priming effects on List 2. Second, JP's baselineidentification rate was near zero on both lists (.042)and lower than that of the control subjects (.136). In

view of this nearly nonexistent level of baseline per-formance, it is difficult to make strong claims aboutnormal priming effects.

To address these points and to obtain more infor-mation about auditory priming in JP, we performed anadditional experiment using a different priming task,one in which used words are made difficult to identifyby degrading them with a low-pass filter. We haverecently observed robust priming effects on the low-pass filter identification task in college students(Church & Schacter, 1992). In addition to allowing usto obtain more observations on JP and thereby assessthe reliability and generality of his priming, the low-pass filter task has two attractive features. First, iden-tification difficulty can be varied by manipulating thelevel of decibel reduction that is produced by thefiltering process. Thus, we hoped to raise JP's baselineidentification rate well above the near-zero levels ob-served in Experiment 1. Second, our previous workwith college students showed that priming on the low-pass filter identification task is influenced by study-to-test changes in the speaker's voice (Church &Schacter, 1992). Accordingly, we were able examinewhether JP, like normal subjects, would show evi-dence of voice-specific priming.

Experiment 2

Method

Subjects. JP and the same four matched control subjectsfrom Experiment 1 participated in the experiment.

Design, materials, and procedure. All features of themain experiment were similar to Experiment 1, with thefollowing exceptions. The main difference was that thelow-pass filter task was used to assess auditory identifica-tion. All target items were put through a low-pass filter inthe SoundEdit program on a Macintosh computer. For pa-tient JP, the filter reduced the decibel level of a distributionof frequencies between 1000 Hz and 2000 Hz by 0 to 20decibels and reduced all frequencies above 2000 Hz by 20decibels. The effect of the filtering was a slight muffling ofthe word. For control subjects, target words were degradedfurther in an attempt to produce more comparable levels ofbaseline performance: The maximal level of decibel reduc-tion was 60 dB instead of 20 dB, and the filtered wordssounded muffled. As in Experiment 1, six different speak-ers (three men and three women) were used, and eachword was recorded in both a male and a female voice. Twonew sets of moderate to low frequency words were used astarget materials. Each set consisted of 48 words and wasdivided into two subsets of 24 studied and 24 nonstudieditems. For each set, the two subsets were matched asclosely as possible for frequency, length, and concreteness.

Because the purpose of this experiment was to assess

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.


the reliability of JP's priming, and because he respondedno to all items on the recognition test in Experiment 1, wedropped the recognition test from Experiment 2. A finalchange involved the assessment of JP's comprehension ofstudied words: Because the definition-production task andrelatedness-judgment task yielded similar patterns of re-sults in Experiment 1, we assessed comprehension onlywith the definition-production task. In addition, becausecontrol subjects scored at ceiling levels on the definition-production task, we administered the task only to JP.

All testing was done in a single session. Instructionsand procedure for the encoding task were as described forExperiment 1. Following completion of the same brief dis-tractor task that was used in the first experiment, subjectswere told that they would be exposed to a series of filteredwords that would sound somewhat muffled, that we wereinterested in their subjective perception of what they heard,and that they should respond by writing down the firstword that came to mind. After a short break, the same pro-cedure was repeated for the second list. In addition, JP wasgiven the definition-production task at the conclusion oftesting for each list.

Results and Discussion

Word identification. The data from the low-passfilter task are presented in Table 5. Three pointsabout these data are worth noting. First, JP exhibiteda much higher level of baseline performance on bothlists (.250) than he did in Experiment 1, although itwas still lower than that of the control subjects. Sec-ond, JP showed a large overall priming effect, identi-fying .416 of the studied words correctly. As in Ex-periment 1, JP's performance varied across lists: Heshowed a large priming effect on List 2 and a weakpriming effect on List 1. Third, the overall magni-tude of priming exhibited by JP (.167) was quitesimilar to the mean priming score of the control sub-jects (.198), and as in Experiment 1, he scored wellwithin the range of the control subjects. JP showedmore priming than two controls, whose primingscores were .125 and .145, respectively, and lesspriming than the other two controls, whose primingscores were .270 and .250, respectively. Although

JP's priming score on the first list was quite small(.042), one of the control subjects failed to show anypriming on this list. Thus, both JP and the controlsubjects showed more robust priming on List 2 thanon List 1.

We also examined identification of studied wordsas a function of the voice manipulation. JP showedstrong evidence of voice-specific priming: His over-all priming score in the same-voice condition (.292)was much higher than in the different-voice condi-tion (.042). For control subjects, overall primingscores were also higher in the same-voice condition(.235) than in the different-voice condition (.162),but the magnitude of the voice-change effect was notas large. For both JP and the control subjects, nearlyall of the evidence for voice-specific priming wasprovided by List 2. On List 1, JP identified one moreitem in the same-voice than in the different-voicecondition, and the controls identified on averageone item less in the same-voice than in the different-voice condition. In view of the fact that strongvoice-change effects were observed on only one ofthe lists, and that the number of observations percell is quite small for these comparisons, we mustbe cautious about interpreting the evidence on voice-specific priming. It seems safe to conclude, however,that JP and the controls showed the same overallpattern of voice-specific priming (i.e., greater voice-change effect on List 2 than on List 1) and thatthere is suggestive evidence that JP was more sus-ceptible to voice-change effects than were the con-trol subjects.

Comprehension of studied words. JP's perfor-mance on the definition-production task for studiedwords was scored by the two raters from Experiment1 in the same manner as described previously, andthe same proportional score described in Experiment1 was computed from their ratings. As in Experiment1, there was relatively high agreement between thetwo raters (.875), so their mean scores were used.Once again, JP exhibited little understanding of thetarget words, achieving a mean proportional score of

Table 5Proportion of Studied and Nonstudied Words Identified Correctly by JP and Control Subjectson the Auditory Identification Test in Experiment 2

Subject

JPControls

Studied

.292

.635

List 1

Non-studied

.250

.500

Primingscore3

.042

.135

Studied

.542

.792

List 2

Non-studied

.250

.510

Primingscore3

.292

.282

Studied

.417

.703

Overall

Non-studied

.250

.505

Primingscore"

.167

.198

' Computed by subtracting proportion of nonstudied words identified from the proportion of studied words identified.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.


.099. His scores were low both for words identifiedcorrectly (.113) and words that were not identifiedcorrectly (.089).

General Discussion

The critical finding of the present study was that JPshowed intact priming of spoken words on two dif-ferent kinds of auditory identification tests despite hiscomprehension deficit. Even though he exhibited amarked inability to understand the study-list words, asindicated by his severe impairment on the definition-production and semantic-matching tests, he showedjust as much priming as did the controls: Collapsedacross the four separate study and test sessions thatconstituted the two experiments, IP's mean primingscore was .136, and the control subjects' mean prim-ing score was .143. These results provide confirma-tion of the prediction made by the PRS account ofauditory priming and are also consistent with relatedtheoretical views that hold that priming on identifica-tion and similar implicit tasks depends heavily onperceptual processes (cf. Keane et al., 1991; Roediger,1990; Squire et al., 1992). Taken together withSchacter and Church's (1992) finding that priming onthe auditory identification test in normal subjects isrobust following nonsemantic study tasks, our resultsprovide converging empirical support for the idea thatpriming on perceptual or data-driven tasks such asauditory identification depends heavily on a prese-mantic system.

Although our data are entirely consistent with thePRS account, a number of qualifications and cautionsshould be noted. First, because this study is based ona single subject, we must reserve judgment concern-ing the generality of our results pending the investi-gation of additional patients. Second, as noted earlier,JP's baseline level of identification performance waslower than that of the control subjects. The compari-son of priming scores at different levels of baselineperformance is not entirely straightforward and callsfor interpretive caution (for a discussion, see Keaneet al., 1991; Schacter, Kihlstrom, Kaszniak, & Val-diserri, in press). Note, however, that JP showed ro-bust priming both in Experiment 1, when his baselinescore was quite low (.042), and in Experiment 2, whenit was considerably higher (.250). Moreover, the mag-nitude of priming exhibited by JP was as large as thatexhibited by the control subjects both when absolutepriming scores were considered, as we did in bothexperiments, and when proportional priming scores(i.e., percentage increase over baseline) were consid-

ered. Thus, it appears that the priming exhibited by JPwas not strongly related to baseline performance.Nevertheless, the fact that JP's baseline performancewas substantially lower than that of the control sub-jects in both experiments means that we must be cau-tious about claiming that he exhibited entirely normalpriming. What we can state confidently, however, isthat JP showed robust priming despite his comprehen-sion deficit.

A third cautionary note concerns the extent towhich JP's priming can be characterized as prese-mantic. We have argued that the combination of JP'srobust priming and impaired auditory comprehensionsupport the notion that priming of auditory wordidentification operates at a presemantic level. Thefinding in both experiments that JP exhibited im-paired comprehension of studied words that he iden-tified correctly is consistent with this view. Nonethe-less, it is possible that even when JP failed tounderstand a spoken word explicitly, some semanticinformation about it was activated. For example, pre-vious research has demonstrated that Wernicke'saphasics show indirect or semantic priming of wordsthat they fail to comprehend explicitly (cf. Blum-stein, Milberg, & Shrier, 1982; Milberg & Blum-stein, 1981; Milberg, Blumstein, & Dworetzky,1988), and it is quite possible that JP would showsimilar evidence for implicit semantic activation.Given that this possibility exists, the appropriateconclusion from the present study is that primingof auditory word identification does not dependon explicit comprehension of a spoken word andin that sense can be thought of as a presemanticphenomenon.

If we assume that priming in JP depends on a pre-semantic auditory PRS, an important question con-cerns the neural basis of this system and the extent towhich it is spared by JP's lesion. Recent data fromneuroimaging studies using positron emission tomog-raphy (PET) indicate that cortical areas in the vicinityof the left angular gyrus and supramarginal gyrus playan important role in the processing and representationof auditory word-form information (Petersen, Fox,Posner, Mintun, & Raichle, 1989). We noted earlierthat MRI evidence indicates that JP's major lesion isquite extensive in the anterior portions of Wernicke'sarea but becomes patchy in the regions of the angularand supramarginal gyri. Thus, it seems plausible tosuggest that JP's spared priming is mediated by alargely preserved auditory word-form system thatdepends critically on the left angular and supramar-ginal gyri. The extensive lesion in other regions of

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.


Wernicke's area may have effectively disconnected theoutput of the auditory word-form system from regionsinvolved in semantic representation, which may be de-pendent on anterior cortical regions in the temporal andfrontal lobes (e.g., Petersen et al., 1989).

It is also possible, of course, that JP's intact primingis mediated by his spared right hemisphere. We(Schacter, 1992a, 1992b; Schacter & Church, 1992)have suggested that there may be two components toauditory priming: a voice-specific component that de-pends primarily on the right hemisphere and an ab-stract, voice-independent component that depends pri-marily on the left hemisphere. This argument is basedon three different types of observations: (a) Study-to-test changes in speaker's voice have no effect on prim-ing when priming is assessed with the identification-in-noise test used in Experiment 1, whereas voice-change effects are observed when priming is assessedwith implicit tests that do not use white noise (e.g.,identification with low-pass filter or auditory stemcompletion); (b) the right hemisphere's ability to pro-cess spoken words seems to be especially impaired bybackground noise (Zaidel, 1978); and (c) in a recentstudy of auditory stem completion priming that used adichotic listening procedure, it was observed that theleft hemisphere showed significant priming in bothsame- and different-voice conditions, whereas theright hemisphere showed significant priming inthe same-voice condition but failed to show anypriming in the different-voice condition (Schacter,Aminoff, & Church, 1992; cf. Marsolek, Kosslyn, &Squire, 1992). The fact that JP exhibited a large voice-change effect in Experiment 2 is consistent with theidea that his right hemisphere played a role in the ob-served priming, and the apparent trend for larger voice-change effects in JP than in the controls raises the pos-sibility that priming in JP depends to a greater extenton the right hemisphere than it does in the control sub-jects. However, when words are embedded in noise onthe auditory identification test, as they were in Exper-iment 1, the right hemisphere may be effectively ex-cluded from contributing to priming effects. BecauseJP showed robust priming with the identification-in-noise procedure, it seems unlikely that priming de-pended entirely on right hemisphere involvement.Nevertheless, we cannot rule out this possibilityconclusively.

Our observations concerning auditory priming in JPraise the question of whether similar effects wouldbe observed in patients who exhibit form versus se-mantic dissociations in visual domains, where primingmay be mediated by modality-specific PRS sub-

systems. For example, patients have been describedwho can read regular and irregular printed words butdo not understand them (cf. Funnell, 1983; Sartori,Masterson, & Job, 1987; Schwartz, Saffran, & Marin,1980), and these observations have been cited in sup-port of the idea that a presemantic visual word-formsystem supports visual word priming on a variety ofimplicit tasks (Schacter, 1990). Similarly, dissocia-tions between relatively intact access to knowledge ofthe structure of visual objects despite impaired accessto knowledge of their functional and associative prop-erties (cf. Riddoch & Humphreys, 1987; Warrington,1975, 1982; Warrington & Taylor, 1978) have pro-vided a basis for arguing that a presemantic structuraldescription system supports priming effects for visualobjects (Schacter, 1990; Schacter et al., 1990). Itwould be of considerable interest to determinewhether such patients, like JP, show robust perceptualpriming for visual words or objects that they do notcomprehend. Such research would yield informationconcerning the generality of the dissociation that wehave observed, provide a strong test of the PRS ac-count of priming, and illuminate the nature of therelation between perceptual and conceptual memoryprocesses.

References

Allport, D. A., & Funnell, E. (1981). Components of themental lexicon. Philosophical Transactions of the RoyalSociety of London (Series B), 295, 397-^10.

Bassili, J. N., Smith, M. C., & MacLeod, C. M. (1989).Auditory and visual word stem completion: Separatingdata-driven and conceptually driven processes. QuarterlyJournal of Experimental Psychology, 41 A, 439^453.

Benton, A. L., & Hamsher, K. de S. (1976). MultilingualAphasia Examination (rev. ed.) [manual]. Iowa City: Uni-versity of Iowa.

Blumstein, S. E., Milberg, W., & Shrier, R. (1982). Semanticprocessing in aphasia: Evidence from an auditory lexicaldecision task. Brain and Language, 17, 301-315.

Bowers, J. S., & Schacter, D. L. (1990). Implicit memoryand test awareness. Journal of Experimental Psychology:Learning, Memory, and Cognition, 16, 404^16.

Church, B. A., & Schacter, D. L. (1992, November). Implicitmemory for speaker's voice: Intonation and fundamentalfrequency. Paper presented at the 33rd Annual Meeting ofthe Psychonomic Society, St. Louis, MO.

Dunn, L. M. (1965). Expanded manual for the PeabodyPicture Vocabulary Test. Minneapolis, MN: AmericanGuidance Service.

Ellis, A. W. (1984). Introduction to Byrom Bramwell's(1887) case of word-meaning deafness. Cognitive Neu-ropsychology, I, 245-248.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.


Ellis, A. W., & Young, A. W. (1988). Human cognitiveneuropsychology. Hove, England: Erlbaum.

Folstein, M. R, Folstein, S. R, & McHugh, P. R. (1975)."Mini-Mental State": A practical method for grading thecognitive state of patients for the clinician. Journal ofPsychiatric Research, 12, 189-198.

Funnell, E. (1983). Phonological processes in reading: Newevidence from acquired dyslexia. British Journal of Psy-chology, 74, 159-180.

Gabrieli, J. D. E., Milberg, W., Keane, M. M., & Corkin, S.(1990). Intact priming of patterns despite impaired mem-ory. Neuropsychologia, 28, 417^28.

Goodglass, H., & Kaplan, E. (1972). Assessment of aphasiaand related disorders. Philadelphia: Lea & Febiger.

Graf, P., & Mandler, G. (1984). Activation makes wordsmore accessible, but not necessarily more retrievable.Journal of Verbal Learning and Verbal Behavior, 23, 553-568.

Graf, P., & Schacter, D. L. (1985). Implicit and explicitmemory for new associations in normal subjects and am-nesic patients. Journal of Experimental Psychology:Learning, Memory, and Cognition, 11, 501-518.

Hart, R. P., Kwentas, J. P., Wade, J. B., & Taylor, J. R.(1988). Modified Wisconsin Sorting Test in elderly de-pressed and demented subjects. The Clinical Neuropsy-chologist, 2, 49-56.

Hooper, H. E. (1958). The Hooper Visual Organization Test[manual]. Los Angeles: Western Psychological Services.

Jackson, A., & Morton, J. (1984). Facilitation of auditoryword recognition. Memory and Cognition, 12, 568-574.

Jacoby, L. L., & Dallas, M. (1981). On the relationshipbetween autobiographical memory and perceptual learn-ing. Journal of Experimental Psychology: General, 110,306-340.

Kaplan, E. R, Goodglass, H., & Weintraub, S. (1978).The Boston Naming Test. Boston: E. F. Kaplan & H.Goodglass.

Keane, M. M., Gabrieli, J. D. E., Fennema, A. C., Growdon,J. H., & Corkin, S. (1991). Evidence for a dissociationbetween perceptual and conceptual priming in Alzhe-imer's disease. Behavioral Neuroscience, 105, 326-342.

Kohn, S. E., & Friedman, R. B. (1986). Word-meaningdeafness: A phonological-semantic dissociation. Cogni-tive Neuropsychology, 3, 291-308.

Marsolek, C. J., Kosslyn, S. M., & Squire, L. R. (1992).Form-specific priming in the right cerebral hemisphere.Journal of Experimental Psychology: Learning, Memory,& Cognition, 18, 492-508.

Milberg, W., & Blumstein, S. E. (1981). Lexical decisionand aphasia: Evidence for semantic processing. Brain andLanguage, 14, 371-385.

Milberg, W., Blumstein, S., & Dworetzky, B. (1988).Phonological processing and lexical access in aphasia.Brain and Language, 34, 279-293.

Milner, B. (1968). Visual recognition and recall after righttemporal lobe excision in memory. Neuropsychologia, 6,191-209.

Nelson, H. E. (1976). A modified card sorting test sensitiveto frontal lobe deficits. Cortex, 12, 313-324.

Petersen, S. E., Fox, P. T., Posner, M. I., Mintun, M. A., &Raichle, M. E. (1989). Positron emission tomographicstudies of the processing of single words. Journal ofCognitive Neuroscience, 1, 153-170.

Rey, A. (1964). L'examen clinique en psychologic [Theclinical examination in psychology]. Paris: Presses Uni-versitaires de France.

Richardson-Klavehn, A., & Bjork, R. A. (1988). Measuresof memory. Annual Review of Psychology, 36, 475-543.

Riddoch, M. J., & Humphreys, G. W. (1987). Visual objectprocessing in optic aphasia: A case of semantic accessagnosia. Cognitive Neuropsychology, 4, 131-186.

Roediger, H. L. III. (1990). Implicit memory: Retentionwithout remembering. American Psychologist, 45, 1043-1056.

Roediger, H. L. Ill, Weldon, M. S., Stadler, M. L., & Rie-gler, G. L. (1992). Direct comparison of two implicitmemory tests: Word fragment and word stem completion.Journal of Experimental Psychology: Learning, Memory,and Cognition, 18, 1251-1269.

Sartori, G., Masterson, J., & Job, R. (1987). Direct-routereading and the locus of lexical decision. In M. Coltheart,G. Sartori, & R. Job (Eds.), The cognitive neuropsychol-ogy of language (pp. 59-78). London: Erlbaum.

Schacter, D. L. (1987). Implicit memory: History and cur-rent status. Journal of Experimental Psychology: Learn-ing, Memory, and Cognition, 13, 501-518.

Schacter, D. L. (1990). Perceptual representation systemsand implicit memory: Toward a resolution of the multiplememory systems debate. In A. Diamond (Ed.), Annals ofthe New York Academy of Sciences: Vol. 608. Develop-ment and neural bases of higher cognition (pp. 543-571).New York: New York Academy of Sciences.

Schacter, D. L. (1992a). Priming and multiple memory sys-tems: Perceptual mechanisms of implicit memory. Jour-nal of Cognitive Neuroscience, 4, 244-256.

Schacter, D. L. (1992b). Understanding implicit memory: Acognitive neuroscience approach. American Psychologist,47, 559-569.

Schacter, D. L., Aminoff, A., & Church, B. A. (1992). [Adichotic listening study of auditory stem completion].Unpublished raw data.

Schacter, D. L., Chiu, C. Y. P., & Ochsner, K. N. (in press).Implicit memory: A selective review. Annual Review ofNeuroscience.

Schacter, D. L., & Church, B. (1992). Auditory priming:Implicit and explicit memory for words and voices. Jour-nal of Experimental Psychology: Learning, Memory, &Cognition, 18, 915-930.

Schacter, D. L., Cooper, L. A., & Delaney, S. M. (1990).Implicit memory for unfamiliar objects depends on accessto structural descriptions. Journal of Experimental Psy-chology: General, 119, 5-24.

Schacter, D. L., Kihlstrom, J. R, Kaszniak, A. K., & Vald-iserri, M. (in press). Preserved and impaired memory

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.


functions in elderly adults. In J. Cerella, W. Hoyer, J.Rybash, & M. Commons (Eds.), Adult information pro-cessing: Limits on loss. San Diego, CA: Academic Press.

Schacter, D. L., & McGlynn, S. (1989). Implicit memory:Effects of elaboration depend on unitization. AmericanJournal of Psychology, 102, 151-181.

Schwartz, M. F, Saffran, E. M., & Marin, O. S. M. (1980).Fractionating the reading process in dementia: Evidencefor word specific print-to-sound associations. In M. Colt-heart, K. Patterson, & J. C. Marshall (Eds.), Deep dyslexia(pp. 259-269). London: Routledge & Kegan Paul.

Shimamura, A. P. (1986). Priming effects in amnesia: Evi-dence for a dissociable memory function. Quarterly Jour-nal of Experimental Psychology, 38A, 619-644.

Spreen, O., & Benton, A. L. (1969). Neurosensory CenterComprehensive Examination for Aphasia. Victoria, Brit-ish Columbia, Canada: University of Victoria, Depart-memt of Psychology, Neuropsychological Laboratory.

Squire, L. R., Ojemann, J. E., Miezin, F. M., Petersen, S. E.,Videen, T. O., & Raichle, M. E. (1992). Activation ofhippocampus in normal humans: A functional anatomicalstudy of memory. Proceedings of the National Academyof Sciences, 89, 1837-1841.

Tulving, E., & Schacter, D. L. (1990). Priming and hur- nmemory systems. Science, 247, 301-306.

Warrington, E. K. (1975). The selective impairment of se-mantic memory. Quarterly Journal of Experimental Psy-chology, 27, 635-657.

Warrington, E. K. (1982). Neuropsychological studies ofobject recognition. Philosophical Transactions of theRoyal Society of London (Series B), 289, 15-33.

Warrington, E. K., & Taylor, A. M. (1978). Two categoricalstages of object recognition. Perception, 7, 695-705.

Wechsler, D. (1981). Wechsler Adult Intelligence Scale-Revised [manual]. New York: Psychological Corp.

Wechsler, D. (1987). Wechsler Memory Scale—Revised[manual]. New York: Psychological Corp.

Zaidel, E. (1978). Concepts of cerebral dominance in thesplit brain. In P. A. Buser & A. Rougel-Buser (Eds.),Cerebral correlates of conscious experience (pp. 263-284). Amsterdam: Elsevier.

Received June 12, 1992Revision received August 19, 1992

Accepted August 25, 1992 •

Correction to Editorial

In the editorial "Some Comments on the Goals and Direction ofNeuropsychology" by Nelson Butters (Neuropsychology, 1993, Vol. 7,No. 1, pp. 3-4), the name of the National Academy of Neuropsychologywas misspelled. This was due to an incorrect change by the printer at a latestage in production, after the proofs were properly reviewed by the Editor.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

copyright 1993 by the american psychological...

Documents