Perceptual Priming of Proper Names in Young and Older Normal ...

10 downloads 47 Views 1013KB Size Report
In Experiment 1, normative data were collected on 4 proper-name priming tasks. The variables of .... Such a deficit at the output stage would necessarily weaken.
Neuropsychology 1997, Vol. 11, No. 2, 232-242

Copyright 1997 by the American Psychological Association, Inc. 0894-4105/97/$3.00

Perceptual Priming of Proper Names in Young and Older Normal Adults and a Patient With Prosopanomia Anat Geva

Morris Moscovitch

University of Toronto

University of Toronto and Baycrest Center for Geriatric Care

Larry Leach Baycrest Center for Geriatric Care The purpose of this study was to determine if normal participants and a patient with prosopanomia can be perceptually primed for proper names. To this end, 2 experiments were conducted. In Experiment 1, normative data were collected on 4 proper-name priming tasks. The variables of levels of processing at encoding and age were m.anipulated. Robust priming results were obtained that were not influenced by either of these variables. The results of Experiment 2 indicated that prosopanomic patient N.G. demonstrated normal repetition priming, despite her marked impairment in deliberate retrieval of person and city names. These results are interpreted in terms of a dissociation between explicit and implicit memory for proper names and suggest that the deficit in prosopanomia only involves deliberate access to the name; access to presemantic representations of the visual word form remains intact.

It is well established that naming people is difficult, whether it is learning new names or retrieving familiar ones. This problem, exacerbated in normal old age, becomes insurmountable in some cases of neurological damage. The name given to this disorder is prosopanomia I (Carney & Temple, 1993), a deficit that refers to an inability to retrieve proper names of people, despite preserved intelligence and memory and relatively preserved retrieval of common nouns. Such a selective deficit can occur even though the ability to recognize the name and the capacity to provide a great deal of other information about the individual's identity are intact. Together these observations imply that the underlying naming deficit in prosopanomia is one of access, not of storage. The question addressed in this study is whether access to proper names can be facilitated by priming at a presemantic orthographic level. A positive finding would suggest that, at this presemantic level of representation, the distinction between common and proper names does not exist and would support the notion that the

deficit in prosopanomia occurs late in the naming process. In addition, the study would provide information about priming of proper names in young and older adults, for which no data are currently available. Proper names are more difficult to learn than other biographical information about a person. When participants are asked to learn the names, occupations, and hobbies associated with unfamiliar faces, person names take longer to learn and are recalled less well (Cohen, 1990; Cohen & Faulkner, 1986; Stanhope & Cohen, 1993), with older adults consistently performing more poorly than young adults (Cohen & Faulkner, 1984; Crook & West, 1990; Jones & Rabbitt, 1994). Even when the occupations and name are the same word (e.g., Baker vs. baker; McWeeny, Young, Hay, & Ellis, 1987), recall of occupation is superior to recall of names. Proper names of familiar people are also more susceptible to retrieval failures (Yarmey, 1973). Hanley and Cowell (1988) demonstrated that participants are often able to recognize a famous face as familiar and to retrieve a great deal of semantic information about the individual but are unable to recall the person's name. Parallel findings have been found in diary studies of everyday errors in person identification (Young, Ellis, & Flude, 1988; Young, Hay, & Ellis, 1985). Proper-name retrieval deficits have also been observed in the tip-of-the-tongue (TOT) phenomenon (Burke, Mackay,

Anat Geva, Department of Psychology, University of Toronto, Toronto, Ontario, Canada; Morris Moscovitch, Department of Psychology, University of Toronto, and Department of Psychology and Rotman Research Institute, Baycrest Center for Geriatric Care, Toronto, Ontario, Canada; Larry Leach, Department of Psychology, Baycrest Center for Geriatric Care, Toronto, Ontario, Canada. This research was supported by a grant from the Medical Research Council of Canada and an Ontario Mental Health Research Foundation Associateship. We are grateful to Brenda Melo for technical assistance. Correspondence concerning this article should be addressed to Anat Geva, who is now at the Department of Psychology, University of Michigan, 525 East University Avenue, Ann Arbor, Michigan 48109-1109. Electronic mall may be sent via Internet to either Anat Geva at [email protected] or Morris Moscovitch at momos @credit.erin.utoronto.ca.

1 Strictly speaking, the term prosopanomia refers only to a difficulty in naming faces, but the term has been extended to cover cases of proper-name anomia where the primary difficulty is one of naming people. Because the term has gained some currency, we use it in this article with the understanding that the deficit to which it refers involves people and some other types of proper names, such as countries and cities. 232

PERCEPTUAL PRIMING OF PROPER NAMES Worthley, & Wade, 1991). In the TOT state, a person is unable to produce a word, despite being able to demonstrate its intact representation in memory by various means, for example by correct recognition. TOTs for people names have been observed in experimental studies (Brennen, Baguely, Bright, & Bruce, 1990; Maylor, 1990) and occur frequently in everyday life for names of acquaintances (Brown, 1991; Burke et al., 1991). TOTs for proper names are presumed to occur at the stage of access to phonological representation of the name (Brennen et al., 1990; Hanley & Cowell, 1988), because even when the name is not accessible, participants in a TOT state can provide semantic information associated with the name (Cohen & Faulkner, 1986). Moreover, phonemic cuing (e.g., the person's initials) is effective in resolving proper-name TOTs (Brennen et al., 1990). Thus, the failure of normal adults to retrieve proper names may be characterized by intact access to semantic information but impaired access to the phonological or perhaps the lexical level of proper-name representation. Such a deficit at the output stage would necessarily weaken the mechanism by which names are deliberately retrieved. The peculiar difficulty in spontaneously retrieving proper names of people from memory has been further confirmed by clinical observations of patients with selective anomia for person names, despite the relatively preserved capacity to name common nouns, retrieve person-specific semantic information, and recognize the name (Flude, Ellis, & Kay, 1989). Patients showing a word-finding difficulty primarily confined to person proper names have been described by Carney and Temple (1993), Fery, Vincent, and Bredart (1995), Hittmair-Delazer, Denes, Semenza, and Mantovan (1994), and Lucchelli and de Renzi (1992). In some cases, the deficit may generalize to include geographical names (e.g., Semenza & Zettin, 1988, 1989). In this article, prosopanomic patient N.G. is described. N.G. is selectively impaired at retrieving person names from photographs, from description, or in word generation by category. Despite being unable to generate person names spontaneously, N.G. provides unequivocal evidence of preserved semantic knowledge and name recognition. She is able to provide semantic information about famous people and shows intact person-name recognition in tasks requiring her to point to the photograph named by the examiner among several alternatives. N.G.'s impairment also involves geographical names. Though she can provide names for countries, mountains, and rivers, she is impaired at naming cities. Evidence that N.G. retains an intact store of proper names and their related concepts is indicated by the following: (a) she is able to provide semantic information about the very same items she cannot name, and (b) she shows intact recognition of person and city names, despite impaired performance on proper-name free-recall tasks. This implies that the underlying naming deficit is one of deliberate access or retrieval, not of storage. The question addressed in this article is to what extent proper names can be accessed at a presemantic orthographic

233

level. Perceptual repetition priming for common words is presumed to involve access to presemantic representation of visual word-forms (Moscovitch, 1992; Moscovitch, GoshenGottstein, & Vriezen, 1994; Moscovitch, Vriezen, & GoshenGottstein, 1993; Schacter, 1992). In the prototypical repetition-priming study, participants are presented, in the first phase, with stimulus targets to study (e.g., a word list). In the second (test) stage, participants are not asked to retrieve intentionally the studied items, as would be required in an explicit retrieval task. Rather, they may be asked to identify a word from minimal cues or to categorize it without any reference to the study episode. Priming is demonstrated if performance is superior for the previously encountered targets in comparison to the nonstudied lures. The magnitude of the priming effect is the difference between performance on the target items and the nonstudied items. Priming on such implicit memory tasks has been dissociated from recall or recognition on explicit tests (e.g., Graf & Mandler, 1984; Jacoby & Dallas, 1981; Schacter & Graf, 1986). In contrast to conscious recollection, repetition priming is sensitive to manipulation of perceptual but not semantic variables (Roediger & McDermott, 1993). For example, a common finding is that levels-of-processing effects are minimal or nonexistent in perceptual-priming tasks (Graf & Mandler, 1984; Jacoby & Dallas, 1981; Roediger, Weldon, Stadler, & Riegler, 1992). Based on these findings, a number of investigators proposed that priming involves reactivation of stored presemantic representations in perceptual modules (Moscovitch, 1992; Moscovitch & Umilta, 1990) or perceptual representational systems (Schacter, 1990, 1992; Tulving & Schacter, 1990). These systems are less susceptible to aging and neurological damage associated with amnesia. Indeed, repetition priming is relatively preserved with these participants, whereas conscious recollection is impaired (Cermack, Talbot, Chandler, & Wolbarst, 1985; Graf, Squire, & Mandler, 1984; Greenbaum & Graf, 1989; Jacoby & Witherspoon, 1982; Light & Singh, 1987; Moscovitch, 1982; Parkin & Streete, 1988; Warrington & Weiskrantz, 1968, 1970). The purpose of this study was to determine whether comparable presemantic processes are involved in repetition priming of proper names. To our knowledge, no studies have been conducted on this topic. Because retrieval of proper names can be impaired selectively and may involve different processes than retrieval of common names, it is of interest to know whether repetition priming of proper names is similar to that of common nouns. To this end, two experiments were conducted. If repetition priming for proper names is also presemantic, then it should not be influenced by age and levels-of-processing manipulations in normal people. If it is mediated by perceptual representations in word-form systems, then perceptual priming should be intact in prosopanomia because the word-form system is intact. In Experiment 1, normative data on young and older adults were collected on four proper-name priming tasks. Experiment 2 examined whether N.G. would demonstrate proper-name priming despite her marked impairment at person-name free recall.

234

GEVA, MOSCOVITCH, AND LEACH Experiment 1

Method Participants. Two groups of participants were included in Experiment h a group of normal young controls (n = 12, mean age = 22 years) and a group of normal older controls (n = 12, mean age = 60 years). The older participants were matched for age and education (at least an undergraduate degree) to the prosopanomic patient N.G. The young participants were upper-year undergraduate students, receiving course credit for their participation. The older participants received monetary compensation. Materials. Person names were selected from a larger pool of 250 famous faces that N.G. recognized and had previously attempted to name. On a comparable famous faces test, normal individuals were able to retrieve the names of about 90% of the test items (E. Sekuler & M. Behrmann, personal communication,April 1993). The person names were highly familiar to Canadians (e.g., Elizabeth Taylor, Pierre Trudeau). City names were chosen from a list that N.G. had recognized and had attempted to name from an atlas (e.g., Jerusalem, Boston; for more details, see case description in the following discussion). There were three lists of person names and one list of city names. Each list corresponded to a different testing condition (person fragment completion only, person fragment completion accompanied by semantic cuing, person stem completion, and city fragment completion) and consisted of 60 items. During the study phase, 20 of the items were presented. At test, 60 items were shown, 20 of which had been previously encountered at study. Procedure. Participants were tested for person-name priming in three test conditions and city name priming in one test condition. Each condition consisted of a study phase immediately followed by a test phase. Each study list consisted of 20 person names that were randomly selected from the corresponding test list of 60 names. In each corresponding test phase, participants were presented with 60 proper-name fragments (or stems, depending on the test condition) to complete, 20 of which they had previously seen. Participants were tested individually. They were informed at study that they would be shown a list of 20 proper names. The names were shown one at a time in the center of a white screen on an Apple Macintosh Plus computer. Each name appeared on the screen for 2 s. There was a 2-s interval between adjacent items. There were two possible 20-item study lists for each test list. The study lists were counterbalanced across participants. In all four test conditions, the study phase consisted of a levels-of-processing manipulation. In the three person-name conditions, half the participants were instructed to count the number of times the letter A appeared in each name (shallow processing), while the other half categorized the person names as an actor, author, musician, politician, or athlete (deep processing). Likewise, in the city name condition, half the participants counted the number of times the letter A appeared in each name (shallow processing);

however, the remaining half categorized the city names by country (deep processing). Participants were instructed to say study responses aloud, though responses were not recorded. Each study phase was followed immediately by a test phase. At the test phase, priming for person proper names was assessed in one of three conditions: word fragment completion only, word fragment completion accompanied by semantic cuing, or stem completion only. Priming for city names was assessed in a fourth test condition that required participants to complete city word fragments. Examples of the different test items appear in Table 1. The 60 proper-name fragments were presented on the screen at a 12-s rate. In the word fragment completion task, participants were told that they would see word fragments with missing letters that corresponded to proper names (of people in the person-name condition, consisting of both the given and family name, and city names in the city name condition). Participants were asked to write down the first name that came to mind that successfully completed the fragment on the test sheets that were provided. In the fragment completion accompanied by semantic-cuingcondition, participants were asked to write the first name that came to mind that corresponded to the semantic information provided simultaneously with the presentation of the fragment. Participants in the word stem completion task were instructed to write the first name that successfully matched the first few letters of both the given name and the family name. Participants were told that they would have 12 s to solve each fragment or stem and that they should not return to old items. For each test condition, the study phase was immediately followed by its corresponding test phase. The four testing conditions were presented in a fixed order: person fragment completion only, person fragment completion accompanied by semantic cuing, person stem completion, and city fragment completion. However, the proper-name lists were presented in a counterbalanced order. Three orders of presentation were used. Each person-name list appeared either in the first test condition (fragment completion only), the second test condition (fragment completion accompanied by semantic cuing), or the third test condition (stem completion). The city name list always appeared in the last test condition (city fragment completion).

Results The results for each priming test condition are presented in Tables 2 and 3. The main findings were that (a) a reliable priming effect was found in all study conditions, with completion rates being higher for studied than nonstudied items in both age groups; (b) the amount of priming varied little as a function of age; and (c) despite the variation of level of processing at study, there was no effect of this variable on priming in fragment or stem completions. These trends were supported by subsequent statistical analysis.

Table 1

Examples of Study and Test Items Used in Different Conditions in Experiment I Condition Person name Fragment Semantic cue Stem City name Fragment

Name

Test

margaret thatcher elizabeth taylor nancy reagan

ar a e t_a_c_e_ li b t t 1o + "starred in Cleopatra" na rea

montreal

on

e 1

PERCEPTUAL PRIMING OF PROPER NAMES Table 2 Young Adults: The Percentage of ltems Completed, Lures Completed (Baseline), and the Difference Between Them (Priming)for Words Processed in a Deep or Shallow Level of Encoding Condition

Primed M SD

Lures

M

Difference

SD

M

SD

Shallow processing Person name Fragment Semantic cue Stem City name Fragment

48 78 70 63

21 12 14

19 43 35

12 18 8

29 35 35

17 9 15

22

30

9

33

16

Deep processing Person name Fragment Semantic cue Stem City name Fragment

42 68 63

21 3 9

19 39 37

12 10 9

23 29 26

21 11 8

68

12

29

14

39

14

Note. Testing was either by fragment completion, semantic cuing plus fragment completion, and stem completion for person names, or by fragment completion for city names. A five-factor analysis of variance (ANOVA) was conducted with age (young vs. old), level of processing at study (deep vs. shallow), and study list (List A vs. List B) as between-subject factors and priming (prime vs. lure) and test condition (person fragment completion vs. fragment completion accompanied by semantic information vs. stem completion vs. city fragment completion) as within-subject factors. The dependent variable for this analysis was the difference in completion rates between primed and lure items. The combined ANOVA revealed a significant priming effect, F(1, 16) = 300.62, p < .0001, MSE = 3,3943.40, indicating that primed items were completed at a higher rate than lures. An insignificant Test Condition X Priming interaction, F(3, 48) = 1.46, p > .23, MSE = 137.47, demonstrated that priming magnitude did not vary as a function of testing condition. Another finding of interest was that no levels-of-processing effect was found on fragment or stem completion rates, F(1, 16) = 0.03, p > .86, MSE = 3.13. Moreover, age had an insignificant effect on priming, F(1, 16) = 2.43. p > .13, MSE = 273.88. We performed a power analysis to determine whether there was enough power to detect an age difference had one been present. It was revealed that with a Type I error of .05, the power of the ANOVA test was .89, indicating that the failure to find an age effect was a reliable result. No other significant main effects or interactions were found.

Summary The purpose of Experiment 1 was to demonstrate repetition priming for proper names. We found robust priming that was not influenced by levels of processing or by age. Moreover, though semantic cuing did increase the overall number of fragments completed (primes and lures), it did not

235

increase the priming rate. The present results are consistent with the hypothesis that repetition priming of proper names, like that of common nouns, involves reactivation of stored information, namely the word-form, from a presemantic perceptual system. It is likely that this process operates independently of conscious retrieval mechanisms. Because explicit tests of memory were not included in our study, we cannot assert unequivocally that this is the case. However, had explicit processes been involved, performance should have improved with depth of processing and been better in young participants than in older adults (Craik & Jermings, 1992). The absence of such effects suggests strongly that performance on the implicit test was not contaminated to any great extent by explicit memory. The magnitude of priming was equivalent across all test conditions. The person-name priming effect was equivalent for both the stem and the fragment completion tasks. Furthermore, the insignificant difference in the priming effect found between person-name fragment completion accompanied by semantic cuing and person-name fragment completion alone suggests that semantic representations and processes seem not to be implicated in such an implicit test of memory; conceptual information did not confer an advantage to previously studied items. Experiment 2

Method Having demonstrated perceptual repetition priming for proper names in normal people, we wished to determine whether it would be preserved in prosopanomia. There is reason to believe that the word-form system for proper names is intact in prosopanomia, because prosopanomic people can read and recognize written names and match them to corresponding pictures. Consequently, if Table 3 Older Adults: The Percentage of ltems Completed, Lures Completed (Baseline), and the Difference Between Them (Priming)for Words Processed in a Deep or Shallow Level of Encoding Primed

Lures

Condition

M

Person name Fragment Semantic cue Stem City name Fragment

43 78 73

20 17 18

23 55 45

64

17

38

Difference

SD M SD Shallow prbcessing

M

SD

10 20 13

20 23 28

13 14 12

13

26

13

Deep processing Person name Fragment Semantic cue Stem City name Fragment

34 74 68

15 13 17

14 42 40

12 11 10

20 32 28

11 7 16

51

20

21

13

30

13

Note. Testing was either by fragment completion, semantic cuing plus fragment completion, and stem completion for person names, or by fragment completion for city names.

236

GEVA, MOSCOVITCH, AND LEACH

repetition priming operates through this system, then the prosopanomic patient N.G. should show normal proper-name priming. Nevertheless, even if the system is intact, N.G. might be impaired if, at test, the phonology of the item cannot be accessed from the orthographic representation of the proper-name fragment. Because the priming magnitude did not vary across test conditions and no levels-of-processing effect was found with the normal controls, N.G. was assessed only on word fragments in the shallow condition. We chose to test her on fragment completion, because there was reason to believe from other literature (see Winocur, Moscovitch, & Stuss, 1996) that, compared to stem completion, priming on fragment completion was least contaminated by nonperceptual variables. Participant. The prosopanomic patient N.G. was tested. Case report. N.G. was born in England in 1938 and was educated both in England and Canada. She obtained a bachelor's degree in theology and history and a master's degree in education. Immediately before becoming ill, she had decided to return to school to obtain a doctorate in education. N.G.'s problems began shortly after returning to Canada from a vacation in England in October 1987. She felt fatigued but was able to return to work as a hospital administrator for a period of 2 days. One morning she woke with a severe bifrontal headache and felt quite ill. She was able to take herself to the emergency room of a hospital in the Toronto area, but, following her admission, her condition deteriorated over a 3- to 4-day period. She became delirious, confused, agitated, and finally lost consciousness. A computed tomography scan was read as normal, but a radioisotope scan showed an area of increased uptake in the left petrous region; a subsequent MRI scan showed a lesion in the left temporal area. Acute and convalescent cerebrospinal fluid titres were positive for herpes simplex virus antibodies. She was treated with Cytovere and Decadron and gradually recovered. Upon recovery, the neurological examination was normal except for anosmia in the left nostril. Mental status examination revealed evidence of a fluent dysphasia with significant word-finding problems both for proper and common names. Following discharge from hospital, N.G. became aware of other problems, including failure to remember recent events and frequently used phone numbers. She attempted to return to work but had difficulties assuming her responsibilities. Her speech was fluent but digressive, and she had difficulty focusing on the topic of conversation. N.G. was also more irritable, verbally aggressive, and disinhibited. She had difficulty planning and organizing her work. Her supervisor described her as having difficulty discriminating salient and important features of a job, and she was unable to prioritize tasks. She was forgetful and did not follow through on assignments. Eventually, N.G. was forced to take a permanent leave of absence. N.G.'s behavior, disposition, and competency have markedly improved since then. Although she has not returned to work, she has completed a number of courses at an art institute, she is socially appropriate, and she has settled into a role of a vigorous, active matriarch. Between December of 1987 and June of 1994, N.G. underwent neuropsychological evaluation five times. She was always very cooperative with the testing. She exhibited an acute awareness of her problems. She was friendly and talkative and initially exhibited frequent word-finding problems. However, N.G.'s last assessment indicates a significant improvement on most neuropsychological measures. The results of the assessments are given in Table 4. Our experimental tests were conducted between 1993 and 1995, when her word-finding problems were restricted primarily to proper names. Of 189 pictures from Snodgrass and Vanderwort (1980), N.G. named 166 without cues and an additional 18 with phonemic cues. She was slightly better at naming nonliving than living objects,

Table 4

Results From N.G. 's First and Two Most Recent Neuropsychological Assessments Assessment date Test

December June November 1987 1990 1995

IQ a

Verbal IQ Performance IQ Vocabulary IQ Full IQ Boston Naming Testb Verbal Fluencyc FAS Animals Wechsler Memory Scale--Revisedd California Verbal Learning Teste Trial 5 acquisition Total 5 trial acquisition Rey-Osterreith Complex Figuref Copy Immediate Delayed

96 102 11 98 14

100 125 13 111 42

125 132 16 133 45

28 7 89

46 15 104

52 16 128

2 17

15 62

13 58

35 18.5 14.5

35 24 21

----

Note.

In addition to the neuropsychological tests listed, N.G.'s performance on the Wisconsin Card Sorting Test was evaluated three times (September 1988, November 1989, and June 1990). These scores were basically equivalent each time. During her most recent assessment, she achieved six categories and made seven perseverative responses. Her error rate revealed five perseverative and five nonperseverative errors. aWechsler (1981). Verbal IQ scores represent IQ; Vocabulary IQ scores are age-corrected scaled scores, bKaplan et al. (1983). The 60-item version was used. cKaplan et al. (1983). dWechsler (1987). eDeiis,Kramer, Kaplan, and Ober (1987). fLezak(1995); Spreen and Strauss (1991). 95% versus 76%. On the short version of the Boston Naming Test (Kaplan, Goodglass, & Weintraub, 1983), she scored 45/60, which also is low for her intelligence and education. Normal people score virtually 100% correct on this test (E. Sekuler & M. Behrmann, personal communication, April, 1993). Of the 250 pictures of famous people that N.G. recognized, she could name only 90 of them. At the time the priming experiment was conducted, her condition had stabilized so that she was consistent in the people she could name and those she could not. There was no difference in her ability to name people who became famous in the remote past as compared to those who achieved fame more recently, even since her illness (but see Shallice & Kartsounis, 1993, for such an anterograde naming deficit). On comparable tests of naming, controls named 90% of the people they recognized (E. Sekuler & M. Behrmann, personal communication, April 1993). Thus, although N.G. is severely anomic for proper names (of people), she also has a slight anomia even for common names. Despite these difficulties, she could match written names of people to their pictures perfectly. On standard tests of verbal fluency conducted at the same time as the present experiment, N.G. performed in the normal or superior range (see Table 4). When fluency for familiar people was tested (she had to generate both the first and last names), her performance was extremely impaired. As Table 5 indicates, she did much better when required simply to generate first or last names that did not refer to any particular individual. She also had difficulties with other categories indicating that her anomia for proper names is not restricted only to people. Table 6 shows N.G.'s performance on tests of free recall of 10

237

PERCEPTUAL PRIMING OF PROPER NAMES Table 5

Fluency: Items Generated by N.G. in 1 Minute Category

Number of items

Per so n n a l n e s

First and last Last only First only Countries Cities Books Movies

3 19 28 19 5 1 3

items. On this test, names were read to her at a rate of one per second, and she had to repeat each one after she heard it. After the 10th name, a signal was given (a rap on the table) for her to recall all the names she could remember in any order in the next 45 s. Applying Tulving and Colotla's (1970) procedure for distinguishing items from primary and secondary memory, we note that for people N.G. cannot name from photographs, almost all the names are recalled from primary memory. Estimates of her primary memory are in the normal range (see Watldns, 1975), but her secondary memory is very defective. In contrast, for people she can name, for first and last names only that do not refer to any individual, and for countries, she recalls items from both primary and secondary memory. Fictitious names, created by combining first and last names from the single-names list, behave like real names. Despite her poor performance on free recall, N.G.'s performance on recognition is perfect when she is required to choose the names she just heard from among 10 additional distracters. The results of the tests of memory are consistent with N.G.'s performance on tests of fluency and with her anomia. Clearly, the information is available to her, but she simply cannot access its phonological representation from long-term memory. If the name is held in short-term memory, she seems able to access its phonological representation without difficulty. More will be said of this in the final discussion. Materials~ In a fashion similar to ~ e first experiment, Experiment 2 conslsted of three lists of person names and one list of city names. Each list corresponded to a different fragment completion test and consisted of 60 items. Twenty names were presented at study and 601at test. Person names were~.selected from a larger pool of 250 famous faces that N.G. had recognized and had previously attempted to name. At each study session, 't0 of the names belonged to person stimuli that N,G. could both identify andname while the remaining 10 belonged tO faces N.G. could identify but was unable to name. At each test phase, half (30) of the fragments consisted of items N.G. had both identified and named, while the other half (30) consisted of items she could only identify. City items were chosen in a similar fashion. N.G. had previously demonstrated that she could only name half (10) of the city items chosen to be presented at each study phase, although she was familiar with the remaining half. Moreover, she was only able to name half (30) of the cities corresponding to the city fragments, despite her demonstration of in-depth knowledge about the remaining 30 cities. In this manner, N.G. served as her own control for items (i.e., persons and cities) she could name and those she could not. As in Experiment 1, the study phase was immediately followed by its corresponding test phase. Procedure. N.G. was tested on fragment completion only. N.G.'s priming for person names was examined in three person fragment completion tasks. Priming for city names was assessed in

one city fragment completion test. At study, she was presented with a list of 20 proper names. The names were shown one at a time in the center of a white screen on an Apple Macintosh Plus computer. Each name appeared on the screen for 2 s. There was a 2-s interval between adjacent items. N.G. processed all study items in a shallow manner, counting the number of times the letter A appeared in both person and city names. She was instructed to say study responses aloud. Each study phase was followed immediately by a test phase. The 60 proper-name fragments, 20 of which she had studied, were presented on the screen at a 12-s rate. N.G. was told that she would see word fragments with missing letters that corresponded to proper names (of people in the three person-name conditions, given name and family name, and cities in the city name condition). She was instructed to write down the first name that came to mind that successfully completed the fragment on the provided test sheets. N.G. was told that she would have 12 s to solve each fragment and that she should not return to old items. The four tests were presented in a fixed order: three person-name fragment completion tasks followed by one city name fragment completion task.

Results Experiment 2 revealed the following: (a) a priming effect was found, with N.G.'s fragment completion rate being higher for studied than for nonstudied items; (b) N . G ? s priming rates were similar to those of the matched older controls (within 1 SD); and (c) N.G.'s completion rate for the lures varied dramatically as a function o f her ability to name them. Table 7 shows the mean percentage of fragments(across three tests) completed for both studied and nonstudied person-name items, as well as city name items. Primed and unprimed person-name items were completed with average probabilities o f .33 and .13, respectively. As indicated by Figure 1, this person-name priming rate o f .20 was identical to the older matched controls' rate. Primed city items were completed w i t h a probability of .45 compared with a .15 probability of lures. Again, this priming rate of .30 was similar to the .26 priming rate observed in the older controls. N.G. showed priming for studied items that was comparable with that o f controls, regardless of her ability to name the stimuli. In fact, N.G. showed person- and city-name priming of equal magnitude for those items she could name and for those for which she could only provide semantic information. O f the 6.6 person-name primed items comTable 6

Number of Items Recalled From Memory During Free Recall Names

Primary memory

Secondary memory

Person names Unnamable (first and last) Fictitious (first and last) Namable (first and las0 First names Last names Country

.2.7 2.5 3.0 3.3 3.3 1.5

0.6 0.3 2.5 3.8 2.8 5.5

Note. The total maximum is 10.

238

GEVA, MOSCOVITCH, AND LEACH 10o

Table 7

Percentage of Primed Items Completed, Lures Completed (Baseline), and the Difference Between Them (Priming) for N. G. Condition

Primed

Lures

Difference

Person name Fragment 1 Fragment 2 Fragment 3 M City name Fragment

35 35 30 33

18 13 8 13

17 22 22 20

45

15

30

908070

6O o

50

Note. Results are reported for three person-name fragment completion tasks and one city-name fragment completion task. The mean of the three person-name priming conditions is also reported. The values represent percentages of correctly completed fragments (primes and lures) and the difference between them. person prime personlure cityprime pleted (average across the three person-name fragment completion tasks), 50% were namable items and the remaining 50% could only be identified. Of the nine city primes completed, 55% (five) were namable and the remaining 45% (four) were not. However, her completion rate of the lures varied as a function of her ability to name them. Of the five person-name lures completed, N.G. was able to name four out of five (80%) of the items but merely identify one out of five (20%) of them. Similarly, of the six city name lures completed, N.G. demonstrated intact naming for five out of six (83%) of the items but could only identify one out of six (17%) of the items she could not name. These results are summarized in Figure 2.

Summary Experiment 2 demonstrated normal proper-name priming in prosopanomia. N.G. showed person and city priming,

lOO 9080-

city lure

Condition

Figure 2. Percentage of person and city names completed by N.G. as a function of preexperimental ability to name the items and as a function of priming. The stippled area designates items she could only identify and not name (id), and the black area designates items she could identify and name.

despite an impairment in spontaneously retrieving the very same items. The magnitude of priming demonstrated by N.G. was similar to that of controls. Higher completion rates were found for studied than for nonstudied items. Priming was comparable for items that she could name and those she could not. However, lure completion was affected by N.G.'s naming abilities. N.G. performed considerably worse on lures she could not name, indicating that partial orthographic cues alone are not sufficient to lead to naming. Instead, the cues are effective only if they reactivate a complete orthographic representation, as occurs with repetition priming, which in turn activates the appropriate phonological form. In short, the results indicate that naming that is dependent on access to presemantic orthographic representations of names is preserved in prosopanomia.

70-

Discussion

60-

~

30 20

10 0 N.G.

Older controls

Figure 1. Percentage of person names (stippled) and city names (black) completed above baseline for older controls (n = 12) and N.G.

The purpose of the present series of experiments was to determine if normal participants and a patient with prosopanomia can be perceptually primed for proper names. Experiment 1 showed that young and older normal adults can access proper names presemantically. Priming was unaffected by levels of processing and age. Based on this evidence, we predicted, and our results confirmed, that proper-name priming would be intact in prosopanomia. The results of Experiment 2 indicated that N.G. demonstrated repetition priming that was comparable with that of controls, despite her marked impairment in deliberate retrieval of person and city names. The priming observed did not vary as a function of her ability to name the stimulus; however, lure completion was affected by this variable.

PERCEPTUAL PRIMING OF PROPER NAMES

Proper-Name Priming in Young and Older Adults It is postulated that during deliberate retrieval, proper names are processed in a unique manner, as evidenced by the disproportionate forgetting rate displayed by normal participants (Cohen & Burke, 1993; Cohen & Faulkner, 1986; Jones & Rabbitt, 1994; Young et al., 1985) and the categoryspecific deficit of prosopanomia. To investigate whether differences between proper and common names exist at early levels of representation (namely, at the presemantic level of the visual word-form), we designed Experiment 1 to manipulate age and levels of processing. The findings were consistent with the idea that, at the presemantic level, proper names are treated in a manner comparable to common names. Like common names, there was no effect of semantic processing on proper-name repetition priming. Neither levels of processing at encoding nor semantic cuing at retrieval affected the priming rate. Another major finding that is consistent with the observation that semantic processing contributed little to priming is that proper-name priming was resistant to age effects, suggesting that the effects observed are not likely due to conscious recollection that typically improves with semantic processing. The failure to find an age effect in Experiment 1 is congruent with previous findings showing that, unlike explicit memory, implicit memory is relatively unaffected by this variable (Roediger & McDermott, 1993). For example, despite the well-known fact that tasks reflecting explicit memory decline with age (Light, 1991), priming remains relatively stable with aging. Young and older participants have demonstrated facilitation in a lexical decision task that used repetition priming (Moscovitch, 1982). Moreover, the magnitude of this facilitation is equal for both young and older participants (Mitchell, 1989). This pattern of relatively preserved implicit memory performance in the older participant has also been demonstrated for word fragment and stem completion tasks (Light & Singh, 1987; Light, Singh, & Capps, 1986). Even in cases in which loss of priming has been observed, it is very mild and much smaller than that on explicit tests (for a review, see Light, & LaVoie, 1993; Howard, 1988). Accordingly, the failure to find an age effect in repetition priming of proper names suggests that it involves unconscious retrieval processes. Evidence implicating presemantic access of the visual word-form in repetition priming of proper names comes from the lack of a semantic-processing effect, at either encoding or retrieval. Several studies, examining memory for common nouns, have demonstrated that tasks manipulating levels of processing at study have differential effects on implicit and explicit memory. Jacoby and Dallas (1981) compared the effects of levels of processing on implicit (word identification) and explicit (yes-no recognition) tasks. As expected, the orienting task performed at study influenced recognition memory but had no effect on implicit memory. Similar results were found by Graf and Mandler (1984) and Roediger et al. (1992). Using the identical task (e.g., stem completion) as an index of both explicit and implicit memory for common nouns, they found a depth-of-

239

processing effect on explicit tests of memory, but not on the implicit versions of the same test. Thus, perceptual priming of proper names, like priming of other common nouns, operates primarily at an unconscious presemantic perceptual level. The observation that age and levels of processing do not affect priming performance suggests that explicit memory, which is known to be influenced by these factors, does not contribute to priming of proper names. This conclusion is tempered somewhat by the absence of explicit tests of memory against which performance on the implicit tests could be evaluated directly.

Priming in Prosopanomia Though Experiment 1 showed that repetition priming for proper names accesses presemantic representations of the name, the question of whether prosopanomic patient N.G. would show sensitivity to repetition priming of proper names remained unresolved. N.G. is impaired at deliberately retrieving exemplars from this class of nouns. Such a category-specific anomia may arise due to a breakdown at any of the following levels of processing: (a) visual processing, (b) organization within the semantic system, (c) the store of names, or (d) name activation. However, it is highly improbable that the name-retrieval deficit that characterizes prosopanomia is the result of a deficit earlier in the naming system. N.G., as well as other patients described in the prosopanomia literature, show intact face perception and recognition (e.g., Carney & Temple, 1993; Semenza & Zettin, 1989). Moreover, regardless of the test cue (e.g., written name or picture), these patients can provide a great deal of information on nearly all tested items, be it city or person names (e.g., HittmairDelazer et al., 1994; Semenza & Zettin, 1988). This suggests that the semantic features defining the person or city to be named are available to the participant, despite the failure to provide the correct name. It also seems to be the case that prosopanomia is not the result of a loss of the proper names themselves; knowledge at the phonological or lexical level of the name appears to be intact. Patients with prosopanomia can retrieve appropriate semantic information to a spoken or written proper noun and have no difficulty matching proper names with the real persons or with the pictures of famous persons (or, as the case may be, to city items) in four alternating multiple choices (e.g., Fery et al., 1995; Lucchelli & de Renzi, 1992). As well, N.G. shows normal retrieval of names from auditory-verbal short-term memory, indicating that it is access to phonological representations from long-term memory that is impaired. Taken together, these studies indicate that in prosopanomia, information critical' to the naming process, such as concept identity knowledge and the representation of the name itself, remains unimpaired. Prosopanomia may, in fact, be characterized as an acute manifestation of the propername TOT state exhibited by normal participants. In the face of spontaneous name-retrieval failure, both normal participants and prosopanomics display intact conceptual knowledge of the relevant proper noun and preserved storage of the name itself. It is, therefore, conceivable that the deficit is

240

GEVA, MOSCOVITCH, AND LEACH .

one of deliberately accessing proper names from long-term memory. Intact phonological representations of the name contrast with a dramatic failure to activate the name itself. As a consequence, the following question arises: Despite the failure of prosopanomics deliberately to retrieve proper names from long-term, semantic store, can proper names be perceptually primed? Results from Experiment 2 indicate that despite the underlying deficit in prosopanomia, these patients do show proper-name repetition priming. Prosopanomic patient N.G. displayed a proper-name priming rate that was comparable with that of the matched older normal controls. Primed items were completed at a higher rate than lures, indicating that the test cues themselves were not sufficient in activating the name store, unless they corresponded to the perceptual trace of a primed item. This implies that naming that is dependent on access to presemantic orthographic representations of the name is preserved in prosopanomia. In fact, completion of primed items did not vary as a function of N.G.'s deliberate naming ability, further substantiating the claim that at the presemantic level of representation, the deficit underlying prosopanomia does not distinguish between those items that can be deliberately retrieved and those that cannot. In contrast, completion of lures was affected by naming ability. N.G. completed a larger proportion of lures for which she could provide the name than those for which she could not. This is not surprising, given that N.G,'s impairment involves an inability to retrieve voluntarily person and city names. Because the lures were not primed, N.G. had no perceptual orthographic trace on which to rely. Instead, lure completion depended on an intentional search through semantic memory, the very process for which N.G. is impaired.

The visual trace of the proper name may, in turn, activate the necessary phonology for priming output. In fact, this phonological representation may be automatically activated from the prelexical orthographic form (Ferrand & Grainger, 1992; Perfetti & Bell, 1991; Perfetti, Bell, & Delaney, 1988) and operates without reference to the impaired semantic route of phonological access.

Conclusions To our knowledge, this is the first demonstration that repetition priming for proper names is perceptually based. Semantic processing at encoding (levels of processing effect) and retrieval (semantic cuing) had no effect on priming in young and older adults. The equivalent performance of young and older adults suggests that explicit memory did not contribute to priming. The preserved priming in prosopanomia suggests that naming that is dependent on access to a presemantic perceptual representation is normal in prosopanomic patients. Equivalent levels of priming were found for those items N.G. could name spontaneously and those she could not. These findings indicate that the source of the access deficit in prosopanomia involves either access to a phonological representation of the name or a lexical representation via a semantic route. Though the deficiency in prosopanomia is selective for proper names, the nature of this deficit may be similar to that found in other naming disorders. Thus, this is also the first demonstration of preserved repetition priming in anomia of any sort.

References

Priming and the Locus of the Deficit in Prosopanomia These findings support evidence that suggests that, depending on the nature of the task (e.g., speech production v s . perception, recognition of spoken vs. written words), phonological information may be encoded from distinct representational systems (Morais & Kolinksy, 1994; Peterson, Fox, Posner, Mintun, & Raichle, 1989; Shallice, McLeod, & Lewis, 1985). In the case of word production (e.g., naming task), semantic processing is independent of the stage at which the phonological form of the word is retrieved and occurs prior to it (Dell, 1986; Levelt, 1989, 1994). Thus, the failure of patients with prosopanomia to name proper nouns deliberately may reflect a failure to access this phonological system from an intact semantic representation. In other words, there is a deficit at the level of output phonology. On the other hand, to a large extent, priming is a presemantic phenomenon that likely functions independently of phonological representations accessed from conceptual knowledge. Instead, it depends largely upon structural processing of surface feature information, that is, the visual word-form (Moscovitch et al., 1993, 1994; Tulving & Schacter, 1990; Schacter, Rapscak, Rubens, Tharan, & Laguna, 1990). It is this intact preexisting, presemantic orthographic representation that supports proper-name priming in prosopanomia.

Brennen, T., Baguely, T., Bright, J., & Bruce, V. (1990). Resolving semantically induced tip-of-the-tongue states for proper nouns. Memory and Cognition, 18, 339-347. Brown, A. S. (1991). A review of the tip-of-the-tongue experience. Psychological Bulletin, 109, 204-223. Burke, D: M., Mackay, D. G., Worthley, J. S., & Wade, E. (1991). On the tip-of-the-t~ngue: What causes word finding failures in young and older adults. Journal of Memory and Language, 30, 542-579. Carney, R., & Temple, C. M. (1993). Prosopanomia? A possible category-specific anomia for faces. Cognitive Neuropsychology, 10, 185-195. Cermak, L. S., Talbot, N., Chandler, K., & Wolbarst, L. R. (1985). The perceptual priming phenomenon in amnesia. Neuropsychologia, 23, 615-622. Cohen, G. (1990). Why is it difficult to put names to faces? British Journal of Psychology, 81, 287-297. Cohen, G., & Burke, D. M. (1993). Memory for proper names: A review. Memory, 1, 249-263. Cohen, G., & Fanlkner, D. 0984). Memory in old age: "Good in parts." New Scientist, 11, 49-51. Cohen, G., & Faulkner, D. (1986). Memory for proper names: Age differences in retrieval. British Journal of Psychology, 4, 187197. Craik, E I. M., & Jennings, J. M. (1992). Human memory. In E I. M. Craik & T. A. Salthouse (Eds.), The handbook of aging and cognition (pp. 51-110). Hillsdale, NJ: Erlbanm.

PERCEPTUAL PRIMING OF PROPER NAMES Crook, T. H., & West, R. L. (1990). Name recall performance across the adult life span. British Journal of Psychology, 81, 335-349. Delis, D. C., Kramer, J. H., Kaplan, E., & Ober, B. A. (1987). California Verbal Learning Test, Adult Version. San Antonio, TX: The Psychological Corporation. Dell, G. S. (1986). A spreading activation theory of retrieval in sentence production. Psychological Review, 93, 283-321. Ferrand, L., & Grainger, J. (1992). Phonology and orthography in visual word recognition: Evidence from masked nonword priming. Quarterly Journal of Experimental Psychology, 45A, 353372. Fery, P., Vincent, E., & Bredart, S. (1995). Personal name anomia: A single case study. Cortex, 31, 191-198. Flude, B., Ellis, A. W., & Kay, J. (1989). Face processing and name retrieval in an anomic aphasic: Names are stored separately from semantic information about familiar people. Brain and Cognition, 11, 60-72. Graf, P., & Mandler, G. (1984). Activation makes words more accessible, but not necessarily more retrievable. Journal of Verbal Learning and Verbal Behavior, 23, 553-568. Graf, P., Squire, A., & Mandler, G. (1984). The information that amnesic patients do not forget. Journal of Experimental Psychology: Learning, Memory, and Cognition, 10, 164-178. Greenbaum, J. L., & Graf, P. (1989). Preschool period development of implicit and explicit remembering. Bulletin of the Psychonomic Society, 27, 417-420. Hanley, J. R., & Cowell, E. S. (1988). The effects of different types of retrieval cues on the recall of names of famous faces. Memory and Cognition, 16, 545-565. Hittmair-Delazer, M., Denes, G., Semenza, C., & Mantovan, M. C. (1994). Anomia for people's names. Neuropsychologia, 32, 465-476. Howard, D. V. (1988). Aging and memory activation: The priming of semantic and episodic memories. In L. L. Light & D. M. Burke (Eds.), Language, memory, and aging (pp. 77-99). New York: Cambridge University Press. Jacoby, L, L., & Dallas, M. (1981). On the relationship between autobiographical memory and perceptual learning. Journal of Experimental Psychology: General, 110, 306-340. Jacoby, L. L., & Witherspoon, D. (1982). Remembering without awareness. Canadian Journal of Psychology, 32, 300-324. Jones, S, J., & Rabbitt, E M. A. (1994). Effects of age on the ability to remember common and rare proper names. The Quarterly Journal of Experimental Psychology, 47A, 1001-1014. Kaplan, E. E, Goodglass, H., & Weintraub, S. (1983). The Boston Naming Test (2nd ed.). Philadelphia: Lea & Febiger. Levelt, W. J. M. (1989). Speaking: From intention to articulation. Cambridge, MA: MIT Press. Levelt, W. J. M. (1994). The skill of speaking. In E Bertelson, E Eelen, & G. d'Ydewalle (Eds.), International perspectives on psychological science: Vol. 1. Leading theories (pp. 89-103). Hove, England: Erlbaum. Lezak, M. D. (1995). Neuropsychological assessment (3rd ed.). New York: Oxford University Press. Light, L. L. (1991). Memory and aging: Four hypotheses in search of data. In M. R. Rosenzweig & L. W. Porter (Eds.), Annual review of psychology (Vol. 42, pp. 333-376). Stanford, CA: Annual Reviews. Light, L. L. & LaVoie, D. (1993). Direct and indirect measures of memory in old age. In E Graf & M. E. J. Masson (Eds.), Implicit memory: New directions in cognition, development and neuropsychology (pp. 207-230). Hillsdale, NJ: Erlbaum.

241

Light, L. L., & Singh, A. (1987). Implicit and explicit memory in young and older adults. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13, 531-541. Light, L. L., Singh, A., & Capps, J. L. (1986). Dissociation of memory and awareness in young and older adults. Journal of Clinical and Experimental Neuropsychology, 8, 62-74. Lucchelli, E, & de Renzi, E. (1992). Proper name anomia. Cortex, 28, 221-230. Maylor, E. A. (1990). Recognizing and naming faces: Aging, memory retrieval and the tip of the tongue state. Journal of Gerentology : Psychological Sciences, 45, 215-225. McWeeny, K. H., Young, A. W., Hay, D. C., & Ellis, A. W. (1987). Putting names to faces. British Journal of Psychology, 78, 143-149. Mitchell, D. B. (1989). How many memory systems? Evidence form aging. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 31-49. Morais, J., & Kolinsky, R. (1994). Perception and awareness in phonological processing: The case of the phoneme. Cognition, 50, 287-297. Moscovitch, M. (1982). Multiple dissociations of function in amnesia. In L. S. Cermak (Ed.), Human memory and amnesia (pp. 337-370). Hillsdale, NJ: Erlbaum. Moscovitch, M. (1992). Memory and working-with-memory: A component process model based on modules and central systems. Journal of Cognitive Neuroscience, 4, 257-267. Moscovitch, M., Goshen-Gottstein, Y., & Vriezen, E. (1994). Memory without recollection: A tutorial review from a neuropsychological perspective. In C. Umilta & M. Moscovitch (Eds.), Attention and performance 15: Conscious and nonconscious information processing (pp. 619-660). Cambridge, MA: M1T Press. Moscovitch, M., & Umilta, C. 0990). Modularity and neuropsychology: Implications for the organization of attention and memory in normal and brain damaged people. In M. Schwartz (Ed.), Modular processes in dementia. Cambridge, MA: MIT Press. Moscovitch, M., Vriezen, E., & Goshen-Gottstein, Y (1993). Implicit tests of memory in patients with focal lesions and degenerative brain disorders. In H. Spinnler & E Boiler (Eds.), Handbook ofneuropsychology (Vol. 8, pp. 133-173). Amsterdam: Elsevier. Parkin, A. J., & Streete, S. (1988). Implicit and explicit memory in young children and adults. British Journal of Psychology, 79, 361-369. Perfetti, C. A., & Bell, L. (1991). Phonetic activation during the first 40 ms of word identification: Evidence from backward masking and priming. Journal of Memory and Language, 30, 473--485. Perfetti, C. A., Bell, L., & Delaney, S. M. (1988). Automatic (prelexical) phonemic activation in silent word reading: Evidence from backward masking. Journal of Memory and Language, 27, 59-70. Peterson, S. E., Fox, P. T., Posner, M. I., Mintun, M. E., & Raichle, J. (1989). Positron emission tomographic studies of the processing of single words. Journal of Cognitive Neurosciences, 1, 153-170. Roediger, H. L., & McDermott, K. B. (1993). Implicit memory in normal human subjects. In H. Spinnler & E Boiler (Eds.), Handbook ofneuropsychology (Vol. 8, pp. 63-131). Amsterdam: Elsevier. Roediger, H. L., Weldon, M. S., Stadler, M. A., & Riegler, G. H. (1992). Direct comparison of word stems and word fragments in implicit and explicit retention tests. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18, 1251-1269.

242

GEVA, MOSCOVITCH, AND LEACH

Schacter, D. L. (1990). Perceptual representation systems and implicit memory: Toward a resolution of the multiple memory systems debate. In A. Diamond (Ed.), Development and neural bases of higher cognition (pp. 543-571). New York: Annals of the New York Academy of Sciences. Schacter, D. (1992). Priming and multiple memory systems: Perceptual mechanisms of implicit memory. Journal of Cognitive Neuroscience, 4, 2A.a, 256. Schacter, D. L., & Graf, P. (1986). Effects of elaborative processing on implicit and explicit memory of new associations. Journal of

Experimental Psychology: Learning, Memory, and Cognition, 12, 432--444. Schacter, D. L., Rapscak, S. Z., Rubens, A. B., Tharan, M., & Laguna, J. (1990). Priming effects in a letter-by-letter reader depends on access to the word-form system. Neuropsychologia, 28, 1079-1094. Semenza, C., & Zettin, M. (1988). Generating proper names: A case of selective inability. Cognitive Neuropsychology, 5, 711721. Semenza, C., & Zettin, M. (1989). Evidence from aphasia for the role of pure referring expressions. Nature, 32, 678-679. Shallice, T., & Kartsounis, L. D. (1993). Selective impairment of retrieving people's names: A category specific disorder? Cortex, 29, 281-291. Shallice, T., McLeod, P., & Lewis, K. (1985). Isolating cognitive modules with the dual task paradigm: Are speech perception and production separate processes? Quarterly Journal of Experimental Psychology, 37A, 507-532. Snodgrass, J. G., & Vanderwort, M. (1980). A standardized test of 260 pictures: Norms for name agreement, image agreement, familiarity and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174-215. Spreen, O., & Strauss, E. (1991). A compendium ofneuropsychological tests. New York: Oxford University Press.

Stanhope, N., & Cohen, G. (1993). Retrieval of proper names: Testing the models. British Journal of Psychology, 84, 51-65. Tulving, E., & Colotla, V. A. (1970). Free recall of trilingual lists. Cognitive Psychology, 1, 86-98. Tulving, E., & Schacter, D. L. (1990). Priming and human memory systems. Science, 247, 301-305. Warrington, E. K., & Weiskrantz, L. (1968). New method of testing long term retention with special reference to amnesic patients. Nature, 217, 972-974. Warrington, E. K., & Weiskrantz, L. (1970). Amnesic syndrome: Consolidation or retrieval? Nature, 228, 629-630. Watkins, M. J. (1975). Inhibition in recall with extralist "cues." Journal of Verbal Learning and Verbal Behavior, 14, 294-303. Wechsler, D. (1981). Wechsler Adult Intelligence Scale--Revised. New York: The Psychological Corporation. Wechsler, D. (1987). WechslerMemory Scale--Revised. San Antonio, TX: The Psychological Corporation. Winocur, G., Moscovitch, M., & Stuss, D. T. (1996). Explicit and implicit memory in the elderly: Evidence for double dissociation involving medial temporal- and frontal-lobe functions. Neuropsychology, 10, 57-65. Yarmey, A. D. (1973). I recognize your face but I can't remember your name: Further evidence of the tip-of-the-tongue phenomenon. Memory and Cognition, 1, 287-290. Young, A. W., Ellis, A. W., & Flude, M. B. (1988). Accessing stored information about people. PsychologicaIResearch, 50, 111-115. Young, A. W., Hay, D. C., & Ellis, A. W. (1985). The faces that launched a thousand slips: Everyday difficulties and errors in recognizing people. British Journal of Psychology, 76, 495-523. Received March 25, 1996 Revision received September 17, 1996 Accepted September 18, 1996 •