Platel, H., Baron, J. C., Desgranges, B., Bernard, F., & Eustache, F. (2003). Semantic and Episodic...

7/29/2019 Platel, H., Baron, J. C., Desgranges, B., Bernard, F., & Eustache, F. (2003). Semantic and Episodic Memory of Mus…

http://slidepdf.com/reader/full/platel-h-baron-j-c-desgranges-b-bernard-f-eustache-f-2003 1/13

Semantic and episodic memory of music are subservedby distinct neural networks

Herve Platel,a,* Jean-Claude Baron,b Beatrice Desgranges,a

Frederic Bernard,a and Francis Eustachea,c

a Inserm E.0218-Universite de Caen, Gip Cyceron, Laboratoire de Neuropsychologie, CHU Cote de Nacre, Caen, Franceb Department of Neurology, University of Cambridge, Cambridge UK

c Ecole Pratique des Hautes Etudes, CNRS 8581, Universite Rene Descartes, Paris 5, France

Received 26 September 2002; revised 14 February 2003; accepted 16 May 2003

Abstract

Numerous functional imaging studies have shown that retrieval from semantic and episodic memory is subserved by distinct neural

networks. However, these results were essentially obtained with verbal and visuospatial material. The aim of this work was to determine

the neural substrates underlying the semantic and episodic components of music using familiar and nonfamiliar melodic tunes. To study

musical semantic memory, we designed a task in which the instruction was to judge whether or not the musical extract was felt as “familiar.”

To study musical episodic memory, we constructed two delayed recognition tasks, one containing only familiar and the other only

nonfamiliar items. For each recognition task, half of the extracts (targets) were presented in the prior semantic task. The episodic and

semantic tasks were to be contrasted by a comparison to two perceptive control tasks and to one another. Cerebral blood flow was assessed

by means of the oxygen-15-labeled water injection method, using high-resolution PET. Distinct patterns of activations were found. First,

regarding the episodic memory condition, bilateral activations of the middle and superior frontal gyri and precuneus (more prominent on

the right side) were observed. Second, the semantic memory condition disclosed extensive activations in the medial and orbital frontal cortexbilaterally, the left angular gyrus, and predominantly the left anterior part of the middle temporal gyri. The findings from this study are

discussed in light of the available neuropsychological data obtained in brain-damaged subjects and functional neuroimaging studies.

© 2003 Elsevier Inc. All rights reserved.

Keywords: PET; Music; Semantic memory; Episodic memory; Frontal lobes; HERA model

Introduction

Many moments of our life are associated with a song or

a particular melody. However, the neural substrates of mu-

sical perception and musical memory processes are still

little known. Although it is common belief that perceptionof music is a specific ability of the right hemisphere, studies

in brain-damaged subjects demonstrate that musical capa-

bilities are distributed in both hemispheres (Lechevalier et

al., 1995; Peretz, 1994, 2001; Platel, 2002). Thus, in the

brain there is no “center for music” but a number of neural

networks that deal with the different components of music

perception (e.g., pitch, timber, rhythm, intensity); these neu-

ral networks are very sensitive to musical training and

expertise and could differ from one subject to another (Al-

tenmuller, 2001; Schlaug, 2001). Over the past 10 years, themost relevant information relating to the musical processes

have clearly emerged from functional neuroimaging studies,

but most of these have addressed perceptual treatments

(Zatorre et al., 1992; Sergent et al., 1992b; Platel et al.,

1997; Griffiths et al., 1999). So far, very few neuropsycho-

logical, psychophysical, or imaging studies have related to

the mnemonic aspects of music, and the concept of a “mod-

ular” musical memory is little considered in the psycholog-

ical and neuropsychological literatures.

* Corresponding author. EMI-E.0218 Inserm-Universite de Caen,

U.F.R. de Psychologie, Universite de Caen, Esplanade de la Paix, 14032

Caen Cedex, France.

E-mail address: [email protected] (H. Platel).

NeuroImage 20 (2003) 244 –256 www.elsevier.com/locate/ynimg

1053-8119/03/$ – see front matter © 2003 Elsevier Inc. All rights reserved.

doi:10.1016/S1053-8119(03)00287-8



Experimental psychology and psychoacoustic investiga-

tions mainly relate to the elementary characteristics of mu-

sical perception and more rarely on the mnemonic processes

as such, while the rare studies that did so focussed on

working rather than long-term memory processes (Deutsch,

1970; Semal and Demany, 1991, 1993). Experimental in-

vestigations of the recall of well-known or new musicalparts generally concurred in showing the importance of

melodic relative to rhythmic information in both the encod-

ing and the retrieval processes for musical material (Halp-

ern, 1984; Dowling et al., 1995; Hebert and Peretz, 1997).

This supports the idea of a melodic advantage in the con-

stitution of the musical lexicon.

Findings in brain-damaged subjects suggest that musical

memory could have a different cerebral organization from

other kinds of memory, such as verbal and visual memory.

Clinical reports show that although identification and rec-

ognition of a musical piece appears to involve both hemi-

spheres, integrity of the left hemisphere is however critical

(see review in Lechevalier et al., 1995), as illustrated by a

left brain-damaged patient who exhibited impaired melodic

identification despite intact melodic discrimination (Eu-

stache et al., 1990). Interestingly, these disturbances in

melodic identification were found to be dissociated from the

language impairment observed following cerebral lesions

involving specifically the verbal semantic processes (Signo-

ret et al., 1987). In addition, the studies carried out by

Zatorre (1985) and Samson and Zatorre (1991) in patients

with right or left temporal lobectomies revealed a major role

of the right temporal areas in the recognition of unknown

melodies, suggesting that such recognition, which is not

based on a semantic strategy, is subtended by a perceptiveanalysis and a comparison of the melodies. The observation

of patient C.N. (Peretz, 1996) is particularly relevant be-

cause, several years after having sustained bilateral tempo-

ral lesions, this patient was unable to memorize new musical

tunes. Furthermore, there was a lack of priming effect for

musical material in this patient, which suggested abolished

access and encoding specific to music. For the author, this

clinical case argues for the existence of a long-term memory

subsystem specific to musical material.

Another argument in favour of a possible separate long-

term memory subsystem between language and music

comes from event-related brain potentials studies (ERPs;Besson and Schon, 2001), which show differential ERP

effects for semantic processes when subjects focus their

attention only to the lyrics or to the music of opera excerpts.

The aim of this work was to determine the neural substrates

underlying the semantic and episodic components of music

memory. The distinction between episodic and semantic

memory has become very popular since it was first proposed

by Tulving in 1972. Episodic memory is conceived as the

memory of life events linked to their spatial and temporal

context of acquisition, while semantic memory corresponds to

the memory of concepts, transcending a particular context

(Tulving, 1985, 2001). In this work, we define musical seman-

tic memory as that referring to “well-known” excerpts of music

stored in memory without it being possible to retrieve the

temporal or spatial circumstances surrounding their encounter.

Semantic memory allows us to identify or to have a strong

feeling of knowing for familiar songs or melodies. We may

name the tune (composer or performer) or just have the capac-

ity to hum or whistle the subsequent notes of a melody. Mu-sical semantic memory may represent a musical lexicon, sep-

arate of a verbal lexicon, even though strong links certainly

exist between them. Our previous PET study, which dealt with

the music components of music perception, did tackle the

semantic processing of music (Platel et al., 1997). In this

investigation, judging whether or not musical sequences were

familiar induced specific activation of the left inferior frontal

gyrus (area 47 of Brodmann) and anterior part of the left

superior temporal gyrus relative to attention to the other mu-

sical components (i.e., rhythm, pitch, and timber). However, in

their study of mental imagery for familiar melodies, Halpern

and Zatorre (1999) claimed that musical semantic memory

primarily engaged right-sided regions. These findings however

appear somewhat to conflict with the large neuroimaging lit-

erature about recall of verbal or visual material from semantic

memory (see Cabeza and Nyberg, 2000, for review), in which

preferential implication of the left temporal and prefrontal

regions is a regular feature. It may therefore be argued that the

right-sided engagement reported by Halpern and Zatorre

(1999) may reflect the nonverbal nature of the musical material

used by these authors. The neural substrates of musical seman-

tic memory therefore remain uncertain.

Episodic memory for musical information will be re-

ferred here as the capacity to recognize a musical excerpt

(whether familiar or not) for which the spatiotemporal con-text surrounding its former encounter (i.e., when, where,

and how) can be recalled. Neuroimaging studies have con-

firmed the importance of the medial temporal and frontal

lobes in the encoding and retrieval of episodic information

(Cabeza and Nyberg, 2000; Fletcher and Henson, 2001).

Mayes and Montaldi (2001) however note that functional

neuroimaging has been much less successful at confirming

the roles of the midline diencephalon and basal forebrain

structures in episodic memory tasks. Explanatory hypothe-

ses stressed both technical and methodological limitations

(Desgranges et al., 1998). Yet, other structures, such as the

anterior cingulate and the precuneus, have been found toengage during episodic memory tasks (Sanders et al., 2000).

Concerning the prefrontal lobes, the HERA model (Tulving

et al., 1994) ascribes to the left prefrontal cortex a prefer-

ential role in the encoding process of episodic material and

the recall of semantic information, while the right prefrontal

cortex would preferentially operate in the recall of episodic

information. Nyberg (1998) further proposed that the right

anterior prefrontal cortex would be involved in all memory

tasks, and the right posterior prefrontal cortex in the more

dif ficult retrieval conditions. Other studies (Kelley et al.,

1998) have suggested that lateralization of activations may

depend not so much on whether encoding or retrieval are

245 H. Platel et al. / NeuroImage 20 (2003) 244 – 256



primarily being engaged, but on whether verbal or hard-to-

verbalize materials are being processed. So far, however,

experimental data have almost exclusively concerned verbal

or visual material, and practically nothing is known about

the functional anatomy of episodic memory for music.

Based on our previous PET study (Platel et al., 1997), as

well as on the neuropsychological literature, we hypothesizethat musical semantic processes will induce activation of the

left inferior frontal and left anterior temporal areas, but

could also produce right prefrontal activations (Halpern and

Zatorre, 1999). Concerning musical episodic memory, we

anticipate activation of the classic episodic memory net-

work, namely involving the right prefrontal cortex, the an-

terior cingulate gyrus, the precuneus, and potentially also

the hippocampal regions, while whether or not additional

activation specifically related to the use of musical material

would occur cannot be firmly predicted.

Materials and methods

Subjects

Nine young healthy men were selected from a population

of university students. They were all right-handed and free

of any psychiatric or organic pathology and had normal

hearing. To avoid specific cognitive strategies related to

musical expertise, these subjects were selected so as to

belong to class I of Wertheim and Botez (1959), i.e., “Mu-

sical people without theoretical musical studies and musical

knowledge.” We purposely studied nonmusical subjects so

that our findings would generalize to the largest population,whereas musicians have specific strategies for listening to

music, which depend in particular on the played instrument.

The subjects who participated in this research were selected

on the basis of two principal criteria: first, they were to be

“common listeners” (i.e., not music lovers, who tend to

listen to a specific type of music only), and second, they

were to have normal performances in a test of pitch percep-

tion. The subjects selected for the PET study were very

similar to the subjects who took part in the selection of the

familiar and nonfamiliar melodies; thus, the general musical

culture of these two groups was equivalent.

They were told that the experiment in which they were totake part related to the perception of the music, but never

informed that the experiment related to musical memory, so

that they did not train themselves to memorize melodies

before the experiment. All of them gave written informed

consent prior to participation and the research protocol was

approved by the regional ethics committee. The study was

performed in line with the Declaration of Helsinki.

Nature of the musical material

Musical material was specially created for this study and

comprised 128 short melodies (5 s) played without orches-

tration, but with the same timber of instrument (flute). All

were real melodies, extracted from the classic and modern

repertoires, but excluding songs so as to limit verbal asso-

ciations. We also excluded extracts which might spontane-

ously evoke autobiographical memories, such as the “wed-

ding march” or melodies used in popular TV spots. The 128

tunes comprised 64 “familiar” and 64 “nonfamiliar” tunes.The familiar melodies were those judged very familiar by

more than 70% of subjects in a pilot study of 150 subjects

matched with the experimental sample,1 while the nonfa-

miliar melodies were those judged as unknown by more

than 80% of the subjects from the same population. Both

samples were extracted from a larger database of melodies

that was used in this preexperimental study.

Paradigm

In the musical semantic memory task, the instruction was

to classify the melody heard as familiar or nonfamiliar. Half of the stimuli were “familiar” based on the above criteria

(target items), and half were “nonfamiliar.” According to

our design (see below), this semantic judgment will involve

incidental encoding of the stimuli, and the subjects will be

asked to recall the latter in the subsequent episodic memory

tasks. We deliberately eschewed explicit encoding in our

paradigm so as to avoid the subjects from doing a double

task, namely, memorizing the items while judging their

familiarity. To highlight activations specific to musical se-

mantic memory, a perceptive control condition was de-

signed where decisional and motor processes were present.

The instruction was to indicate if the last two notes of each

sequence had the same pitch. Since by necessity the seman-

tic task contained both familiar and nonfamiliar items, two

distinct control tasks had to be constructed, one containing

only familiar and the other only nonfamiliar melodies. None

of the melodies used in these two perceptual control tasks

were employed in the other conditions. So as to equalize

attentional involvement, these control tasks were made to be

dif ficult, with hit rates ranging from 60% to 70% in preex-

perimental studies.

As already explained above, musical episodic memory

was studied by asking the subjects to recognize, among

distractors, melodies (either familiar or nonfamiliar) that

1 Some examples of selected familiar melodies: L. W. Beethoven,

excerpts from “Symphony Nos. 6 and 9”; D. Brubeck, “Take Five”; M.

Mussorgsky, “Pictures at an exhibition”; A. Vivaldi, excerpts from “The

Seasons”; P.I. Tchaikovsky excerpts from “Nutcracker Suite”; J. Williams,

“Raiders March”; S. Prokofiev, excerpts from “Peter and the Wolf ”; A.

Dvorak, “Symphony No. 9, from the New World”; M. Ravel, “Daphnis et

Chloe”; W.A. Mozart, “Eine Kleine Nachtmusik ”; Vangelis, “Conquest of

Paradise”; E. Grieg, excerpts from “Peer Gynt Suite.” Some examples of

selected nonfamiliar melodies: L. W. Beethoven, excerpts from “Corolian”

overture, “Piano sonata No. 14”; W.A. Mozart, “Salzburg symphony No.

1”; R. Wagner “Tannhausser Overture”; C. Debussy, “Syrinx”; Vangelis,

“Antartica”; P.I. Tchaikovsky, excerpts from “Swan Lake”; B. Smetana,

“The Moldau”; F. Schubert, excerpts from “Symphony No. 6 and 8.”




they heard earlier, during the semantic task. Because, in

contrast with the semantic task, there was no obligation for

the episodic task to mix familiar and unfamiliar items and to

be consistent with the control tasks, two episodic tasks were

constructed (i.e., using nonfamiliar and familiar items, to be

referred to as F and NF in what follows). In each, half of theitems were targets (previously heard melodies), and half

were distractors. Thus, both episodic tasks had to be given

after the semantic task.

Brain activations specifically related to musical episodic

retrieval were to be assessed by comparing the episodic

tasks (NF ϩ F) to the control tasks (NF ϩ F), as for the

semantic task. In addition, separating the familiar and non-

familiar items was meant to enable us to assess the effect of

“familiarity,” by comparing each episodic task (NF or F) to

the corresponding control task (NF or F) and each episodic

condition to one another (episodic NF vs. F, and F vs. NF).

To highlight possible automatic semantic processes duringthe episodic and control tasks with familiar melodies (F), we

contrasted these conditions with the nonfamiliar episodic or

control conditions (NF). Moreover, we directly compared

episodic (NF ϩ F) versus semantic and semantic versus

episodic (NF ϩ F) conditions, and finally the pattern of

activations for control tasks (NF ϩ F) was revealed by

comparing these control conditions with rest measurements.

Altogether therefore, five tasks were constructed, i.e.,

one semantic, two episodic, and two control tasks, with, as

stated above, the semantic task always coming before the

episodic tasks. Because our preexperimental studies had

shown that the recall performance for the familiar items was

significantly better than that for nonfamiliar items, (ϳ80

and ϳ70%, respectively), we partly counterbalanced this

difference by giving the episodic task with nonfamiliar

items before that with familiar items, so as to lengthen the

retention interval for the familiar items. In addition, the

items of the semantic tasks of each version were refreshedbetween the scans, before the episodic tasks (see Fig. 1).

In all five conditions, the subjects were instructed to

respond by pressing with their right index finger the “yes”

or “no” button of a computer mouse. Each task lasted 2 min

and consisted of eight target and eight nontarget items. Each

item was Ӎ 5 s long, with an interstimulus interval of 3 s.

Each subject carried out each task twice and thus two

versions of each task, using different items, were con-

structed. The order of the two versions of each task was

counterbalanced. Finally, two “rest” scans (eyes closed and

without auditory stimulation) were acquired. Altogether,

therefore, 12 PET scans were acquired per subject (i.e., 10activation tasks and two rest measurements). With a total of

nine subjects, we thus collected 18 scans per condition. The

entire sequence of scans is illustrated in Fig. 1.

Data acquisition and analysis

Data acquisition

Measurements of regional distribution of radioactivity

were performed with an ECAT HRϩ (Siemens) PET cam-

era with full-volume acquisition allowing the reconstruction

Fig. 1. Experimental paradigm.




of 63 planes. Transmission scans were obtained with a 68Ga

source prior to emission scans. The duration of each scan

was 90 s. About 7 mCi of H2

O15 was administered as a slow

bolus in the left antecubital vein by means of an automatic

infusion pump. Each experimental condition was started

30 s before data acquisition and continued until scan com-

pletion. This process was repeated for each of the 12 scans,for a total injected dose of ϳ80 mCi. The interval between

injections was 6 min 40 s. Subjects were scanned while

lying supine in a darkened and quiet room. The head was

gently immobilized in a dedicated head rest. Head position

was aligned transaxially to the orbitomeatal line with a laser

beam. The position of the head was controlled with the laser

beam prior to each injection.

Data analysis

All calculations and image transformation were per-

formed on Unix System work stations. First, the 12 scans of

each subject were realigned to each other, using the AIR3.0

software. For subsequent data analysis, the Statistical Para-

metric Mapping software (SPM99, Wellcome Department

of Cognitive Neurology) implemented in the MATLAB

environment was used. The images were nonlinearly trans-

formed into standard space, MRI template of SMP99, which

is in Talairach space. The images were smoothed using a

12-mm Gaussian filter. The images were scaled to an overall

CBF grand mean of 50 ml/100 g/min; we therefore refer to

“adjusted rCBF” in what follows. We used a gray matter

threshold of 80% of the whole brain mean; and covariates

were centered before inclusion in the design matrix. An

ANCOVA (analysis of covariance), using global activity asa confounding covariate, was performed on a pixel-by-pixel

basis. The results of t statistic (SPM {t }) were then trans-

formed into a normal standard distribution (SPM { z}). The

significance cutoff was set at P Ͻ 0.05, cluster-level-cor-

rected for multiple comparisons.

Anatomical/cytoarchitechtonic localization of the signif-

icant activations was based on the SPM99 MRI template

and Talairach’s coordinates, which were obtained using M.

Brett’s linear transforms (see http://www.mrc-cbu.cam-

.ac.uk/Imaging/mnispace.html). All the coordinates listed in

the sections below are SPM99 coordinates.

Results

Behavioral data

The average performance across the nine subjects was

consistent with that obtained in the preexperimental popu-

lation, with nearly 70% hit rate in the control tasks, more

than 80% in the semantic task, and 89% in the episodic task

with familiar melodies (Fig. 2). As expected, the perfor-

mance for the episodic task with nonfamiliar melodies (66%

hit rate) was significantly lower than that for the familiar

melodies (P Ͻ 0.001). However, this result is particularly

related to the bad performance of one subject of the group

who did not succeed in recognizing more than two nonfa-

miliar items, whereas his performance for the familiar items

was accurate. If this subject is discarded, the average per-

centage of success becomes 73%, consistent with our stan-

dards. Mean false alarms was rather small for all targetitems of the semantic and episodic tasks (Ͻ20%); as ex-

pected, false alarms on nontargets items were commoner

(mean 31%) for episodic tasks with nonfamiliar melodies.

At debriefing, no subject expressed to have been aware of

the dichotomy of the tasks according to the material (i.e.,

familiar or not). More precisely, the subjects stated not to

have clearly realized that there were only familiar or un-

known melodies for the control as for the episodic tasks.

They did not try to make particular effort to maintain in

memory the material presented. There was no significant

improvement of the performance in the course of time.

PET data

Comparing the semantic with the control tasks (NF ϩ F)

showed extensive activation of the medial frontal areas (BA

11 and 10) bilaterally, left middle temporal gyrus (extend-

ing into the inferior frontal gyrus), and left angular gyrus

(Table 1 and Fig. 3). Activation of these areas was not

observed in the right hemisphere.

Comparing the episodic (NF ϩ F) with the control

(NF ϩ F) tasks revealed bilateral activation of the precu-

neus (BA 7), the middle frontal gyri (BA 9 and 10), and the

medial surface of the frontal lobe at the junction between

BA9 and the anterior cingulate cortex. Although bilateral,these activations were stronger on the right hemisphere

(Table 1 and Fig. 4).

Comparing the episodic (NF ϩ F) with the semantic

tasks revealed activation of the precuneus (BA 7; right side

predominantly), right superior frontal gyrus (BA 11), and

right middle frontal gyrus (BA 8 and 9). All of these acti-

vations were clearly right-sided (Table 1 and Fig. 5).

Comparing the semantic with the episodic tasks (NF ϩ

F) revealed activation of the bilateral medial frontal cortex

(BA 11/10), left inferior and bilateral middle temporal gyrus

(BA 20/21), and right cerebellum (Table 1 and Fig. 6).

Comparing the two control tasks (NFϩ

F) with restrevealed extensive bilateral activation mainly of the lateral

temporal areas (middle and superior temporal gyri), bilat-

eral inferior frontal gyrus, and cerebellum (Table 1 and Fig.

7). The temporal activations included the primary and sec-

ondary auditory cortices and were particularly conspicuous

on the right hemisphere. Activation of the cerebellum, on

the other hand, was more pronounced on the left side.

Comparing the control tasks (F) with control tasks (NF)

revealed activations of left middle and inferior frontal gyrus

(BA 11/47), left precentral gyrus (BA 6, including Broca),

and left medial frontal cortex (BA 8) (Table 1 and Fig. 8).

Comparing the episodic (F) with the control (F) tasks


http://www.mrc-cbu.cam.ac.uk/Imaging/mnispace.html

http://www.mrc-cbu.cam.ac.uk/Imaging/mnispace.html



revealed activation of the right precuneus (BA 7) and rightsuperior frontal gyrus (BA 10) only (Table 1 and Fig. 9).

Comparing the episodic (NF) with the control (NF) tasks

revealed bilateral activation of the superior and middle

frontal gyri (BA 10) bilateraly and of the medial frontal

cortex (BA 8 and 9) (Table 1 and Fig. 10).

Finally, no significant activation was found for the control

(NF) versus control (F) contrast and for the two direct com-

parison between the episodic tasks (NF vs. F or F vs. NF).

Discussion

The activation pattern of the control perceptive tasks (con-

trol NF ϩ F vs. rest) is an awaited result and consolidates the

idea of a right temporal preferential contribution in the percep-

tion of melody and pitch judgment (Zatorre et al., 1994). In

addition to temporal activations, bilateral activation of the

intrasylvian (opercular) inferior frontal areas is observed. As

shown in Fig. 7 (axial cuts), these frontal activations do not

reach the outer surface of the cortex and are not observed in the

comparisons relating to the memory tasks. These frontal acti-

vations could correspond to working memory processes or

contribute to the perceptual analysis of the melodies (Fletcher

and Henson, 2001). Activations of the cerebellum reflect par-

ticularly the motor control, and these were more accentuated inthe left hemisphere because the task response was carried out

with the right-hand fingers. The performance for these control

tasks indicates that these tasks were dif ficult. However, they

could be performed, according to debriefing, without a feeling

of failure, consistent with our goal. Accordingly, compared to

the rest condition, the pattern of activation mainly reflected an

auditory perceptual activity. Consequently, we consider that

our control tasks were appropriate perceptual control condi-

tions to serve as reference for memory tasks.

Compared to the control tasks, the activation patterns

observed for the semantic and episodic tasks were clearly

independent (Figs. 3 and 4). Only a very limited overlap of

activation may be considered in the medial frontal regions.

These distinct and new results, obtained here with musical

material, confirm the significant functional independence

between semantic and episodic memory processes, already

documented with verbal and visuospatial material (Cabeza

and Nyberg, 2000; Mayes and Montaldi, 2001).

Semantic memory

During the semantic task, activation was found to be

extensive in two distinct brain regions: first, the medial

frontal region (involving mainly BA 11 and 10) and, sec-

Fig. 2. Hit rates for the memory tasks and the control tasks: Semantic, semantic tasks carried out during PET scans; Episodic NF, episodic tasks with

non-familiar melodies; Episodic F, episodic tasks with familiar melodies; Control NF, perceptive/control tasks with nonfamiliar melodies; Control F,

perceptive/control tasks with familiar melodies.




Table 1

Activation results

t ( z) Cluster size x, y, z

Semantic vs. control (NF ؉ F)

Bilateral medial frontal cortex (BA 11/10) 7.03 (6.29) 6197 0, 59, 08

6.14 (5.62) Ϫ4, 17, Ϫ16

6.00 (5.51) 6, 15,Ϫ18Left middle temporal gyrus (BA 21) 6.14 (5.61) 2932 Ϫ55,Ϫ9,Ϫ15

5.43 (5.05) Ϫ57,Ϫ27,Ϫ2

Left inferior frontal gyrus (BA 47) 5.54 (5.14) Ϫ38, 13,Ϫ16

Left angular gyrus (BA 39) 4.93 (4.63) 220 Ϫ51,Ϫ59, 29

Episodic (NF ؉ F) vs. control (NF ؉ F)

Right precuneus (BA 7) 5.96 (5.47) 782 2,Ϫ62, 33

Right middle frontal gyrus (BA 10) 5.67 (5.24) 1825 30, 58, Ϫ3

4.97 (4.67) 34, 62, 11

Left middle frontal gyrus (BA 10) 5.23 (4.89) 964 Ϫ34, 56,Ϫ8

4.17 (3.99) Ϫ28, 62, 22

Medial frontal cortex (BA 9) 3.99 (3.83) 439 2, 36, 22

Episodic (NF ؉ F) vs. semantic

Right precuneus (BA 7/BA 19) 4.57 (4.71) 637 36, Ϫ66, 38

Precuneus (BA 7) 4.50 (4.27) 373 4,Ϫ56, 42

Right superior frontal gyrus (BA 11) 4.80 (4.53) 558 34, 52, Ϫ14Right middle frontal gyrus (BA 8/9) 4.35 (4.14) 450 38, 12, 44

Semantic vs. episodic (NF ؉ F)

Bilateral medial frontal cortex (BA 11/10) 6.32 (5.75) 3252 0, 60, 10

5.39 (5.02) Ϫ4, 18,Ϫ18

Right cerebellum 4.59 (4.35) 484 32, Ϫ84,Ϫ38

Right middle temporal gyrus (BA 21) 4.54 (4.31) 363 56, 4,Ϫ24

Left inferior and middle temporal gyri (BA 20/21) 3.84 (3.69) 544 Ϫ48,Ϫ26,Ϫ22

3.79 (3.65) Ϫ54,Ϫ2,Ϫ18

Control (NF ؉ F) vs. rest

Right superior temporal gyrus (BA 22) 12.37 (inf) 3922 63, Ϫ21, 5

9.6 (inf) 53, Ϫ10,Ϫ3

Left superior temporal gyrus (BA 22) 7.54 (6.65) 2788 Ϫ59,Ϫ17, 3

6.73 (6.06) Ϫ51,Ϫ2,Ϫ7

Left cerebellum 7.11 (6.34) 6688 Ϫ38,Ϫ67,Ϫ20

6.90 (6.19) Ϫ36,Ϫ60,Ϫ26Right cerebellum 6.09 (5.58) 32, Ϫ73,Ϫ18

Left thalamus 6.76 (6.09) 3314 Ϫ14,Ϫ17, 17

Left caudate 5.75 (5.31) Ϫ30, 25, 02

Left inferior frontal gyrus (BA 44) 5.48 (5.09) Ϫ44, 3, 20

Right inferior frontal gyrus (BA 45) 4.91 (4.62) 772 40, 22, 10

4.36 (4.15) 34, 20, 17

Right superior frontal gyrus (BA 11) 4.82 (4.54) 403 22, 56,Ϫ16

4.74 (4.48) 26, 49,Ϫ23

Control (F) vs. control (NF)

Left middle and inferior frontal gyrus (BA 11/47) 5.52 (5.13) 875 Ϫ38, 50, Ϫ20

4.00 (3.83) Ϫ42, 24, Ϫ8

Left precentral gyrus (BA 6, Broca area) 4.84 (4.56) 1157 Ϫ42, 0, 36

4.33 (4.12) Ϫ54, 12, 6

Medial frontal cortex (BA 8) 4.37 (4.16) 403 Ϫ10, 36, 40

Episodic (F) vs. control (F)

Right precuneus (BA 7) 5.45 (5.07) 551 5,Ϫ67, 33

Right superior frontal gyrus (BA 10) 4.58 (4.34) 270 28, 57, 0

Episodic (NF) vs. control (NF)

Right superior and middle frontal gyri (BA 10) 5.18 (4.85) 1242 32, 59,Ϫ16

4.37 (4.16) 38, 57, 6

Left superior and middle frontal gyri (BA 10) 5.08 (4.77) 961 Ϫ32, 54, Ϫ8

4.26 (4.06) Ϫ24, 57, 10

Medial frontal cortex (BA 8/9) 4.95 (4.65) 704 10, 33, 48

4.34 (4.13) Ϫ4, 33, 33

Note. Areas significantly activated at PϽ 0.001, but using the PϽ 0.05 corrected for multiple comparisons (cluster level) cutoff. Anatomical localization

of the significant activations and approximate Brodmann’s areas was as described under Materials and methods. Stereotaxic coordinates shown are those listed

in SPM99.




Fig. 3. Semantic tasks versus control tasks (NF ϩ F). Significantly activated regions at the threshold of P Ͻ 0.05 corrected for multiple comparisons,

displayed with surface rendering. Illustrated here is the relative contribution of the different conditions of our paradigm, according to the “effects of interests”

for selected peaks. The contrast was centred around zero, and the ordinate of each plot is the mean size of the effect for each condition Ϯ SEM, within the

peak voxel.

Fig. 4. Episodic tasks (NF ϩ F) versus control tasks (NF ϩ F). Same as Fig. 3.




ond, the left hemisphere involving an extensive band of the

middle temporal gyrus (BA 21) and extending into the

inferior frontal gyrus (BA 47). Further to these two main

activation foci, an isolated activation of the left angular

gyrus was noted. The semantic versus episodic contrast

(Fig. 6), however revealed an additional right middle tem-

poral gyrus activation (BA 21), suggesting a contribution of

these right-sided areas in semantic memory processes. This

activation profile, mainly lateralized to the left hemisphere,

is entirely consistent with previous functional neuroimaging

of semantic memory tasks (Cabeza and Nyberg, 2000;

Fletcher and Henson, 2001) and also with the HERA model

(Tulving et al., 1994). The diagrams showing the effects of

interests (Fig. 3) clearly document no noticeable participa-

tion of the two control and episodic conditions in the finding

of activation in these medial frontal and left middle tempo-

Fig. 5. Episodic tasks (NF ϩ F) versus semantic tasks. Significantly activated regions at the threshold of P Ͻ 0.05 corrected for multiple comparisons,

displayed with surface rendering.

Fig. 6. Semantic tasks versus episodic tasks (NF ϩ F). Significantly activated regions at the threshold of P Ͻ 0.05 corrected for multiple comparisons,

displayed with surface rendering.

Fig. 7. Control tasks (NF ϩ F) versus rest. Significantly activated regions at the threshold of P Ͻ 0.05 corrected for multiple comparisons displayed with

surface rendering (left) and illustrative cuts (right) of the SPM99 T1 weights MRI template.

Fig. 8. Control tasks (F) versus control tasks (NF). Signi ficantly activated regions at the threshold of P Ͻ 0.05 corrected for multiple comparisons, displayed

with surface rendering.




ral regions. On the other hand, there was a substantial level

of activity of these regions during rest, suggesting a heter-

ogeneous cognitive activity during this condition among the

subjects and justifying the importance of designing control

conditions that strongly engage attention, as purposely im-

plemented for this paradigm.

Taken together, these data confirm a functional asymme-

try in favour of the left hemisphere for semantic memory

tasks. Regarding the medial frontal activation, similar find-

ings have been reported occasionally with semantic tasks

(Sergent et al., 1992a; Kapur et al., 1994; Vandenberghe etal., 1996). According to the review of Cabeza and Nyberg

(2000), this activity would be specifically linked to the

process of categorization of semantic information. We

therefore consider that in our study, the medial frontal

activation would specifically reflect the categorization pro-

cess (i.e., the decision on familiar vs. nonfamiliar melodies).

For example, the results obtained by Tempini et al. (1998)

are in close agreement with the present investigation. Their

study, in which the association of names and familiar faces

(famous people) was studied, revealed an activation of the

medial frontal areas (BA 10 and 11), left anterior temporal

areas (BA 21), and angular gyrus. The activation of the leftangular gyrus may contribute to this categorization process

and also possibly to attempts in verbal labeling of the heard

melodies. An activation of the left angular gyrus has also

been observed previously in tasks of verb or word genera-

tion (Frith et al., 1991; Warburton et al., 1996).

Thus, activation of neither the medial frontal areas nor

the left angular gyrus appears to be very specific to musical

semantic memory. In contrast, could the activation of the

left middle temporal and inferior frontal regions character-

ize the access to a musical semantic memory? This latter

finding replicates that we obtained previously in a task of

judgment of musical familiarity (Platel et al., 1997). Fur-

thermore, activation of the left inferior frontal gyrus, al-

though less extensive than in our earlier study, was obtained

here with a different musical material and experimental

paradigm. Additional elements support the hypothesis that

this activation points to cortical regions that underlie musi-

cal semantic memory. One meta-analysis of activation

peaks from PET studies relating to the perception of lan-

guage and music (data from Petersen et al., 1988, and

Sergent et al., 1992b) revealed an absence of complete

overlap of activations (Drury and Van Essen, 1997). Thus,

the perception of simple sounds, melodies, or timber pro-duces mainly temporal and prefrontal activations that only

partially overlap those obtained for the perception of pho-

nemes, logatoms, or real words (Petersen et al., 1988; Ser-

gent et al., 1992b; Demonet et al., 1994; Zatorre et al., 1992;

Platel et al., 1997), and activation peaks obtained with

musical material appeared more anteriorly located than for

linguistic material. We suggest therefore that the left ante-

rior temporal and inferior frontal regions subtend specific

nonverbal auditory processing abilities and that they might

underlie a musical lexicon.

Why should there be a left hemisphere specialization for

musical semantic memory? The prevalent idea that the righthemisphere preferentially underlies the perception of music

arises primarily from experimental studies (in particular

with the dichotic listening technique) using nonfamiliar

melodies (Peretz, 1994). Meanwhile, the few neuropsycho-

logical case studies in which a specific impairment of mu-

sical identification was present (in the absence of perceptual

deficit) almost exclusively concerned left hemisphere le-

sions (Dupre and Nathan, 1911; Souques and Baruk, 1930;

Eustache et al., 1990). The few functional neuroimaging

studies that have focused on musical memory (Zatorre et al.,

1994; Matteis et al., 1997; Holcomb et al., 1998; Halpern

and Zatorre, 1999) showed mainly bilateral, but preferen-

Fig. 9. Episodic tasks (F) versus control tasks (F). Comparison for only familiar melodies. Significantly activated regions at the threshold of P Ͻ 0.05

corrected for multiple comparisons, displayed in SPM99 standard cuts (sagittal, coronal, and transverse).

Fig. 10. Episodic tasks. (NF) versus control tasks (NF). Comparison for only nonfamiliar melodies. Signi ficantly activated regions at the threshold of P Ͻ

0.05 corrected for multiple comparisons, displayed in SPM99 standard cuts (sagittal, coronal, and transverse).




tially right-sided, frontal and anterior temporal activations.

However, apart from the work of Halpern and Zatorre

(1999), these studies did not address musical semantic

memory, but rather recognition processes for nonfamiliar

melodies. In the Halpern and Zatorre study (1999), the

comparison of the second task (to imagine the continuation

of familiar melodies) with the third task (mental repetitionof new melodies) was designed to remove the functional

activity related to auditory analysis, working memory, and

mental imagery and to reveal only the recovery processes in

semantic memory. This comparison revealed activation in

the right inferior and middle frontal gyri, the right inferior

and superior temporal gyri, the right anterior cingulate, and

the parietal cortex. Activation of the left inferior and middle

frontal gyri was also observed, though with less intensity

than on the right side. According to Halpern and Zatorre, the

observed asymmetry in favor of the right hemisphere would

be related to the musical specificity of the cognitive pro-

cesses. There might however be some dif ficulty in the in-

terpretation of this finding, given a contamination by epi-

sodic processes, acknowledged by the authors in their

paradigm. Indeed, in contrast to our study, their subjects

were selected initially so as to have a strong feeling of

familiarity with the chosen musical extracts, and moreover,

each subject was trained to imagine the continuation of

these familiar melodies before the scan sessions. In contrast,

our paradigm was specifically designed for the study of

musical semantic memory, and we believe that the prefer-

ential left-sided activation pattern obtained for the semantic

tasks does not only reflect verbal processing during perfor-

mance of this task. Indeed, all the familiar melodies were

selected so as not to be popular songs (see Materials andmethods) and to be dif ficult to name for nonmusicians. For

example, an extract like the “Toccata and Fugue in D

Minor” from J.S. Bach would be easily judged as familiar

by virtually all nonmusicians subjects, but only few of them

would be able to produce the title and/or the composer of

this musical piece.

Thanks to dichotomizing the familiar and nonfamiliar

material in our design, it was also possible, in addition to the

semantic condition, to assess the brain activity involved in

semantic memory during the control tasks with familiar

melodies. In other words, contrasting the control tasks with

familiar versus nonfamiliar melodies makes it possible tohighlight the cerebral areas dedicated to the implicit access

to musical semantic memory. This contrast (Fig. 8) revealed

significant differences in the left prefrontal areas (in partic-

ular BA 6), as well as the left middle, inferior, and medial

frontal regions. Thus, a left dominance was again present

for this contrast, strengthening the idea that mainly left

cerebral areas support the access to musical semantic mem-

ory. However, activations observed in this comparison

could also reflect processes of humming, if subjects invol-

untarily sung (covertly) those melodies that were familiar to

them.

In agreement with Scott et al. (2000), we think that the

right posterior superior temporal regions play a specific part

in the dynamic treatment of pitch variations (for music and

language) and that the left hemisphere homologous areas

are more particularly dedicated to the phonological treat-

ment and the comprehension of language. However, based

on the present work and our earlier study (Platel et al.,

1997), we believe that the anterior part of the left temporalcortex is particularly involved in nonverbal semantic pro-

cesses and could sustain musical semantic representations.

Overall, therefore, the left middle temporal activations ob-

served in our semantic task would mainly represent the

access to musical semantic memory while the frontal acti-

vations would represent essentially the process of categori-

zation of the presented melodies.

Episodic memory

Recognition of the familiar and nonfamiliar melodies

heard at the time of the semantic task induced a profile of

activation clearly different from that obtained for the se-

mantic task. The observed activation of the rostral-most part

of the middle frontal areas (BA 9 and 10) and the precuneus

(BA 7), although bilateral, predominantly involved the right

hemisphere (Fig. 4). Predominantly right-sided frontal and

prefrontal activation has been frequently reported with re-

trieval of either verbal or nonverbal material (Cabeza and

Nyberg, 2000). Comparing the episodic versus semantic

tasks (Fig. 5) confirmed first the right dominance for the

retrieval processes and second that our musical semantic

memory task mobilized no evident resource in the right

hemisphere given that subtracting the semantic processes

from the episodic tasks did not decrease or remove theseactivations of the right hemisphere.

Regarding activation of the precuneus, although previous

functional imaging studies on perception and processing of

musical material have reported activation near this area

(Sergent et al., 1992b; Platel et al., 1997), similar findings

have concerned nonmusical material, and thus this finding

has been interpreted as reflecting a process of mental im-

agery triggered by the particular task (Kosslyn et al., 1997).

However, precuneus activation has also been observed in

episodic memory tasks, regardless of whether or not the

items possessed imageable characteristics (Krause et al.,

1999). According to Kapur et al. (1995), this cerebral regionwould be particularly involved in episodic retrieval and

more precisely in the success of episodic recall. This inter-

pretation would fit well with our results, as our musical

material did not involve particularly imageable features.

Further arguments in favor of this interpretation would be

the lack of activation of the precuneus in our semantic task

despite use of the same musical material as for the episodic

task and the fact that, at debriefing, no subject mentioned

having employed a specific mental representation strategy

for episodic recall, whether the melodies were familiar or

not. In addition, the separate comparisons between the ep-

isodic and the control tasks for familiar (Fig. 9) and nonfa-




miliar melodies (Fig. 10) show that the recognition of the

familiar melodies is associated with strictly right-lateralized

activations (precuneus and superior frontal regions),

whereas the recognition of nonfamiliar melodies produces

bilateral activations of the frontal areas (medial, superior,

and middle). Since our subjects could easily recognize the

familiar melodies, these data would support the idea that theprecuneus activation was linked with the success of epi-

sodic retrieval. On the other hand, the significant engage-

ment of the frontal regions during the recognition of the

nonfamiliar melodies reflects undoubtedly the dif ficulties

experienced by the subjects to achieve this task and suggests

the occurrence of executive control processes (such as

maintenance, planning, and inhibition). Nolde et al. (1998)

also proposed that the extent of left prefrontal activations

during retrieval probably increases as the executive de-

mands of retrieval increase.

Overall, the results of the present investigation would be

in agreement with the HERA model proposed by Tulving

and collaborators (1994), which is based on functional

asymmetry in favor of the left hemisphere for semantic

memory research and right hemisphere dominance for epi-

sodic retrieval. Despite the fact that the HERA model re-

mains controversial owing to conflicting results from some

functional neuroimaging studies, the review of the literature

carried out by Cabeza and Nyberg (2000) highlighted the

fact that the majority of the published data reported such

more or less marked asymmetry (Blanchet et al. 2001). The

bilaterality of frontal activations observed in many studies

using widely different paradigms does however mitigate the

notion of a strict right-sided predominance for episodic

retrieval. Some works suggest notably that lateralization of activations may not depend so much on whether encoding

or retrieval are primarily being engaged, but more on

whether verbal or hard-to-verbalize materials are being pro-

cessed (Kelley et al., 1998; Mayes and Montaldi, 2001).

This hypothesis fits poorly with our results, showing more

left-sided engagement of frontal regions for episodic re-

trieval of hard-to-verbalize (nonfamiliar) melodies and clear

right-sided activations with easier retrieval of familiar mel-

odies. Our feeling, which concurs with that of Nolde et al.

(1998), is that the extent of left prefrontal activations during

retrieval probably reflects the executive demands.

To conclude, these new findings from a study of seman-tic and episodic memory of musical material are altogether

consistent with the results established earlier with verbal or

visuospatial material. However, some functional specificity

for musical memory does appear to exist. Based on our data,

the left anterior temporal cortex would appear particularly

involved in semantic memory for musical material.

Acknowledgments

We thank the staff of the MRC Cyclotron Unit for scanning

facilities. This work was supported by Inserm U.320.

References

Altenmuller, E., 2001. How many music centers are in the brain? in:

Zatorre, R.J., Peretz, I. (Eds.), The Biological Foundations of Music.

Ann. N.Y. Acad. Sci. 930, 273–280.

Besson, M., Schon, D., 2001. Comparison between language and music, in:

Zatorre, R.J., I. Peretz, I. (Eds), The Biological Foundations of music.

Ann. N.Y. Acad. Sci. 930, 232–258.Blanchet, S., Desgranges, B., Denise, P., Lechevalier, B., Eustache, F.,

Faure, S., 2001. New questions on the hemispheric encoding/retrieval

asymmetry (HERA) model assessed by divided visual-field tachisto-

scopy in normal subjects. Neuropsychologia 39, 502–509.

Cabeza, R., Nyberg, L., 2000. Imaging cognition. II. Empirical review of

275 PET and fMRI studies. J. Cogn. Neurosci. 12, 1 – 47.

Demonet, J.F., Price, C., Wise, R., Frackowiak, R.S.J., 1994. Differential

activation of right and left posterior sylvian regions by semantic and

phonological tasks: a positron emission tomography study in normal

human subjects. Neurosci. Lett. 182, 25–28.

Desgranges, B., Baron, J.C., Eustache, F., 1998. The functional neuroanat-

omy of episodic memory: the role of the frontal lobes, the hippocampal

formation, and others areas. NeuroImage 8, 198 –213.

Deutsch, D., 1970. Tones and numbers: specificity of interference in

immediate memory. Science 168, 1604 –1605.

Dowling, W.J., Kwak, S., Andrews, M.W., 1995. The time course of

recognition of novel melodies. Percept. Psychophys. 57, 136 –149.

Dupre, E., Nathan, M., 1911. Le langage musical, in: Etude medico-

psychologique, Alcan, Paris, pp. 64 –75.

Drury, H.A., Van Essen, D.C., 1997. Functional specializations in human

cerebral cortex analyzed using the visible man surface-based atlas.

Hum. Brain Map. 5, 233–237.

Eustache, F., Lechevalier, B., Viader, F., Lambert, J., 1990. Identification

and discrimination disorders in auditory perception: a report on two

cases. Neuropsychologia 28, 257–270.

Fletcher, P.C., Henson, R.N.A., 2001. Frontal lobes and human memory:

insights from functional neuroimaging. Brain 124, 849 – 881.

Frith, C.D., Friston, K., Liddle, P.F., Frackowiak, R.S.J., 1991. A PET

study of word finding. Neuropsychologia 29, 1137–1148.Grif fiths, T.D., Johnsrude, I., Dean, J.L., Green, G.G., 1999. A common

neural substrate for the analysis of pitch and duration pattern in seg-

mented sound. NeuroReport 18, 3825–3830.

Halpern, A., 1984. Organization in memory for familiar songs. J. Exp.

Psychol. Learn. Mem. Cogn. 3, 496 –512.

Halpern, A., Zatorre, R.J., 1999. When that tune runs through your head:

a PET investigation of auditory imagery for familiar melodies. Cereb.

Cortex 9, 697–704.

Hebert, S., Peretz, I., 1997. Recognition of music in long-term memory: are

melodic and temporal patterns equal partners. Mem. Cogn. 25, 518 –

533.

Holcomb, H., Medoff, D., Caudill, P., Zhao, Z., Lahti, A., Dannals, R.,

Tamminga, C., 1998. Cerebral blood flow relationships associated with

a dif ficult tone recognition task in trained normal volunteers. Cereb.

Cortex 8, 534 –542.Kapur, S., Craik, F.I.M., Jones, C., Brown, G.M., Houle, S., Tulving, E.,

1995. Functional role of the prefrontal cortex in retrieval of memories:

a PET study. NeuroReport 6, 1880 –1884.

Kapur, S., Rose, R., Liddle, P.F., Zipursky, R.B., Brown, G.M., Stuss, D.,

Houle, S., Tulving, E., 1994. The role of left prefrontal cortex in verbal

processing: semantic processing or willed action. NeuroReport 5,

2193–2196.

Kelley, W.M., Miezin, F.M., McDermott, K.B., Buckner, R.L., Raichle,

M.E., Cohen, N.J., et al., 1998. Hemispheric specialization in human

dorsal frontal cortex and medial temporal lobe for verbal and non-

verbal memory encoding. Neuron 20, 927–936.

Kosslyn, S.M., Thompson, W.L., Alpert, N.M., 1997. Neural systems

shared by visual imagery and visual perception: a positron emission

tomography study. NeuroImage 6, 320 –334.




Krause, B.J., Schmidt, D., Mottaghy, F., Taylor, J., Halsband, U., Herzog,

H., Tellmann, L., Muller-Gartner, H.W., 1999. Episodic retrieval acti-

vates the precuneus irrespective of the imagery content of word pair

associates: a PET study. Brain 122, 225–263.

Lechevalier, B., Platel, H., Eustache, F., 1995. Neuropsychologie de

l’identification musicale. Rev. Neurol. 151, 505–510.

Matteis, M., Silvestrini, M., Troisi, E., Cupini, L.M., Caltagirone, C., 1997.

Transcranial Doppler assessment of cerebral flow velocity during per-

ception and recognition of melodies. J. Neurol. Sci. 149, 57– 61.

Mayes, A., Montaldi, D., 2001. Exploring the neural bases of episodic and

semantic memory: the role of structural and functional neuroimaging.

Neurosci. Biobehav. Rev. 25, 555–573.

Nolde, S.F., Johnson, M.K., Raye, C.L., 1998. The role of prefrontal cortex

during tests of episodic memory. Trends Cogn. Neurosci. 2, 399 – 406.

Nyberg, L., 1998. Mapping episodic memory. Behav. Brain Res. 90,

107–114.

Peretz, I., 1994. Les agnosies auditives. in: Seron, X., Jeannerod, M. (Eds.),

Neuropsychologie. Humaine, Mardaga, Liege, pp. 205–216.

Peretz, I., 1996. Can we lose memory for music? A case of music agnosia

in a non musician. J. Cogn. Neurosci. 8, 481– 496.

Peretz, I., 2001. Brain specialization for music: new evidence from con-

genital amusia, in: Zatorre, R.J., Peretz, I. (Eds.). The Biological

Foundations of Music. Ann. N.Y. Acad. Sci. 930, 152–165.Petersen, S.E., Fox, P.T., Posner, M.I., Mintun, M., Raichle, M.E., 1988.

Positron emission tomographic studies of the cortical anatomy of single

word processing. Nature 331, 585–589.

Platel, H., 2002. Neuropsychology of musical perception: new perspec-

tives. Brain 125, 223–224.

Platel, H., Price, C., Baron, J.C., Wise, R., Lambert, J., Frackowiak, R.S.J.,

Lechevalier, B., Eustache, F., 1997. The structural components of

music perception: a functional anatomical study. Brain 120, 229 –243.

Samson, S., Zatorre, R.J., 1991. Recognition memory for text and melody

of songs after unilateral temporal lobe lesion: evidence for dual encod-

ing. J. Exp. Psychol. Learn. Mem. Cogn. 17, 793– 804.

Sanders, A.L., Wheeler, M.E., Buckner, R.L., 2000. Episodic recognition

modulates frontal and parietal cortex activity. J. Cogn. Neurosci. 50a.

Scott, S.K., Blank, C.C., Rosen, S., Wise, R.J., 2000. Identification of a

pathway for intelligible speech in the left temporal lobe. Brain 123,2400 –2406.

Semal, C., Demany, L., 1991. Dissociation of pitch from timbre in auditory

short-term memory. J. Acoust. Soc. Am. 89, 2404 –2410.

Semal, C., Demany, L., 1993. Further evidence for an autonomous pro-

cessing of pitch in auditory short-term memory. J. Acoust. Soc. Am.

94, 1315–1322.

Sergent, J., Otah, S., MacDonald, B., 1992a. Functional neuroanatomy of

face and object processing. Brain 115, 15–36.

Sergent, J., Zuck, E., Terriah, S., MacDonald, B., 1992b. Distributed neural

network underlying musical sight-reading and keyboard performance.

Science 257, 106 –109.

Schlaug, G., 2001. The brain of musicians: a model for functional and

structural adaptation, in: Zatorre, R.J., Peretz, I., (Eds.), The Biological

Foundations of Music. Ann. N.Y. Acad. Sci. 930, 281–299.

Signoret, J.L., Van Eeckout, P., Poncet, M., Castaigne, P., 1987. Aphasie

sans amusie chez un organiste aveugle. Rev. Neurol. 143, 172–181.

Souques, A., Baruk, H., 1930. Autopsie d’un cas d’amusie (avec aphasie)

chez un professeur de piano. Rev. Neurol. 4, 545–557.

Tempini, M.L., Price, C., Josephs, O., Vandenberghe, R., Cappa, S., Kapur,

N., Frackowiak, R.S.J., 1998. The neural systems sustaining face and

proper-name processing. Brain 121, 2103–2118.

Tulving, E., 1985. Memory and consciousness. Can. Psychol. 26, 1 –11.

Tulving, E., 2001. Episodic memory and common sense: how far apart?

Philos. Trans. R Soc. Lond. B Biol. Sci. 356, 1505–1515.

Tulving, E., Kapur, S., Craik, F.I.M., Moscovitch, M., Houle, S., 1994.

Hemispheric encoding/retrieval asymmetry in episodic memory:positron emission tomography findings. Proc. Natl. Acad. Sci. USA 91,

2016 –2020.

Vandenberghe, R., Price, C., Wise, R., Josephs, O., Frackowiak, R.S.J.,

1996. Functional anatomy of a common semantic system for words and

pictures. Nature 383, 254 –256.

Warburton, E., Wise, R., Price, C., Weiller, C., Frackowiak, R.S.J., 1996.

Noun and verb retrieval by normal subjects studies with PET. Brain

119, 159 –179.

Wertheim, N., Botez, M.I., 1959. Plan d’investigation des fonctions mu-

sicales. Encephale 48, 246 –255.

Zatorre, R.J., 1985. Discrimination and recognition of melodies after uni-

cerebral excisions. Neuropsychologia 23, 31– 41.

Zatorre, R.J., Evans, A.C., Meyer, E., 1994. Neural mechanisms underly-

ing melodic perception and memory for pitch. J. Neurosci. 14, 1908 –

1919.Zatorre, R.J., Evans, A.C., Meyer, E., Gjedde, A., 1992. Lateralization of

phonetic and pitch discrimination in speech processing. Science 256,

846 – 849.


Platel, H., Baron, J. C., Desgranges, B., Bernard, F., & Eustache, F. (2003). Semantic and Episodic...

Documents

Transcript of Platel, H., Baron, J. C., Desgranges, B., Bernard, F., & Eustache, F. (2003). Semantic and Episodic...