NEUROPSYCHOLOGICAL AND NEUROIMAGING DATA PERCEPTION OF TONAL AND TEMPORAL STRUCTURES IN CHORD SEQUENCES BY PATIENTS WITH CEREBELLAR DAMAGE

Transcription

1 Pitch and Time Perception in Cerebellar Patients 271 PERCEPTION OF TONAL AND TEMPORAL STRUCTURES IN CHORD SEQUENCES BY PATIENTS WITH CEREBELLAR DAMAGE GÉRALDINE LEBRUN-GUILLAUD & BARBARA TILLMANN Université Claude Bernard Lyon I and CNRS-UMR 5020, Lyon, France TIMOTHY JUSTUS VA Northern California Health Care System, Martinez, California OUR STUDY INVESTIGATED THE PERCEPTION of pitch and time dimensions in chord sequences by patients with cerebellar damage. In eight-chord sequences, tonal relatedness and temporal regularity of the chords were manipulated and their processing was tested with indirect and direct investigation methods (i.e., priming paradigm in Experiment 1; subjective judgments of completion and temporal regularity in Experiments 2 and 3). Experiment 1 replicated a musical relatedness effect despite cerebellar damage (see Tillmann, Justus, & Bigand, 2008) and Experiment 2 extended it to completion judgments. This outcome suggests that an intact cerebellum is not mandatory to access tonal knowledge. However, data on temporal manipulations suggest that the cerebellum is involved in the processing of temporal regularities in music. The comparison between task performances obtained for the same sequences further suggests that the altered processing of temporal structures in patients impairs the rapid development of musical expectations on the time dimension. Received August 5, 2007, accepted October 31, Key words: cerebellar damage, pitch and time, musical expectations, priming, tonal knowledge NEUROPSYCHOLOGICAL AND NEUROIMAGING DATA have shown that the cerebellum is involved not only in motor control and movement but also in higher cognitive functions such as implicit learning, memory, attention, and temporal processing (see Desmond, 2001; Ivry & Spencer, 2004; Justus & Ivry, 2001 for reviews). For auditory perception, neuroimaging studies have demonstrated increased cerebellar activation for a variety of tasks, such as frequency, intensity, and duration discrimination (Belin et al., 2002), as well as processing of complex nonspeech sounds (Vouloumanos, Kiehl, Werker, & Liddle, 2001) and musical structures (e.g., Parsons, 2001; Tillmann, Janata, & Bharucha, 2003). Our study investigated the involvement of the cerebellum in the perception of two main structural dimensions of Western tonal music: pitch and time. The pitch dimension refers to structural regularities concerning the use of the twelve chromatic tones (e.g., their frequency of occurrence and co-occurrence). Nonmusician listeners become sensitive to these regularities by mere exposure to musical pieces obeying this system, and the implicitly acquired tonal knowledge influences their perception of music (Bigand & Poulin-Charronnat, 2006; Francès, 1958; Krumhansl, 1990; Tillmann, Bharucha, & Bigand, 2000). Neuroimaging research has shown the involvement of the cerebellum in tonal structure processing. The reported data do not reflect hemispheric specialization of the cerebellum, but its involvement has been reported respectively in right and left cerebellum as well as bilaterally. Cerebellar activation has been observed not only for music production (i.e., singing or piano performance, Bengtsson & Ullen, 2006; Brown, Martinez, & Hodges, 2004; Brown, Martinez, & Parsons, 2006; Callan et al., 2006; Jeffries, Kritz, & Braun, 2003; Parsons, Sergent, Hodges, & Fox, 2005) but also for music perception (Callan et al., 2006; Pallesen, Brattico, Bailey, Korvenoja, Koivisto, Gjedde, & Carlson, 2005; Parsons, 2001; Rauschecker, 2005; Satoh, Takeda, Nagata, Hatazawa, & Kuzuhara, 2003; Tillmann et al., 2003). For example, increased cerebellar activation was observed when listeners had to detect errors in melodic or harmonic structures of musical pieces (Parsons, 2001) and for unexpected musical events (i.e., less important in the tonal hierarchy; Tillmann et al., 2003). However, recent behavioral data for patients with cerebellar lesions suggest that an intact cerebellum is not mandatory for accessing implicit musical knowledge and for its influence on perception (Tillmann et al., 2008). Despite damage to the cerebellum, patients processed musical structures and showed a musical priming effect Music Perception VOLUME 25, ISSUE 4, PP , ISSN , ELECTRONIC ISSN BY THE REGENTS OF THE UNIVERSITY OF CALIFORNIA. ALL RIGHTS RESERVED. PLEASE DIRECT ALL REQUESTS FOR PERMISSION TO PHOTOCOPY OR REPRODUCE ARTICLE CONTENT THROUGH THE UNIVERSITY OF CALIFORNIA PRESS S RIGHTS AND PERMISSIONS WEBSITE, DOI: /MP

2 272 Géraldine Lebrun-Guillaud, Barbara Tillmann, & Timothy Justus as did healthy college students and matched controls: chord processing was faster when the chord was musically related (and expected based on tonal knowledge) than when it was less related (and less expected). Our study aimed to further investigate tonal structure processing of cerebellar patients by replicating the musical relatedness effect with the priming paradigm and extending it to subjective judgments. This extension tested whether musical structure processing can be observed not only with implicit testing methods (i.e., the priming paradigm), but also with explicit ones. Recently, the musical priming paradigm revealed some preserved implicit knowledge in a case of severe amusia, while explicit music processing (i.e., as measured with subjective completion judgments) was impaired (Tillmann, Peretz, Bigand, & Gosselin, 2007). The time dimension in music refers to the organization of event-onset intervals leading to rhythmic patterns and to the sensation of meter (i.e., regular succession of strong and weak beats), with isochronous sequences being the simplest version of these regularities. Listeners are sensitive to temporal regularities and develop expectations about when future events will occur (e.g., Fraisse, 1974; Jones, 1976; Povel, 1981). Temporal expectations lead to processing advantages for regular, metrical sequences over nonmetrical (or weakly metrical) ones in production (e.g., Essen & Povel, 1985; Patel, Iversen, Chen, & Repp, 2005), perception (e.g., Large & Jones, 1999; Tillmann & Lebrun-Guillaud, 2006; Yee, Holleran, & Jones, 1994) and memory (e.g., Bharucha & Pryor, 1986; Hébert & Cuddy, 2002). Our study investigated the perception of regular, isochronous musical sequences (contrasted to irregular sequences) by cerebellar patients. These temporal manipulations were strong and unmusical, but represent a first attempt to study whether cerebellar damage impairs processing of temporal structures in musical material. The involvement of the cerebellum in temporal processing has been shown for production and perception of speech and nonspeech auditory stimuli (Ackermann, Graber, Hertrich, & Daum, 1997; Ivry & Keele, 1989; Ivry & Spencer, 2004). Beyond impairment in tapping tasks, for example, cerebellar patients were impaired in judging the duration of auditory stimuli, while still being able to judge their intensity (Ivry & Keele, 1989). Similarly, Casini and Ivry (1999) reported impaired performance of cerebellar patients for duration judgments, but not for pitch judgments with the same stimuli. The involvement of the cerebellum in auditory tasks requiring temporal processing has been also observed in neuroimaging studies (see Petacchi, Laird, Fox, & Bower, 2005 for a review). Increased cerebellar activation was reported for discrimination of durations (Belin et al., 2002), of temporal intervals (Jueptner et al., 1995), and of rhythm, tempo, and meter in musical excerpts, as well as for the memory of familiar and unfamiliar rhythms (Parsons, 2001). As for data on the pitch dimension, no specific lateralization pattern of cerebellar involvement emerged, neither from patient nor from imaging data. Our study investigated the perception of pitch and time dimensions in chord sequences by patients with cerebellar damage. In eight-chord sequences, tonal relatedness and temporal regularity of the chords were manipulated (see Figure 1). The last chord of the sequences was either tonally related (i.e., functioning as a tonic chord) or less related (i.e., functioning as a subdominant chord) (Bigand, Poulin, Tillmann, & D Adamo, 2003; Bigand, Tillmann, Poulin, D Adamo, & Madurell, 2001). The first seven chords of the sequences were either played in a regular, isochronous way or with various irregular patterns (Tillmann & Lebrun-Guillaud, 2006). The perception of these manipulations was tested with three tasks. Experiment 1 used the musical priming paradigm to investigate the influence of tonal and temporal structures on the speed of chord processing. Experiments 2 and 3 asked participants to judge the musical sequences by using subjective scales evaluating FIGURE 1. (Top) An example pair of the experimental sequences for the related condition and less related condition. (Bottom) A schematic representation of the temporal manipulations with regular and irregular sequences. (Adapted from Tillmann & Lebrun-Guillaud, 2006, Figure 1)

3 Pitch and Time Perception in Cerebellar Patients 273 either the degree of completion (Experiment 2) or the degree of temporal regularity (Experiment 3). These tasks contrast direct versus indirect measures to investigate the involvement of the cerebellum in pitch and time processing in music. For various neurological disorders (e.g., amnesia, aphasia, prosopagnosia), indirect investigation methods have provided evidence for spared implicit processes in the presence of severe impairments in tasks requiring explicit processing. Experiment 1: Musical Priming Paradigm The musical priming paradigm is an implicit investigation method that tests the influence of tonal and temporal structures on musical event processing (see Tillmann, 2005 for a review). Without judging tonal and temporal manipulations, participants make judgments on a perceptual aspect of target chords, such as sensory consonance (e.g., Bharucha & Stoeckig, 1986), timbre identification (e.g., Tillmann, Bigand, Escoffier, & Lalitte, 2006; Tillmann & Lebrun-Guillaud, 2006) or phoneme identification in sung music (e.g., Bigand et al., 2001). Priming data have shown that listeners develop tonal and temporal expectations (i.e., what kind of tonal event should occur when) that lead to facilitated processing for expected targets occurring at expected time points. For example, processing is faster for tonally related tonic targets than for less related subdominant targets, and faster with regular than with irregular sequences (Tillmann & Lebrun-Guillaud, 2006). Experiment 1 tested cerebellar patients, matched controls, and student controls for tonal and temporal expectations with a timbre identification task (i.e., is the target played by Timbre A or Timbre B). Based on priming data with cerebellar patients for sung music (Tillmann et al., 2008), we predicted faster response times for related targets than for less related targets. Based on reported deficits (or at least alterations) of temporal feature processing for cerebellar patients (Ivry, Keele, & Diener, 1988; Malapani, Dubois, Rancurel, & Gibbon, 1998; Molinari, Leggio, DeMartin, Cerasa, & Thaut, 2003), weaker or no differences were expected for regular versus irregular sequences for cerebellar patients in comparison to matched controls and student controls. Method PARTICIPANTS Six patients with cerebellar dysfunction/lesions and six matched controls (see Table 1 and Figure 2) participated in this study after having given written informed consent. Four patients had bilateral cerebellar degeneration (B10, B9, B8, B3), one patient a left hemisphere focal lesion (L4), and one patient a right hemisphere focal lesion (R1). In addition, Experiment 1 was run with 15 students (mean age: 23.3 years; instrumental practice: 3.0 years ±3.8). MATERIAL Manipulations on pitch and time dimensions were realized in 8-chord sequences (from Bigand et al., 2003; for more details see Tillmann & Lebrun-Guillaud, 2006). For the pitch dimension, the musical relatedness between prime context (first seven chords) and target (eighth chord) was modified: sequences ended either on related targets (tonic) or less related targets (subdominant). For the time dimension, the chords were played either regularly (isochronously) or irregularly (with jittered interchord intervals). 1 The last chord was played on time as expected in regular sequences. 2 Prime chords were played by an acoustic piano sound and targets by a guitar sound (Timbre A) or a harp sound (Timbre B). The experiment was run on Psyscope software (Cohen, MacWhinney, Flatt, & Provost, 1993). PROCEDURE Participants were asked to decide as quickly and as accurately as possible whether the last chord of each sequence was played by Timbre A or Timbre B by pressing one of two keys on the computer keyboard. Error feedback was given. A short random tone sequence was presented after each response to empty sensory memory and avoid tonal carryover effects from trial to trial. Pressing the space bar started the next trial. Short breaks were imposed and participants had the option of taking additional breaks. Each participant judged 96 sequences presented in random order. Results Mean accuracy was 91% for matched controls (ranging from 69% to 100% for Timbre A and from 79% to 98% for Timbre B) and 85% for patients (ranging from 79% to 1 For regular sequences, the onset times for the 8 chords were 0 ms, 640 ms, 1280 ms, 1920 ms, 2560 ms, 3200 ms, 3840 ms, and 4480 ms. For irregular sequences, 12 different patterns of onset times for the 8 chords were used, for example: 0 ms, 520 ms, 1240 ms, 1680 ms, 2000 ms, 2720 ms, 3840 ms, and 4480 ms. Each chord was played for 320 ms, giving the sequences a staccato-like sound. 2 In addition to the influence of harmonic relatedness and regularity, Tillmann and Lebrun-Guillaud (2006) studied targets occurring on time, earlier, and later than expected. Experiment 1 selected only the opposition between regular and irregular sequences, with all targets occurring on time.

4 274 Géraldine Lebrun-Guillaud, Barbara Tillmann, & Timothy Justus TABLE 1. Cerebellar Patients Participating in Experiments 1, 2 and 3. WAIS** Patients Aetiology Age Education Handedness Sex Music ICARS* VIQ PIQ FSIQ WM Index L4 Left hemisphere Right M 4 yrs, percussion stroke (age 48) R1 Right hemisphere Right M 4 yrs, piano, violin stroke (age 66) B10 Bilateral degeneration Right M None (unknown, age 21-) B9 Bilateral degeneration Right F 4 yrs, piano (unknown, age 53-) B8 Bilateral degeneration Right M 2.5 yrs, accordion (unknown, age 42-) B3 Bilateral degeneration Right M None (unknown, age 61-) Mean values ± standard deviation: (±13.2) (±2.4) (±2) Matched Controls (mean values ± standard deviation) right 5 M 0.8 (±13.9) (±2) (minimum 0; 1 F maximum 3) (±1.3) *International cooperative ataxia rating scale (Trouillas et al., 1997). Total ICARS score ranges from 0 (no ataxia) to 100 (most severe ataxia) **Wechsler Adult Intelligence Scale with Verbal IQ (VIQ), Performance IQ (PIQ), Full-Scale IQ (FSIQ) and Working memory (WM) index.

5 Pitch and Time Perception in Cerebellar Patients % for Timbre A, except for R1: 40%, and 73% to 100% for Timbre B). Because of faster mean correct response times for matched controls (761 ms) and strong intersubject variability for patients (1024 ms to 2270 ms), absolute response times were individually normalized to z-scores with a mean of 0 and a standard deviation of 1. These z- scores (see Figure 3) were analyzed separately for Timbre A and Timbre B because a first analysis revealed interactions involving Target timbre. A series of ANOVAs were performed with Group (Patients, Matched controls) as the between-participants factor, Regularity (Regular, Irregular) and Musical Relatedness (Related, Less related) as within-participant factors, and participants (F 1 ) or items (F 2 ) as random variables. 3 For Timbre A, the interaction between Group and Regularity was significant, F 1 (1, 10) = 9.44, p <.05; F 2 (1, 22) = 15.71, p <.001. For patients, response times were faster for regular than irregular sequences, F 1 (1, 5) = 6.69, p <.05 and F 2 (1, 11) = 11.84, p <.01, but not for matched controls. For the participants analyses, the interaction between Regularity and Relatedness was marginally significant, F 1 (1, 10) = 3.91, p =.08, but the differences between related and less related targets were not significant for regular or irregular sequences (p >.28). No other effects were significant. For Timbre B, the main effect of Musical Relatedness was significant, F 1 (1, 10) = 11.59, p <.01 and F 2 (1, 22) = 5.58, p <.05, with faster response times for related than less related targets. For the participants analysis, the interaction between Musical Relatedness and Group was marginally significant, F 1 (1, 10) = 3.82, p =.08: the difference between related and less related targets was stronger for patients, F 1 (1, 5) = 31.88, p <.01, than for matched controls (n.s.). In addition, the interaction between Regularity and Group was marginally significant, F 1 (1, 10) = 3.36; p =.097: response times were faster for regular than for irregular sequences for matched controls, F 1 (1, 5) = 18.56, p <.01, but not for patients (n.s.). No other effects were significant. FIGURE 2. The location of cerebellar damage for the patients with focal lesions. The other patients had cerebellar degeneration and because of their progressive condition, the state of degeneration at the time of the study could not be precisely documented (note that cerebellar degeneration is generally bilateral and symmetrical). Seven horizontal slices through the pons and the cerebellum are shown, with the most superior slice at the top. The corresponding sections in the atlas by Schmahmann, Doyon, Toga, Petrides, and Evans (2000) are approximately: z = 9, 17, 25, 33, 41, 49, and 57. The left side of each slice corresponds to the patient s left side. Darker color indicates a lesion. For further details on these reconstructions, see recons.html STUDENT GROUP Since the data of matched controls did not replicate significantly the data of students (Tillmann & Lebrun- Guillaud, 2006), we considered whether selecting a subset of conditions might have influenced the overall data and ran an additional student group with the present experimental design. The outcome replicated the previous finding, with significant main effects of Relatedness, 3 For each participant group and timbre, the data did not violate normality as indicated by nonsignificant differences in Kolmogorov- Smirnov tests.

6 276 Géraldine Lebrun-Guillaud, Barbara Tillmann, & Timothy Justus FIGURE 3. Correct response times (z-scores) of Experiment 1 for Timbre A targets (top) and Timbre B targets (bottom) for matched controls (C) and patients (P), followed by each patient and student controls (S). The z-scores are presented as a function of regularity (regular, irregular) and musical relatedness (related, less related). Error bars indicate standard errors (over participants for groups and over sequences for each patient). For patients, abbreviations refer to lesion type: L = left hemispheric focal lesion, R = right hemispheric focal lesion, and B = bilateral degeneration.

7 Pitch and Time Perception in Cerebellar Patients 277 F 1 (1, 14) = 14.54, p <.01 and F 2 (1, 11) = 9.96, p <.01, and of Regularity, F 1 (1, 14) = 9.82, p <.01 and F 2 (1, 11) = 12.59, p <.01. In addition, response times were faster for Timbre A than Timbre B, F 1 (1, 14) = 48.29, p <.001 and F 2 (1, 11) = 33.25, p <.001, but Target Timbre did not interact significantly with Relatedness and/or Regularity. Discussion Experiment 1 investigated the perception of pitch and time dimensions in chord sequences with the priming paradigm. Student data replicated the influence of musical relatedness and temporal regularity for both target timbres (Tillmann & Lebrun-Guillaud, 2006): Processing was faster for related than for less related targets, and faster with regular than with irregular sequences. For patients and matched controls, however, the influence of the two dimensions depended on the target timbre. For Timbre B, processing was faster for related than for less related targets. Surprisingly, this effect of musical relatedness reached significance only for patients, but not for matched controls (even if mean response times were shorter for related targets). 4 For each patient, the mean differences reflected the facilitation for related targets. This outcome extends the previously shown musical priming effect in cerebellar patients (Tillmann et al., 2008) from sung phoneme identification to timbre identification and to four other cerebellar patients (L4, B10, B9, B8). The two data sets suggest that cerebellar patients process musical structures and develop tonal expectations, which influence the speed of chord processing. However, no processing advantage for regular over irregular sequences was observed for the patients. This observation suggests that patients were less able to use the regular structure to anticipate the temporal occurrence of the target precisely than were matched controls and students. In contrast to Timbre B targets, only the regularity effect was significant for Timbre A targets in patients: targets were judged faster in regular than in irregular sequences. Timbres A and B were different from the timbre used for the prime context and the faster 4 For matched controls, significance was not reached for all effects (in contrast to students). Mean RT for Timbre B reflected the influence of musical relatedness and regularity, but only the regularity effect was significant. This missing effect strength might be due to the small group size or to age differences leading to more variable response times. Consequently, we refer to deficits for cerebellar patients only when effects were significant for students and matched controls. This was the case for the regularity effect in Timbre B (which then became the focus of Experiments 2 and 3). response times for Timbre A than Timbre B 5 suggest that the timbral dissimilarity with the prime timbre was stronger for Timbre A. This more dissimilar timbre might serve as an acoustic surface marker indicating clearly when to respond, speeding up responses overall. This surface marker indicates the target without requiring listeners to rely on a precise internal clock, useful for the anticipation of the last chord. Observing a processing advantage with Timbre A targets for patients suggests the hypothesis that some internal timing mechanism (which is sensitive to regular structures) is also functioning in patients, but requires some supplementary help (i.e., the acoustic surface marker of Timbre A) to advantage processing speed. However, without this marker (as in Timbre B trials) the finding suggests a deficit or reduced capacity of cerebellar patients to take advantage of the regular structures for processing speed. In sum, patients data for Timbre B trials replicate the musical relatedness effect, but suggest a deficit in the perception of temporal regularity. However, because of the small group of cerebellar patients and the observed variability, Experiments 2 and 3 further investigated the processing of pitch and time dimensions for these Timbre B trials. Explicit tasks were used with the aim to confirm the musical relatedness effect and to investigate whether a regularity effect might be observed with explicit judgments of completion and regularity. Experiment 2: Completion Judgments Cerebellar patients and matched controls judged the degree of completion for musical sequences of Experiment 1 (ending on Timbre B). Based on previous data (Bigand & Pineau, 1997; Tillmann & Lebrun-Guillaud, 2006), related sequences were expected to be judged as more complete than less related ones. The less related subdominant induces a feeling of tension, which contrasts the related tonic inducing a feeling of completion and relaxation. Experiment 2 aimed to investigate whether for cerebellar patients the sensitivity to musical relatedness extends to explicit judgments or is restricted to the implicit priming paradigm, as observed for an amusic patient (Tillmann et al., 2007). Furthermore, the irregularity of the prime context might lead to weaker completion judgments compared to regular prime contexts. 5 Response times were faster overall for Timbre A than Timbre B in student controls (Timbre A: 642 ms; Timbre B: 719 ms), in matched controls (Timbre A: 736 ms; Timbre B: 786 ms) and in patients as a group (Timbre A: 1667 ms; Timbre B: 1810 ms) and individually for L4 (mean RT for B mean RT for A = 15 ms), B10 (= 224 ms), B9 (= 387 ms), B8 (= 218 ms) and B3 (= 28 ms), except for R1 (= 13 ms).

8 278 Géraldine Lebrun-Guillaud, Barbara Tillmann, & Timothy Justus Method PARTICIPANTS Cerebellar patients and matched controls were the same as in Experiment 1. MATERIAL AND PROCEDURE Sequences ending on Timbre B targets of Experiment 1 were used. Participants were asked to rate the degree of completion (i.e., how well the sequence ended) of each sequence by using an 8-point scale (from 1 = incomplete to 8 = very complete ). At the end of each sequence, the scale appeared on the screen. Participants verbally indicated their judgment to the investigator who pressed one of eight keys on the computer keyboard. As the participants were wearing headphones, the investigator could not hear the sequences and could not inadvertently influence their responses. Before the experimental phase, participants received four practice trials to familiarize themselves with the experimental procedure. Short breaks were indicated during the experiment and further breaks could be added. Results Completion judgments (see Figure 4, top) were analyzed by ANOVAs with Group (Patients, Matched Controls) as the between-participants factor, Regularity (Regular, Irregular) and Musical Relatedness (Related, Less related) as within-participant factors, and participants (F 1 ) or items (F 2 ) as random variables. Patient B3 opted not to complete this part of the study. He repeatedly claimed not to be able to tell any difference between completed and less completed sequences and resisted further explanations of the concept of completeness. The main effects of Regularity, F 1 (1, 8) = 6.14, p <.05 and F 2 (1, 22) = 64.05, p <.0001, and of Relatedness, F 1 (1, 8) = 20.99, p <.001 and F 2 (1, 22) = 35.70, p <.0001, were significant: sequences were judged as more complete when ending on related targets than on less related targets and when being played regularly than when played irregularly. For the item analysis, the interaction between Regularity and Group was significant, F 2 (1, 22) = 10.27, p <.01: The difference between regular and irregular sequences was stronger for matched controls, F 2 (1, 11) = 47.40, p <.0001, than patients, albeit still significant, F 2 (1, 11) = 17.05, p <.01. In addition, there was a main effect of group, F 1 (1, 8) = 3.90, p =.08 and F 2 (1, 22) = 26.50, p <.0001: patients judged sequences as being more complete than did matched controls. Discussion Experiment 2 showed that cerebellar patients and matched controls judged sequences as sounding more complete when ending on a related chord than a less related chord and when played regularly than when played irregularly. The relatedness effect was as strong for matched controls as for patients, and all mean differences indicated higher degrees of completion for related sequences (except regular sequences for B10). A possible exception is patient B3, who did not want to continue this particular task beyond the first few trials, claiming that he did not understand the difference between completed and uncompleted sequences. It could be argued that B3 s reluctance to complete this portion of the study resulted from impaired explicit processing of harmonic relations, but based on his behavior we rather suspect that B3 was largely unmotivated to do this particular task. The relatedness effect for the remaining group of cerebellar patients (i.e., those participating in this experiment) is in agreement with priming data of Experiment 1 and points out that cerebellar patients process harmonic relations also at an explicit level. In contrast to the priming data, completion judgments showed that patients were sensitive to the timing manipulation, even if the regularity effect was less strong than for matched controls. Experiment 3: Regularity Judgments Experiment 3 aimed to further investigate the perception of temporal regularity by cerebellar patients using an explicit task on the time dimension. Since these explicit judgments directly focused on the temporal regularity of the sequences, we expected stronger differences between judgments for regular and irregular sequences, which would allow us to further define deficits in the cerebellar group and/or individual cases. Method PARTICIPANTS Cerebellar patients and matched controls were the same as in Experiment 1. MATERIAL AND PROCEDURE Sequences ending on Timbre B targets of Experiment 1 were used. Participants were asked to rate the degree of regularity (i.e., how evenly paced the sequence was) of each sequence by using an 8-point scale (from 1 = irregular to 8 = very regular ). The procedure was as described in Experiment 2.

9 Pitch and Time Perception in Cerebellar Patients 279 FIGURE 4. Completion ratings (from 1 incomplete to 8 very complete) of Experiment 2 (top) and regularity ratings (from 1 irregular to 8 very regular) of Experiment 3 (bottom) for matched controls (C) and patients (P), followed by each patient. Ratings are presented as a function of regularity (regular, irregular) and musical relatedness (related, less related). Error bars indicate standard errors (over participants for groups and over sequences for each patient). For patients, abbreviations refer to lesion type: L = left hemispheric focal lesion, R = right hemispheric focal lesion, and B = bilateral degeneration. Results Regularity judgments (see Figure 4, bottom) were analyzed using ANOVAs with Group (Patients, Matched Controls) as the between-participants factor, Regularity (Regular, Irregular) and Musical Relatedness (Related, Less related) as within-participant factors, and participants (F 1 ) or items (F 2 ) as random variables. The main effect of Regularity was significant, F 1 (1, 10) = 53.68, p <.0001 and F 2 (1, 11) = , p <.0001: regular sequences were judged as being more regular than irregular sequences. For the item analysis, the main

10 280 Géraldine Lebrun-Guillaud, Barbara Tillmann, & Timothy Justus effect of Group was significant, F 2 (1, 22) = 23.93, p <.0001, and interacted with Regularity, F 2 (1, 22) = 50.61, p <.0001: the difference between regular and irregular sequences was stronger for matched controls, F 2 (1, 11) = , p <.0001, than for patients, although still significant, F 2 (1, 11) = , p < Discussion For matched controls and patients, regular sequences were judged as more regular than irregular sequences. This difference was observed for each patient, except for patient B10. The explicit temporal judgments indicated that, despite cerebellar lesions, patients perceive the regularity and irregularity of the sequences. It is important to point out that the difference between regular and irregular sequences was weaker for patients than for matched controls, mainly because patients judged irregular sequences as less irregular than controls (except for B9). General Discussion Our study investigated the perception of tonal relatedness and temporal regularity in musical sequences by cerebellar patients. Experiment 1 used an implicit investigation method with timbre identification and focused on the influence of musical expectations (developed on pitch and time dimensions) on the speed of chord processing. Experiments 2 and 3 used subjective judgments of completion and of temporal regularity and required listeners to give explicit judgments on the musical sequences. Completion judgments are influenced by both pitch structures and temporal structures. Regularity judgments focus listeners judgments on temporal structures only. For the processing of the pitch dimension, the findings suggest that cerebellar patients process musical relatedness comparably to healthy listeners, with facilitated processing for related chords (see also Tillmann et al., 2008) and stronger feelings of completion for sequences ending on related chords. Patients sensitivity to musical relatedness is thus not restricted to implicit investigation methods, but extends to an explicit task. The two studies provide converging evidence that despite damage to the cerebellum, listeners remain able to develop expectations in music based on associative/sequential relationships between tones and chords and that an intact cerebellum is not mandatory for accessing implicit knowledge stored in long-term memory and for its influence on perception. For the processing of the time dimension, however, sensitivity for temporal structures (regular versus irregular) was observed with the explicit tasks, but not with the implicit task (for the same sequences ending on Timbre B). The lack of regularity effect in the priming data cannot be attributed to a missing sensitivity of the dependent variable of this paradigm since response times revealed a processing advantage for musically related targets. In contrast to the priming data, the completion and regularity judgments show that cerebellar patients perceived the temporal manipulations in the chord sequences. However, the cerebellar impairment leads to smaller differences than those observed for matched controls: patients judged the irregular sequences as more complete and less irregular than did matched controls. The comparison of subjective judgments (on completion and regularity) with the priming data suggests that patients even if perceiving the irregularity are not using the temporal organization of the sequence to develop temporal expectations about the precise ending point of the sequence, which would then have led to faster response times for regular sequences. This finding suggests that the patients perception of temporal structures might be less accurate, less fine tuned, and/or less rapidly developed than for healthy listeners. It complements previously reported data patterns showing deficits in perception and production of temporal features in auditory material (Casini & Ivry, 1999; Ivry & Keele, 1989), even if temporal processing is not abolished by cerebellar impairment. The completion and regularity judgments of the sequences can be based on the accumulation of various deviations over time without precise temporal processing. Interestingly, the priming data for Timbre A also suggest some remaining sensitivity of cerebellar patients to regular versus irregular temporal structures, but in order for temporal expectations to influence processing speed, a salient acoustic surface marker for the target seems to be needed. Our study investigated subtle differences in musical relatedness (i.e., related and less related chords are both in-key chords), but strong, unmusical temporal manipulations contrasting regular, isochronous sequences with irregular sequences. For the same sequences (Timbre B trials), cerebellar patients seemed to have a deficit in developing temporal expectations based on isochrony or lack thereof (as suggested by the priming data), while still processing the differences in these temporal structures (as suggested by the subjective judgments). Future research using more subtle temporal manipulations in musical material (also related to more complex temporal representations involving rhythm and meter) will allow us to further specify the deficit of temporal processing and its influence on music perception.

11 Pitch and Time Perception in Cerebellar Patients 281 In addition to the patient group data, it is important to point out that the group of cerebellar patients also revealed individual differences, most particularly for the processing of the temporal manipulations. For the priming data of Timbre B, all patients showed a processing advantage of related over less related targets, but only patient B9 presented an advantage of regular over irregular sequences with both related and less related targets. For the completion task, the individual data revealed that the musical relatedness effect was present in each patient (except for B10 for regular sequences), but only patients L4 and B9 showed a regularity effect comparable in size to that of matched controls. For the temporal regularity task, a difference between regular and irregular sequences was observed in each patient (except for B10), but to varying extents. The group of cerebellar patients included two patients with lesions and four with cerebellar degeneration. For the lesion patients, the completion and regularity data show that patient L4 had a more similar pattern to matched controls than had patient R1. For patient R1, the completion judgments were influenced by musical relatedness, but not by temporal regularity, and the difference for regularity judgments between regular and irregular sequences was reduced in contrast to matched controls. A possible hypothesis about hemispheric differences (and/or lesion location) needs, however, further testing with a larger patient pool with focal lesions. For the degeneration patients, the time elapsed since diagnosis varied between patients and the differences in durations can be linked to the data patterns. Patients B3 and B9 were diagnosed more recently (i.e., 3 years prior to the study) and their data patterns for subjective judgments were close to the data patterns of matched controls. For both patients, the difference in regularity judgments between regular and irregular sequences was comparably strong to the one observed for matched controls. For patient B9 (B3 did not complete this task), completion judgments also showed a strong difference between regular and irregular sequences (even stronger than for matched controls). Interestingly, even B9 s priming data for Timbre B trials showed slower mean response times for irregular than for regular sequences. In contrast to these patients, the completion and regularity judgments for patients B10 and B8 either showed no difference or a reduced difference between regular and irregular sequences. Patients B10 and B8 were diagnosed 16 and 12 years, respectively, prior to the study, and represent the more severe cerebellar degeneration among the patients tested here. The overall data pattern thus suggests the implication of the cerebellum in temporal structure processing, even for our strong temporal manipulations: for more advanced cerebellar degeneration, impairment is observed even for the explicit tasks. Author Note This research was supported in part by the grant program Emergence of the French Rhône-Alpes Region, the Junior Research Team of the French Ministry of Research, and the National Institutes of Health grant NS Thanks to Richard Ivry and the members of his lab for providing access to the cerebellar patients. Correspondence concerning this article should be addressed to Barbara Tillmann, Université Claude Bernard Lyon I, CNRS UMR 5020, Neurosciences Sensorielles Comportement Cognition, 50 Av. Tony Garnier, F Lyon Cedex 07, France, Tel: +33 (0) , Fax: +33 (0) btillmann@olfac. univ-lyon1.fr References ACKERMANN, H., GRABER, S., HERTRICH, I., & DAUM, I. (1999). Cerebellar contributions to the perception of temporal cues within the speech and nonspeech domain. Brain and Language, 67, BELIN, P.,MCADAMS, S., THIVARD, L., SMITH, B.,SAVEL, S., ZILBOVICIUS, M., ET AL. (2002). The neuroanatomical substrate of sound duration discrimination. Neuropsychologia, 40, BENGTSSON, S.L.,& ULLEN, F.(2006). Dissociation between melodic and rhythmic processing during piano performance from musical scores. NeuroImage, 30, BHARUCHA, J.J.,& PRYOR, J.H.(1986). Disrupting the isochrony underlying rhythm: An asymmetry in discrimination. Perception and Psychophysics, 40, BHARUCHA, J.J.,& STOECKIG, K.(1986). Reaction time and musical expectancy: Priming of chords. Journal of Experimental Psychology: Human Perception and Performance, 12, BIGAND, E., & PINEAU, M.(1997). Global context effects on musical expectancy. Perception and Psychophysics, 59, BIGAND, E., POULIN, B.,TILLMANN, B.,& D ADAMO, D. (2003). Cognitive versus sensory components in harmonic

12 282 Géraldine Lebrun-Guillaud, Barbara Tillmann, & Timothy Justus priming effects. Journal of Experimental Psychology: Human Perception and Performance 29, BIGAND, E., & POULIN-CHARRONNAT, B.(2006). Are we experienced listeners? A review of the musical capacities that do not depend on formal musical training. Cognition, 100, BIGAND, E., TILLMANN, B.,POULIN, B.,D ADAMO, D.A., & MADURELL, F.(2001). The effect of harmonic context on phoneme monitoring in vocal music. Cognition, 81, B11-B20. BROWN, S., MARTINEZ, M.J.,& HODGES, D.A.(2004). The song system of the human brain. Cognitive Brain Research, 20, BROWN, S., MARTINEZ, M.J.,& PARSONS, L.M.(2006). Music and language side by side in the brain: A PET study of the generation of melodies and sentences. European Journal of Neuroscience, 23, CALLAN, D.E.,TSYTSAREV, V.,HANAKAWA, T.,CALLAN, A. M., KATSUHARA, M., FUKUYAMA, H., & TURNER, R.(2006). Song and speech: Brain regions involved with perception and covert production. NeuroImage, 31, CASINI, L., & IVRY, R.(1999). Effects of divided attention on temporal processing in patients with lesions of the cerebellum or frontal lobe. Neuropsychology, 13, COHEN, J.,MACWHINNEY, B.,FLATT, M., & PROVOST, J. (1993). PsyScope: An interactive graphic system for designing and controlling experiments in the psychology laboratory using Macintosh computers. Behavior Research Methods, Instruments and Computers, 25, DESMOND, J.E.(2001). Cerebellar involvement in cognitive function: Evidence from neuroimaging. International Review of Psychiatry, 13, ESSEN, P.J.,& POVEL, D.J.(1985). Metrical and non-metrical representations of temporal patterns. Perception and Psychophysics, 37, 1-7. FRAISSE, P.(1974). Memory of drawings and sentences depending on the duration of presentation. L Annee Psychologique, 74, FRANCÈS,R.(1958). La perception de la musique (2nd ed.). Paris: PUF. HÉBERT, S., & CUDDY, L.L.(2002). Detection of metric structure in auditory figural patterns. Perception and Psychophysics, 64, IVRY, R.B.,& KEELE, S.W.(1989). Timing functions of the cerebellum. Journal of Cognitive Neuroscience, 1, IVRY, R.B.,KEELE, S.W.,& DIENER, H.C.(1988). Dissociation of the lateral and medial cerebellum in movement timing and movement execution. Experimental Brain Research, 73, IVRY, R.B.,& SPENCER, R.M.C.(2004). The neural representation of time. Current Opinion in Neurobiology, 14, JEFFRIES, K.J.,FRITZ, J.B.,& BRAUN, A.R.(2003). Words in melody: An H-2 O-15 PET study of brain activation during singing and speaking. Neuroreport, 14, JONES, M.R.(1976). Time, our lost dimension: Toward a new theory of perception, attention, and memory. Psychological Review, 83, JUEPTNER, M., RIJNTJES, M., WEILLER, C., FAISS, J.H., TIMMANN, D.,MUELLER, S.P.,& DIENER, H.C.(1995). Localization of a cerebellar timing process using PET. Neurology, 45, JUSTUS, T.C.,& IVRY, R.B.(2001). The cognitive neuropsychology of the cerebellum. International Review of Psychiatry, 13, KRUMHANSL, C.L.(1990). Cognitive foundations of musical pitch.oxford: Oxford University Press. LARGE, E.W.,& JONES, M.R.(1999). The dynamics of attending: How people track time-varying events. Psychological Review, 106, MALAPANI, C.,DUBOIS, B.,RANCUREL, G.,& GIBBON, J. (1998). Cerebellar dysfunctions of temporal processing in the seconds range in humans. Neuroreport, 9, MOLINARI, M., LEGGIO, M.G.,DEMARTIN, M., CERASA, A., & THAUT, M.(2003). Neurobiology of rhythmic motor entrainment. Annals of the New York Academy of Sciences, 999, PALLESEN, K.J., BRATTICO, E., BAILEY, C., KORVENOJA, A., KOIVISTO, J.,GJEDDE, A., & CARLSON, S.(2005). Emotion processing of major, minor, and dissonant chords : A functional magnetic resonance imaging study. Annals of the New York Academy of Sciences, 1060, PARSONS, L.M.(2001). Exploring the functional neuroanatomy of music performance, perception and comprehension. Annals of the New York Academy of Sciences, 930, PARSONS, L.M.,SERGENT, J.,HODGES, D.A., & FOX, P.T. (2005). The brain basis of piano performance. Neuropsychologia, 43, PATEL, A.D.,IVERSEN, J.R.,CHEN, Y.,& REPP, B.H.(2005). The influence of metricality and modality on synchronization with a beat. Experimental Brain Research, 163, PETACCHI, A., LAIRD, A.R., FOX, P.T.,& BOWER, J.M. (2005). Cerebellum and auditory function: An ALE metaanalysis of functional neuroimaging studies. Human Brain Mapping, 25, POVEL, D.J.(1981). Internal representations of simple temporal patterns. Journal of Experimental Psychology: Human Perception and Performance, 7, RAUSCHECKER, J.P.(2005). Neural encoding and retrieval of sound sequences. Annals of the New York Academy of Sciences, 1060, SATOH, M., TAKEDA, K., NAGATA, K., HATAZAWA, J.,& KUZUHARA, S.(2003). The anterior portion of the bilateral temporal lobes participates in music perception: A positron emission tomography study. American Journal of Neuroradiology, 24,

13 Pitch and Time Perception in Cerebellar Patients 283 SCHMAHMANN, J.D.,DOYON, J.,TOGA,A.W.,PETRIDES, M., & EVANS, A.C.(2000). MRI atlas of the human cerebellum. San Diego, CA: Academic Press. TILLMANN, B.(2005). Implicit investigations of tonal knowledge in nonmusician listeners. Annals of the New York Academy of Sciences, 1060, TILLMANN, B.,BHARUCHA, J.J.,& BIGAND, E.(2000). Implicit learning of tonality: A self-organizing approach. Psychological Review, 107, TILLMANN, B.,BIGAND, E., ESCOFFIER, N., & LALITTE, P.(2006). The influence of musical relatedness on timbre discrimination. European Journal of Cognitive Psychology, 18, TILLMANN, B.,JANATA, P.,& BHARUCHA, J.J.(2003). Activation of the inferior frontal cortex in musical priming. Cognitive Brain Research, 16, TILLMANN, B.,JUSTUS, T.,& BIGAND, E.(2008). Cerebellar patients demonstrate preserved implicit knowledge of association strengths in musical sequences. Brain and Cognition, 66, TILLMANN, B.,KOELSCH, S., ESCOFFIER, N., BIGAND, E., LALITTE, P.,FRIEDERICI, A.D.,& VON CRAMON, D.Y. (2006). Cognitive priming in sung and instrumental music: Activation of inferior frontal cortex. NeuroImage, 31, TILLMANN, B.,& LEBRUN-GUILLAUD, G.(2006). Influence of tonal and temporal expectations on chord processing and on completion judgments of chord sequences. Psychological Research, 70, TILLMANN, B.,PERETZ, I., BIGAND, E., & GOSSELIN, N. (2007). Harmonic priming in an amusic patient: The power of implicit tasks. Cognitive Neuropsychology, 24, TROUILLAS, P.,TAKAYANAGI, T.,HALLETT, M., CURRIER, R. D., SUBRAMONY, S.H.,WESSEL, K., ET AL. (1997). International cooperative ataxia rating scale for pharmacological assessment of the cerebellar syndrome. Journal of the Neurological Sciences, 45, VOULOUMANOS, A., KIEHL, K.A., WERKER, J.F.,& LIDDLE, P.F. (2001). Detection of sounds in the auditory stream: Eventrelated fmri evidence for differential activation to speech and nonspeech. Journal of Cognitive Neuroscience, 13, YEE, W.,HOLLERAN, S., & JONES, M.R.(1994). Sensitivity to event timing in regular and irregular sequences: influences of musical skill. Perception and Psychophysics, 56,

14