Specialist musical training and the octave illusion: analytical listening and veridical perception by pipe organists

Specialist musical training and the octave illusion:analytical listening and veridical perception by

pipe organists

David Brennan, Catherine Stevens *

School of Psychology, Macarthur Auditory Research Centre, University of Western Sydney at Bankstown,

Locked Bag 1797, Penrith South DC, NSW 1797, Australia

Received 15 February 2001; received in revised form 29 June 2001; accepted 29 June 2001

Abstract

The octave illusion is a useful tool for investigation of the contribution of specialist training

to auditory perception. The stimulus that induces the illusion involves two tones with a fre-

quency ratio of 2:1, presented dichotically, and with ear of presentation reversed every 250

ms. Most listeners report hearing a single tone that alternates from high in the right ear to

low in the left ear [Scientific American 233 (1975) 92–104]. The first experiment investigated

the hypothesis that musical training contributes to veridical perception of an ambiguous stim-

ulus. As hypothesized, participants with the highest level of musical training were more likely

to perceive the stimulus veridically. Exploring the effects of specialist training, Experiment 2

contrasted expert pipe organists with other instrumentalists. As hypothesized, participants ex-

pert in playing pipe organ – an instrument with harmonic and spatial features similar to those

of the octave illusion – were more likely to perceive the stimulus veridically. The results have

implications for plasticity of the auditory system and the analytical listening that accompanies

specialist, intensive training and rehearsal. � 2002 Elsevier Science B.V. All rights reserved.

PsycINFO classification: 2326

Keywords: Auditory illusion; Auditory perception; Expertise; Musical training; Spatial perception

*Corresponding author. Tel.: +61-2-9772-6324; fax: +61-2-9772-6736.

E-mail addresses: [email protected] (D. Brennan), [email protected] (C. Stevens).

Acta Psychologica 109 (2002) 301–314

www.elsevier.com/locate/actpsy

0001-6918/02/$ - see front matter � 2002 Elsevier Science B.V. All rights reserved.

PII: S0001-6918 (01 )00063-4

1. Introduction

Deutsch’s (1975) octave illusion confronts the student of perceptual processeswith an extreme of perceptive ‘‘reality’’. The octave illusion occurs when two toneswith a frequency ratio of 2:1 (usually 800 and 400 Hz and preferably sinusoidal) arepresented dichotically, and ear-of-presentation is reversed every 250 ms. The major-ity of respondents (right-handed) report hearing a single tone that alternates fromhigh in the right ear to low in the left ear (see Fig. 1). By whatever mechanism, atone is identified by most people as emanating from the opposite headphone toits actual physical location (Deutsch, 1975, 1999; Deutsch & Roll, 1976; ten Ho-open, 1996). Although apparently quite rare, there are individuals who perceivethe stimulus veridically (in its true physical form). This study seeks to elaborateon the findings of Craig (1979) who suggests that musical training enhances veridicalperception of the stimulus. As the two-tone stimulus for the illusion contains littlemusical content it seems probable that the advantage might involve acoustic as wellas perceptual components. It has already been shown that intensive musical trainingattunes perception to tonal, rhythmic and harmonic features and structures (e.g.,Beal, 1985; Davidson, Power, & Michie, 1987; Handel, 1989; Wolpert, 1990). Thepresent study tests the assumption that specialist experience of musicians in the anal-ysis, decomposition and re-composition of sound may further attune perception toatomistic acoustic components. Specifically, we compare percepts reported by musi-cians experienced with an instrument (the pipe organ) that in some ways simu-lates the octave illusion stimulus with reports from musicians trained on otherinstruments.

Studies that compare musicians and non-musicians, their attributes and abilities,and whether such attributes are innate or learned, abound (e.g., Beal, 1985; Chew,

Fig. 1. Octave illusion stimulus (a) and common percept (b).

302 D. Brennan, C. Stevens / Acta Psychologica 109 (2002) 301–314

Larkey, Soli, Blount, & Jenkins, 1982; Howe, Davidson, & Sloboda, 1998; Tan,1979; Wolpert, 1990). They range from analysis of fundamental parameters suchas hand skill asymmetry (Jancke, Schlaug, & Steinmetz, 1997) and absolute pitch(Hurata, Kuriki, & Pantev, 1999), to more esoteric studies such as those by Hassler,Nieschlag, and De La Motte (1990) that measured the salivary testosterone levels ofcomposers. In general, these approaches attempt to identify a relationship betweenmusical skills and concomitant physical or cognitive attributes. An alternative ap-proach is to identify the effect of musical training on physical and cognitive attri-butes. One example is that of Elbert, Pantev, Weinbruch, Rockstroh, and Taub(1995) where string players were studied and increased cortical representations ofthe fingers of the left hand were noted – possibly a product of extensive practiceon the instrument. Put simply, the former approach seeks to identify those physicaland cognitive attributes that might lead to musical skill acquisition whereas the lat-ter approach studies the physical and cognitive attributes that are products of theacquisition of musical skills. In studies that use adult participants, these two ex-tremes are difficult to completely disjoin. Musically trained individuals may haveshown an initial aptitude that prompts musical study and eventual mastery of theinstrument. The confound can be minimized by sampling expert and non-expertpopulations from a range of ages, by matching musicians on level of training ratherthan age, and by comparing effects of similar amounts of training on different instru-ment types.

With this qualification in mind, the present study investigates whether re-hearsal and exposure to musical training contributes to relatively low-level audi-tory perception. One method that can be used to test this assumption is scrutinyof participants’ perception of certain auditory illusions such as the scale and octaveillusions (Deutsch, 1975). We can ask: are individuals trained in the analysis of soundmore likely to perceive the stimulus veridically compared with untrained listeners(see Craig, 1979; Davidson, Power, & Michie, 1987)? If such a hypothesis is upheldthen experiments that use musical training as an independent variable and use per-ception of such illusions as the dependent variable will be of value in the study ofcognitive development and auditory learning.

1.1. The octave illusion

Deutsch’s (1975) octave illusion with its minimal musical features (i.e., lackingvariety in melody or rhythm) is a simple and effective test stimulus. Deutsch(1982) described the illusion as ‘‘melodic channelling by spatial location’’ and sug-gested that it was a result of different selection mechanisms for pitch and locali-sation. On the one hand, frequencies arriving at one ear are followed and stimulipresented to the other ear, suppressed. On the other hand, the tones are localisedat the ear receiving the high frequency signal. Bregman and Steiger (1980) proposedthat the auditory system treated the 800 Hz tone as a harmonic of the 400 Hz andlocalised to the ‘‘more reliable higher harmonic’’. McClurkin and Hall (1981) re-placed the low (400 Hz) pure tone with a complex (1000–1200–1400 Hz) tone whose

D. Brennan, C. Stevens / Acta Psychologica 109 (2002) 301–314 303

residual pitch was perceived as the low stimulus. 1 They found that the illusionwas pitch-, not frequency-based and that the effect of timbre was negligible.More recently, Ross, Tervaniemi, and Naatanen (1996) mixed the illusion toneswith bursts of illusion-mimicking tones and observed a mismatch negativity inevent-related potentials in auditory cortex. This led them to surmise that the stim-ulus of the octave illusion is encoded according to physical (spectral) rather thanperceptual (illusory) components, and that the illusion is generated beyond audi-tory cortex. As ten Hoopen (1996) notes, different attributes appear to be ana-lyzed by different mechanisms and the output from these analyses is integratedat a later stage. The fact that there is not yet a definitive explanation of the under-lying mechanism does not affect the present study. All that is required is the knowl-edge that some participants are capable of veridical perception of the octavestimulus.

1.2. The effect of musical training and expertise on perception of the octave illusion

Studying the effect of musical training and cerebral asymmetry, Craig (1979) usedperception of the octave illusion as a dependent measure and found that all right-handed musically ‘‘na€ııve’’ participants reported the normal illusory percept. How-ever, over 30% of the right-handed, musically ‘‘experienced’’ participants reportedthe stimulus accurately. A similar finding, although one that involved the more com-plex scale illusion, was reported by Davidson et al. (1987). They found that the scaleillusion stimulus was perceived more accurately by contemporary composers of ato-nal music than four other categories of participants – non-musicians, traditional lis-teners, contemporary listeners and traditional composers. Specifically, participantshighly trained to perceive diatonic tunes (i.e., melodies composed in major and mi-nor keys and conforming to conventions of Western tonality) perceived the scalestimulus in the same diatonic way and therefore produced the illusion, whereas thosetrained in a less traditionally-melodic idiom (contemporary, atonal composition)were more likely to be able to produce the non-melodic veridical percept. Thesetwo studies suggest that both amount and type of musical training can affect percep-tion of auditory illusions.

To further examine effects of training, the level of musical training achieved willbe manipulated and three groups compared: Grades 1–4, Grades 5–8 and Associateof Music, Australia (AMusA). The grades refer to Australian Music ExaminationsBoard standards. AMusA is one of the highest performance qualifications awardedby the Board. The present study used only musically trained participants in an at-tempt to minimise extraneous variables such as differences in motivation and task

1 McClurkin and Hall (1981) note that the residue pitch of a complex (1000–1200–1400 Hz) tone is

generally matched to 200 Hz. They employed this stimulus instead of a complex tone with a 400 Hz

fundamental because the pitch of the latter was often judged as similar to the 800 Hz pure tone in a pilot

study. McClurkin and Hall argued that the phenomenon of octave ambiguity for residue pitch was

avoided by using the complex stimulus with a 200 Hz fundamental.


familiarity that can occur in musician vs. non-musician studies (Gromko, 1993).Craig’s (1979) quasi-experimental design consisted of na€ııve musicians (less thanthree years) and experienced musicians (more than three years). Inclusion of threelevels of musical training in the present study was felt to be a further refinement. Ad-ditionally, by using the musical grade attained rather than age, a more accurate mea-sure of time spent attending to individual stimuli becomes available. Davidson et al.(1987) allude to the idea that level and type of training attained is very different fromtime spent simply listening to sounds or music:

....in the experiencing of music, familiarity is the most important variable...More, however, is necessary for veridical perception at the atomistic level.For one to be able to report accurately on the physical properties of the individ-ual stimuli...it would appear that one needs to have training that involves anal-ysis and manipulation of the individual stimuli (pp. 606–607).

1.3. Aim and hypotheses

Two experiments were designed to examine the implications of results from Craig(1979) and Davidson et al. (1987). Experiment 1 examined the relationship betweenamount of musical training and the likelihood of veridical perception of the octavestimulus. In Experiment 2 we examined the hypothesis that participants who had ex-tensive training with musical stimuli that share acoustic and spatial properties withthe octave illusion are able to produce the spatially-correct, veridical percept. For ex-ample, the near sine-wave quality of many organ pipes and the spatial (dichotic) ar-rangement of organ pipes mimic crucial features of the octave illusion stimulus. Thespecialist training of the pipe organist entails a conscious kind of ‘‘reverse Fourier’’process where sound is built by selecting drawstops that frequently represent individ-ual harmonics (Sumner, 1975). Moreover, the pipes are usually arrayed across theconsole on either side of the keyboard and, for structural reasons, each successivepitch is likely to come from the opposite side of the console. Based on the experienceand training unique to pipe organists and acknowledging their self-selection in study-ing the instrument, it is hypothesized that such specialist training in composing andde-composing sound with respect to both pitch and space contributes to veridicalperception of the octave stimulus.

2. Experiment 1

2.1. Method

2.1.1. DesignThe experiment comprised a one-way between subjects design with three levels of

the musical training variable (Grades 1–4, Grades 5–8, and AMusA). The dependentvariable was the reported percept in response to a sine wave two-tone dichoticstimulus.


2.1.2. ParticipantsThe 44 participants were approximately equal numbers of musically trained males

and females divided into three groups for musical training: Grades 1–4 (14), Grades5–8 (15) and AMusA (15). Participants were selected at random from various musicteaching situations with an effort being made to obtain a wide cross-section of ageand musical training. All participants were right-handed (by self-report) and noone with a known hearing defect was tested.

2.1.3. StimulusThe stimulus consisted of a continuous DAT tape presenting a dichotic chord

consisting of a 400 and 800 Hz tone alternated each 300 ms between the right andleft ear. The tones were linearly ramped up to maximum amplitude during the initial10 ms of each alternation and maintained that amplitude until linearly ramped downto minimum during the last 10 ms of that alternation to minimise clicks.

The stimulus was originally generated using Wavemaker software on a Macintosh5300ce Powerbook laptop computer with a Yamaha CBXD5 Digital Recording Pro-cessor. Sample rate was 44,100 Hz and sample resolution was 16 bit. The stimuluswas then transferred digitally to a Fostex D5 Digital Master Recorder and recordedonto Denon Digital Reality DAT mastering tape.

2.1.4. Equipment and materialsThe stimulus tape was played on a Tascam DA-PI Portable DAT recorder

through a set of Sennheiser HD580 stereo headphones and adjusted to what wasconsidered a comfortable listening level. It was indicated to participants that the le-vel could be altered if required but no participants availed themselves of this option.Response options were presented on a sheet that consisted of four musical examplesand a fifth blank stave (to notate an alternative percept) from which the participanthad to identify the sound they were hearing. The response options (A–E) are shownin Fig. 2. Option B reflects the actual stimulus. The options were selected after pre-tests identified them as the most common percepts. Four different versions of thequestionnaire were used with the response options rotated to minimise positionalbias.

2.1.5. ProcedureParticipants could listen to the stimulus tape for any length of time up to a max-

imum of 30 s. They were asked to circle the musical example that was most similar tothe pattern that they had heard. If participants felt none of the options appropriatethey were asked to notate their own interpretation on the blank staff (see Fig. 2). Al-though actual pitches were indicated on the response sheet staves, it was made clearthat the aim of the experiment was to establish the high/low tone relational percept,not test participants’ ability to identify individual pitches. On completion, partici-pants were asked to answer questions at the base of the questionnaire that categor-ised them by level of musical training and age group. To minimise any bias,headphones were reversed for each alternate participant.


2.2. Results

As hypothesized, non-veridical perception of the stimulus occurred. Fig. 3 showsfrequency of percepts by listeners with different levels of musical training. The verid-ical percept, B, was selected by just 11% of the sample (5/44 participants). 34% of thesample reported Percept A (15/44) and 41% of the sample indicated Percept C (18/44). These responses are the common illusory percepts with Percept C being mostcharacteristic of right-handed participants (Craig, 1979; Deutsch, 1999). 14% ofthe sample checked Percept E (6/44). Participants with extensive musical trainingwere more likely to report veridical perception. Of the five veridical percepts, fourwere reported by participants trained to AMusA level (80%). To test whether theproportion of participants in the AMusA group who reported veridical perception

Fig. 2. Response sheet. Example B is the veridical percept. Examples A and C are the illusory percepts

reported frequently in past experiments with C (higher tone perceived in right ear) being most often re-

ported by right-handed participants.


is greater than the proportion of participants with less musical training who reportedveridical perception, the z-test of proportions was calculated (Freund, 1979). The testindicated that the AMusA proportion (4/15) was significantly greater than theGrades 1–4 and 5–8 proportions of veridical percepts (1/29), z ¼ 2:55, p < 0:05.

In sum, the results show that the majority of listeners experienced the octave illu-sion and that veridical perception was most likely to occur among participants withextensive musical training (AMusA level) and experience.

2.3. Discussion

The results support the hypothesis that listeners with extensive musical training,are more likely to report veridical perception of the stimulus. The effect does not ap-pear to be the result of musical training alone as only one participant with moderatemusical training (to Grades 1–4) reported veridical perception. The octave illusionwas obtained with 75% of participants reporting a percept consisting of a single tonethat alternated between high in the one ear to low in the other ear (Percepts A and C,36/44). It is interesting that, given all participants were right-handed, non-veridicalpercepts were distributed evenly between high/right, low/left (41%) and low/right,high/left (34%). Although Deutsch (1975) referred to the percept switching sides dur-ing prolonged listening, the majority of reports are of almost completely high/right,low/left responses (Deutsch, 1982, 1986, 1999). Percept D (see Fig. 2) was the onlyresponse item not indicated by the pretests. This option was included to providean alternate choice to the veridical for those who identified the presence of tonesat both ears simultaneously. The fact that Percept D was never selected seems to in-

Fig. 3. Experiment 1: Frequency of reported percepts with Grades 1–4 and 5–8 collapsed. Percept B in-

dicates veridical perception. Percept A refers to the higher tone in the left ear followed by the lower tone

in the right ear. Percept C refers to the higher tone in the right ear followed by the lower tone in the left

ear. Percept D refers to the higher tone followed by the lower tone in both left and right ears. Percept E

refers to responses other than A–D and notated by the participant.


dicate that once the presence of two tones one in either ear is perceived, participantsare able to go on and accurately identify the frequency of the tones. The majority ofPercept E (self notation) responses consisted of a veridical high/low tone percept inone ear and a steady high or low tone in the other. This reported percept is similar toCraig’s (1979) findings for right-handed skilled musicians who did not perceive thestimulus veridically.

While various kinds of instrumentalists were tested (pipe organ, piano, saxo-phone, clarinet, trumpet, flute, violin, guitar, and French horn), both of the pipe or-ganists perceived the stimulus veridically. We believe that this is a new result andraises the question as to whether some components of the illusion stimulus mightbe more familiar to pipe organists than other musicians. Studies of differences inthe neural development of musicians have suggested that these variations extendonly as far as the actual sonic components attended to in rehearsal and study. Pantevet al. (1998) measuring cortical representations in musicians found that piano tonesevoked a 21–28% increase in dipole moments among musicians when compared to anon-musician control group. However, there was no significant difference found be-tween the two groups in response to pure tones.

Considering the sonic components of the illusion and the sound generating char-acteristics of pipe organs, we have identified three factors that could explain the pipeorganists’ superior performance. First, many organ pipes generate near sinusoidalwaveforms. Even some of the more complex tones are formed by simultaneouslykeying a number of individual pure toned pipes of different frequencies (and spatiallocation) that constitute the tone’s harmonic structure. Second, the notes sustainconstant amplitude for the duration of the keystroke. Third, the pipes are usually ar-rayed across the console on either side of the keyboard and for structural reasonseach successive pitch level is likely to come from the opposite side of the console.This third feature is akin to dichotic presentation of constant amplitude sine wavesof either high or low frequency to either ear – a description almost identical to thestimulus of the illusion. It is argued that if a relationship between veridical percep-tion of a stimulus and the sonic structure of the participants’ primary instrumentcan be established then this would support the notion that more analytical auditoryperception has been acquired through rehearsal and training. Based on these as-sumptions, it is hypothesized that musicians trained on instruments whose soniccharacteristics are similar to the octave illusion are more likely to perceive the stim-ulus veridically.

3. Experiment 2

Experiment 2 used similar stimuli, equipment, materials and procedure to Exper-iment 1 but was altered in that the level of musical training was limited to a highlytrained AMusA group. This time, the principal instrument of participants was ma-nipulated as the independent variable. It consisted of two levels: pipe organists andall other instruments. The dependent variable remained participants’ choice of one offive possible percepts of a sine wave two-tone stimulus (Fig. 2).


3.1. Method

3.1.1. ParticipantsThe 30 participants had attained approximately equal levels of musical training

(AMusA) and were divided into two equal groups according to their principal instru-ment: pipe organ or other instrument (piano, saxophone, clarinet, trumpet andFrench horn). Only right-handers were considered and no one with a known hearingdefect was tested. No participants from Experiment 1 were tested.

The stimulus, equipment, materials and procedure were identical to those used inExperiment 1.

3.2. Results

The octave illusion was obtained with 53% (16/30) of participants indicating apercept other than the veridical percept including 13 participants reporting the com-mon illusory percept, A or C (see Fig. 2). Fig. 4 shows the frequency of responsesacross the five possible items for participants trained on pipe organ vs. other instru-ments. As hypothesized, more veridical percepts were reported by pipe organists (9/15 or 60%) than those trained on other instruments (5/15 or 33%), z ¼ 1:69, p < 0:05.

4. General discussion

The results of Experiment 2 support the hypothesis that musicians trained on in-struments whose sonic characteristics are similar to the stimulus are more likely to

Fig. 4. Experiment 2: Frequency of reported percepts comparing musicians highly trained on the pipe or-

gan vs. other instruments. Percept B indicates veridical perception. Percept A refers to the higher tone in

the left ear followed by the lower tone in the right ear. Percept C refers to the higher tone in the right ear

followed by the lower tone in the left ear. Percept D refers to the higher tone followed by the lower tone in

both left and right ears. Percept E refers to responses other than A–D and notated by the participant.


perceive it veridically. Taken together, the results of Experiments 1 and 2 indicatethat the probability of veridical perception of the octave stimulus increases as a func-tion of the level of musical training attained. The findings accord with those of Craig(1979) and Shepard (1964) where musical training seems to enhance auditory acuitywith the qualification that there may be factors that predispose individuals to pursuemusical training to a high level and to select a particular instrument (such as the pipeorgan). The results indicate that veridical perception is even more likely when a highlevel of training and experience has involved an instrument whose tonal characteris-tics are similar to that of the stimulus. Experienced pipe organ performers are able toveridically couple auditory dimensions (Deutsch, 1980). Together, these findings sup-port the hypothesis that a substantial part of the musician’s ability to perceive thestimulus veridically can be attributed to rehearsal and training.

The presence of two pipe organists in the Experiment 1 sample was fortuitous.Their inclusion was due to the first authors’ involvement in this specialized fieldand it seems unlikely that such specialists have been included in earlier experimentsamples. In retrospect, it is remarkable how closely the pipe organ can resemblethe stimulus of the octave illusion: consideration of possible alternatives failed toidentify any other instrument with similar attributes. Even the flute generates a muchmore harmonically-rich and varied waveform than the most fundamental of organflue pipes (Grey, 1977; Rasch & Plomp, 1999). The pipe organists’ advantage how-ever, extends further than their instrument’s ability to simulate the octave illusionstimulus. Pipe organists are trained to attain the required timbre from their instru-ment by the practiced and conscious addition of individual harmonics. This ‘‘reverseFourier’’ process is achieved by selecting drawstops which frequently represent indi-vidual harmonics (fundamental (80)¼ 1st harmonic, octave (40)¼ 2nd harmonic, 12th(2 2/30)¼ 3rd harmonic, 15th (20)¼ 4th harmonic, tierce (1 3/50)¼ 5th harmonic, 19th(1 1/30)¼ sixth harmonic etc. – Sumner, 1975, p. 268) at the side of the console and issufficiently isolated from the actual playing of the keys to constitute a separate pro-cess. Most other instruments allow tonal variation but incorporate it as an integraland ideally subliminal component of the actual ‘‘playing’’ process. Further, this pro-cess is not quantified harmonic by harmonic but is usually applied to goals such as‘‘brighter’’ and ‘‘darker’’. The only other instrument bearing any similarity to thepipe organ in the area of tonal structuring separate to performance is the synthesizer(in fact the pipe organ could almost in every way be considered the first synthesizer).However, in the majority of cases contemporary synthesizers use some form of sub-tractive frequency, 2 not harmonic specific synthesis (filter), which could not be as-sumed as similar to the discrete additive structures on which pipe organists aretrained.

The disparity between the pipe organists and other instrumentalists was surpris-ing. Continuity, however, can be found in the data of Davidson et al. (1987) whose

2 Subtractive synthesis refers to complex waveforms that are spectrally shaped by filters and

dynamically shaped using envelope generators. The method is subtractive because the waveform that is

rich in harmonics has its amplitude and frequency content attenuated to some desired level.


results for the group of contemporary composers (80% veridical perception of thescale stimulus) was three times more frequent than any other group tested. Further,during the current experiments a small number of professional recording engineerswere asked to listen to the illusion tape. As they were technically not musicians (acategorisation with which they would disagree!) their results were not included.However, of three tested, two perceived the stimulus veridically. The recording engi-neer results lend support to the earlier suggestion and motivates test of the hypoth-esis that experience with the dichotic presentation of either high or low frequencies toeither ear may be an advantage in veridical perception of the octave stimulus. A pos-sible commonality between these three groups of participants – pipe organists, con-temporary composers, and sound engineers – is their training in isolating individualharmonic components in an overall soundscape.

A secondary goal of the experiments reported here was to assess whether the oc-tave illusion has promise as a tool for measuring auditory analytical capacity and tosuggest its further use in the study of auditory cognition. One way to refine the oc-tave illusion as a tool involves subdivision of the veridical percept. Davidson et al.(1987) propose that the illusion is the result of a schema that has developed to con-serve cognitive resources. Accepting this proposal, we can ask whether those partic-ipants displaying veridical perception (using, presumably, a greater amount ofcognitive resources) revert to the more common illusory percept when other de-mands are placed upon those resources. In interviews after the present experiments,several participants who had perceived the stimulus veridically indicated that theycould also ‘‘hear’’ the common illusion occasionally. One participant reported thatwhen concentrating on the stimulus their perception was veridical but when dis-tracted or not as focused, they would ‘‘slip’’ back to the illusion. Another participantclaimed the ability to switch between the percepts at will. An experiment involvingthe imposition of cross-modal distracters such as that conducted by Hafter, Bonnel,Gallun, and Cohen (1998) could measure the role of cognitive resources. Systemat-ically varying the degree of distraction could allow those perceiving the stimulus ve-ridically to be graded according to the amount of distraction required until veridicalperception of the stimulus breaks down.

Perhaps the frequently documented early accomplishments of musical prodigieswhere their achievements are as much feats of perception as musicality, could bere-conceptualised not only as the development of superior perceptual processes butthe avoidance of ‘‘practical schemas’’ and resource-saving strategies. Although noclaim is made that improved perception correlates with improved musicality, the pre-sent results demonstrate the way in which specialist training contributes to relativelylow-level auditory perception.

Acknowledgements

This research was supported by a Macarthur Auditory Research Centre Sydney(MARCS) summer student research scholarship awarded to the first author. The au-thors thank Professors Gert ten Hoopen, Johan Wagemans, and an anonymous re-


viewer for helpful suggestions. We also thank Professor Roderick Power and Dr Bar-bara David for their advice on an earlier draft.

References

Beal, A. L. (1985). The skill of recognizing musical structures. Memory and Cognition, 13, 405–

412.

Bregman, A. S., & Steiger, H. (1980). Auditory streaming vertical localisation: Interdependence of ‘‘what’’

and ‘‘where’’ decisions in audition. Perception and Psychophysics, 28, 539–546.

Chew, S. L., Larkey, L. S., Soli, S. D., Blount, J., & Jenkins, J. J. (1982). The abstraction of musical ideas.

Memory and Cognition, 10, 413–423.

Craig, J. D. (1979). The effect of musical training and cerebral asymmetries on perception of an auditory

illusion. Cortex, 15, 671–677.

Davidson, B., Power, R. P., & Michie, P. T. (1987). The effects of familiarity and previous training on

perception of an ambiguous musical figure. Perception and Psychophysics, 41, 601–608.

Deutsch, D. (1975). Musical illusions. Scientific American, 233, 92–104.

Deutsch, D. (1980). The octave illusion and the what–where connection. In R. S. Nickerson (Ed.),

Attention and performance VIII. Hillsdale, NJ: Erlbaum.

Deutsch, D. (1982). Grouping mechanisms in music. In D. Deutsch (Ed.), The psychology of music (pp. 99–

134). New York: Academic Press.

Deutsch, D. (1986). Auditory pattern recognition. In K. B. Boff, L. Kaufman, & J. P. Thomas (Eds.),

Handbook of perception and human performance, Vol. II: Cognitive processes and performance (Chapter

32). New York: Wiley.

Deutsch, D. (1999). Grouping mechanisms in music. In D. Deutsch (Ed.), The psychology of music (2nd

ed., pp. 299–348). San Diego: Academic Press.

Deutsch, D., & Roll, P. L. (1976). Separate ‘‘what’’ and ‘‘where’’ decision mechanisms in processing a

dichotic tonal sequence. Journal of Experimental Psychology: Human Perception and Performance, 2,

23–29.

Elbert, T., Pantev, C., Weinbruch, C., Rockstroh, B., & Taub, E. (1995). Increased cortical representation

of the fingers of the left hand in string players. Science, 270, 305–307.

Freund, J. E. (1979). Modern elementary statistics (5th ed.). Englewood Cliffs, NJ: Prentice-Hall.

Grey, J. M. (1977). Multidimensional perceptual scaling of musical timbres. Journal of the Acoustical

Society of America, 61, 1270–1277.

Gromko, J. E. (1993). Perceptual differences between expert and novice music listeners: A multidimen-

sional scaling analysis. Psychology of Music, 21, 34–37.

Hafter, E. R., Bonnel, A. M., Gallun, E., & Cohen, E. (1998). A role for memory in divided attention

between two independent stimuli. In A. R. Palmer, A. Rees, A. Q. Summerfield, & R. Meddis (Eds.),

Psychophysical and psychological advances in hearing (pp. 228–238). London: Whurr Publishing.

Handel, S. (1989). Listening: An introduction to the perception of auditory events. Cambridge, MA: MIT

Press.

Hassler, M., Nieschlag, E., & De La Motte, D. (1990). Creative musical talent, cognitive functioning and

gender: Psychobiological aspects. Music Perception, 8, 35–48.

Howe, M. J., Davidson, J. W., & Sloboda, J. A. (1998). Innate talents: Reality or myth? Behavioral and

Brain Sciences, 21, 399–442.

Hurata, Y., Kuriki, S., & Pantev, C. (1999). Musicians with absolute pitch show distinct neural action in

the auditory cortex. Neuroreport, 10, 999–1002.

Jancke, L., Schlaug, G., & Steinmetz, H. (1997). Hand skill asymmetry in professional musicians. Brain

and Cognition, 34, 424–432.

McClurkin, R. H., & Hall, J. W. (1981). Pitch and timbre in a two-tone dichotic auditory illusion. Journal

of the Acoustical Society of America, 69, 592–594.

Pantev, C., Oostenveld, R., Engelien, A., Ross, B., Roberts, L. E., & Hoke, M. (1998). Increased auditory

cortical representations in musicians. Nature, 392, 811–814.


Rasch, R., & Plomp, R. (1999). The perception of musical tones. In D. Deutsch (Ed.), The psychology of

music (2nd ed., pp. 89–112). San Diego: Academic Press.

Ross, J., Tervaniemi, M., & Naatanen, R. (1996). Neural mechanisms of the octave illusion:

Electrophysiological evidence for central origin. Neuroreport, 8, 303–306.

Shepard, R. N. (1964). Circularity in judgments of relative pitch. Journal of the Acoustical Society of

America, 35, 2346–2353.

Sumner, W. L. (1975). The organ: Its evolution, principles of construction and use. London: Macdonald.

Tan, N. (1979). Tonal organisation in the perception of melodies. Psychology of Music, 7, 3–11.

ten Hoopen, G. (1996). Auditory attention. In O. Neumann, & A. F. Sanders (Eds.), Handbook of

perception and action: Vol. 3: Attention (pp. 79–112). London: Academic Press.

Wolpert, R. S. (1990). Recognition of melody, harmonic accompaniment, and instrumentation: Musicians

vs. nonmusicians. Music Perception, 8, 95–105.


Documents

Specialist musical training and the octave illusion: analytical listening and veridical perception by pipe organists