Steinke, W. R., Cuddy, L. L., & Holden, R. R. (1997). Dissociation of Musical Tonality and Pitch Memory From Nonmusical Cognitive Abilities. Canadian Journal of Experimental Psychology,

7/29/2019 Steinke, W. R., Cuddy, L. L., & Holden, R. R. (1997). Dissociation of Musical Tonality and Pitch Memory From Non

1/20

Dissociation of Musical Tonality and Pitch M emory fromNonm usical Cognitive AbilitiesW.R. STEINKE, L.L. CUDDY, and R.R. HOLDEN

Queen's University, Kingston, Ontario

Abst r ac t The main purposes of this study were toreplicate, validate, and extend measures of sensitivity tomusical pitch and to determine whether performance ontests of tonal structure and pitch memory was related to,or dissociated from, performance on tests of nonmusicalcognitive skills standardized tests of cognitive abstrac-tion, vocabulary, and m e m o r y for digits and nonrepre-sentational figures. Factor analyses of data from 100neurologically intact participants revealed a dissociationbetween music and nonm usic variables, bo th for the fulldata set and a set for which the possible contribution oflevels of music training was statistically removed. Aneurologically impaired participant, C.N., scored withinthe range of matched controls on nonmusic tests butmuch lower than controls on music tests. The studyprovides further evidence of a functional specificity formusical pitch abilities.

How var ious types of cognitive abilities are related, andfurther, how musical ab ilities are related to other abilitiesare questions with both a long history and current im por-tance in cognitive psychology and the psychology ofmusic. Music listening and performance engage a varietyof processing levels from elementary sensory-motorencoding to higher-level relational and s ymb olic represen-tations. Music perception and cognition invite compari-sons to other perceptual and cognitive processes, both interms of commonalit ies and of differences.For example, both music and speech are highly struc-tured forms of comm unication processed by the auditory

system. It has been suggested that precocious abilities inmusic and speech emerge from common origins (e.g.,Davidson & Scripp, 1988; Lyn ch, Sh ort , & Chua , 1995;T r e h u b & Tr ain or, 1993). W arren (1993) rema rks tha t"our use of speech and our product ion and enjoyment ofmusic are based on an elaboration of global organizationalskills possessed by our prelinguistic ancestors" (p. 64).Bigand (1993) com m ents on general cognitive constraintsinfluencing not only hierarchical organization in music

and speech but symbolic processing in general.However , the ongoing search for and description offunctional music modules (e.g., Deliege, 1995) illustratesconcern for the differentiation of musical abilities fromone another and from other cognitive abilities. Distinctneurological processes revealed by brain electrical activity(e.g., Besson, 1997), cerebral bloo d flow p attern s measuredwith positron emission tomography (e.g., Zatorre, Evans,& M eyer, 1994; Zato rre, Halp ern, Perry, Meyer, & Evans,1996), and patte rns of dissociation found in neurologicallycompromised individuals (e.g., Patel & Peretz, 1997)indicate mental operations specific to the domain ofmusic. Taken together, therefore, accounts of bothintegration and differentiation have been proposed. Afurther illustration, one of the earliest and most pertinentto the present study, concerns the relation of music andintelligence.Music and intelligenceEarlier in this century, Spearman (1904,1927) concludedthat music shared the com mon g or general function withall othe r branches of intellectual activity, but allowed thata specific music factor s was operating in its own right.Within the next few decades, researchers identified a musicgroup factor beyond g, but as Vernon (1950) noted, thefactor was p o o r in reliability. Moreover, no consistentsub-grouping of musical factors such as pitch, rh yth m , andtonal memory was found.

Subsequent studies, however, continued to provideencouragement for the notion that musical abilities wereseparable from general intelligence. Shuter-Dyson andGabriel (1981) summarized a large number of studies(involving some 16,000 participants) that examined therelations between intelligence and musical abilities asmeasured by standard musical aptitude tests assessing awide variety of musical skills. All reported correlations,though posit ive, were low. The authors concluded that ,although intelligence may play a role in musical develop-ment, measures of intellectual efficiency are weak indica-tors of musical aptitude and ability. M ore recently, H ow e

Canadian Journal of Experimental Psychology, 1997, 51:4,316-334


2/20

Dissociation 317(1990) also summarized the literature on the relationbetween intelligence and abilities, including music, andcame to a conclusion similar to Shuter-Dyson and G abriel(1981).

W hen em pirical evidence failed to supp ort th e no tionof a unitary construct of intelligence, the notion ofseparate intelligences was put forward. Gardner's (1983)theory of multiple intelligences, for example, states thatmusic intelligence is one of seven separate domains ofintelligence. In a related fashion, Fodor (1983), Jackendoff(1987), and Pe retz an d Morals (1989) have suggested thatthe human cognitive system may comprise dist inct"modules," or physically separate subsystems each "en-dowed with a specific corpus of procedural and declarativeknowledg e" (Peretz & M orais, 1989, pp . 279-280).Support for the dist inctiveness of components orsubskills of music has also increased. In a recent com pre-hensive review of the literature on human cognitiveabilities, Ca rroll (1993) concluded th at several indepe nden tmusical factors within a factor called Broad AuditoryPerception are suggested (italics Carroll, 1993, p. 393) bycurrent research evidence. These include discrimination oftones and sequences of tones on pitch, intensity, d uration ,and rhythmic dimensions, judgments of complex relationsamong tonal patterns, and discrimination and judgment oftonal patterns in musicali ty with respect to melodic,harmonic, and expressive aspects. Carroll cautions that amore definitive list awaits further research. Carroll alsoconcluded that a higher-order factor of general intelligencedominates the Broad Auditory Perception factor as well asthe musical factors listed above. Put differently, the

variance associated w ith a musical factor ma y b e partia llyaccounted for by a unique component and partiallyaccounted for by a general or shared intelligence.The suggestions above arise from test results obtainedfrom neurologically intact individuals wit h vary ing levelsof music training and ability. Evidence that musical abilitycomprises distinct components of music separate fromeach other and from other cognitive abilities has also beenobtained from neurologically compromised individuals inthe form of single-patient studies. Individuals with braininjuries have demonstrated unique patterns of selectiveloss and sparing for musical factors such as melodyrecognition (Steinke, Cuddy, & Jakobson, 1996), contourprocessing (Peretz, 1993a), and timbre discrimination(Samson & Zato rre, 1994). M elody and rhy thm may bedissociated (Peretz & Kolinsky, 1993). Brain injury anddegenerative brain disease have also been shown todifferentially affect musical abilities in comparison tointellectual and linguistic abilities (Peretz, 1993b; Polk &Kertesz, 1993). Finally, high degrees of musical abilityhave been reported for idiot savants who display excep-tional skills in some limited field but are otherwisedefective (Howe, 1990; Judd, 1988).

Sloboda (1985), in reviewing some of the studies on thesupposed location and independence of music in the brain,cites evidence from both normal and brain-damagedindividuals to conclude that "various sub-skills of musichave a certain degree of neural independence. There islittle evidence for a single 'music centre' in the brain" (p.265). While Sloboda tentatively supports the notion ofmultiple modules within music, he agrees with Marin(1982) and writes that fu rther progress is not likely to bemade in this area until "the categories and distinctionsbetween musical activities made on psychological andmusic-theoretic grounds are taken seriously by research-ers" (p. 265).Purposes of the Present StudyThe present study had three purposes. The f irst purposewas to examine that category of musical activity kno wn asa sense of tonality to replicate, validate, and extendmeasures of sensitivity to tona l structure . Th e second w asto determine whether performance on the tests of tonalstructure (or a subset of the tests) was related to, ordissociated from, performance on tests of nonmusicalcognitive skills. Data were collected from a large group ofvolunteers (w = 100) from the general community andstatistically analysed in two stages corresponding to thefirst tw o purposes of the study. Th e third purpo se was toexamine the performance of a neurologically compro-mised individual, C.N., previously assessed as a case ofatonalia (Peretz, 1993a). C.N.'s test results were expectedto provide a direct assessment of dissociation that could becompared with the statist ical solution obtained for thegeneral sample.

Stages 1 and 2 the General SampleTh e purpose of Stage 1 was to test the conv ergent validityof a number of tonality measures. The purpose of Stage 2was to examine the relation between sense of tonality inmusic and selected nonm usic abilities. Th e d ata for Stages1 and 2 were collected from the same group of partici-pants. The overall rationale for test selection, and thegeneral method, will be presented followed by the resultsfor each stage.Stage 1Music TestsSeven music tests we re directed tow ard assessing sensitiv-ity to tonality. A n eighth test assessed me mo ry for tonallyunrelated pitches. Materials for the first four tests werebased on previously available musical constructions.Materials for the remaining tests were constructed byW.R. Steinke, following rules consistent with traditionalmusic theory. T he musical validity of the rules and theirapplication to test construction was verified by a professorof composit ion, C. Crawley, at the School of Music atQueen's University.


3/20

318Sensitivity to tonality. A sense of tonali ty, an importantcom ponent of the "gramm ar" of music, is presumed neces-sary for musical understanding and enjoyment, and istherefore one of the most important and basic aspects ofWestern music. "Without the framework p rovided by thetonic (tonality in general), a note or chord is not integra-ted and remains merely a sound" (Handel, 1989, p. 342).

Our approach to measuring the sense of tonali ty wasinformed by both psychological theory and evidence (forreviews see Bigand, 1993; Dow ling & H arw ood , 1986;Frances, 1958/1988; Krumhansl, 1990a) and by musicthe ory (Lerdahl, 1988; Lerdah l & Jackendoff, 1983;Meyer, 1956; Piston, 1987). In the Western tonal-har-mo nic idiom, tona lity is defined in terms of th e hierarchi-cal organization of pitch relations. Hierarchical pitchrelations exist at three interrelated levels, that of tone,chord, and key. Pitch relations are described in terms ofstability. In any given musical context, some tones,chords, and keys are considered mo re stable or unstablethan others .The concept of the tonal hierarchy describes therelation am ong the single tones w ithin a key. O ne singletone , the tonic, forms a reference point for all tones in thekey. Each of the remaining ton es is located in a hierarchi-cal relation with the tonic. Similarly, the chords within akey may be described in terms of a harmonic hierarchy.The chord buil t on the tonic note, the tonic cho rd, formsa reference point for all other chords in the key. Individ-ual tones and chords reference the tonic in an ongoingfashion as music unfolds. T he listener is presum ed t o abs-tract a sense of tonality from the individual melodic andharmo nic cues with in the overall context of the music.Sensitivity to the hierarchy of pitch relations, alongwith other mental operations tho ugh t to reflect ton ali ty,tap the resources of a complex s ystem . Th e different testsadministered in the present study sought convergingevidence through addressing different aspects of conven-tional notions of tonali ty. A variety of methods andcontexts was employed that included both melodic andharmonic structures. Three assumptions were involved.First, it was assumed that prototypic instances of tonalitycould be created (Cuddy, 1991; Jon es, 1981, 1982, 1991;Krumhansl, 1990a). Second, it was assumed that four or

five distinct levels (at least) could be discriminated along atonal i ty cont inuum (Cuddy, Cohen, & Mewhor t , 1981;see also Croonen, 1994; Dowling, 1991). The sense oftonality does not merely involve a categorical distinctionbetween tonali ty and absence of tonali ty. Third, i t wasassumed that differentiation among the levels requiresaccess to tonal knowledge.The bulk of research findings suggests that the tonalhierarchy construct is psycholog ically valid. Listeners areable to abstract underlying or global aspects of music in

Steinke, Cuddy, and Holdenspite of surface variations, distortions, or transform ationsof various kinds. N o previous studies, however, sought tovalidate different tests of tonality against each oth er.Probe-tone tests. Three tests implemented the probe-tonemethod (Krumhansl, 1990a; Krumhansl & Kessler, 1982;Krumhansl & Shepard, 1979) to assess recovery of thetonal hierarchy for three different contexts. Each was akey-defining context, according to traditional musictheory. On each tr ial , the context was followed by aprobe tone one of the 12 chromatic scale tones, ran-domly selected. Each probe tone was rated on a 10-pointscale for degree of goodness-of-fit of the p ro be t on e to thecontext. The tonal hierarchy is said to be recovered ifhighest ratings are given to the tonic note, next highestratings to the other notes of the tonic triad, lower ratingsto other scale notes, and lowest ratings to the remainingnonsc ale note s. Th ese levels are coded A, B, C, and D inTable 1.

For the Probe-tone Cadences tests, contexts were major(TV-V-I) and m ino r (iv-V-i) cho rd caden ces in th e key s ofC major and C mino r. Each probe to ne was presentedtwice, for a total of 24 presentations of the chord cadenceand probe tone for each major and minor cadence. Theduration of each chord in the cadence and the probe tonewas 1.1 s. Th e cadence and th e pro be to ne w ere separatedby a pause of .5 s. The context for the Probe-tone Melodytest was the "March of King Laois" (as transcribed andrhythmically simplified for experimental purposes byJoh nst on , 1985; see also Cudd y, 1993). Th e me lody, a16th-century Celtic tun e characterized b y simp le elabora-tions of the tonic triad, was chosen because it is highlytonal according to music-theoretic descriptions andbecause it was not likely to be familiar t o any participan t.Th e me lody contains 60 notes of equal value; the durationof each note in the melody w as .2 s. Th e duration of theprobe tone and the pause between the end of the m elodyand the probe tone was 1 s. See App endix A for the musicnotation of the Probe-tone Melody.Melody completion ratings. The fourth and fifth testsrequired participants to rate the last note of a tonalmelody on how well i t completed the melody. Severalstudies have used ratings of goodness and comp leteness ofphrase endings to assess tonal knowledge (e.g., Abe &Hoshino, 1990; Boltz, 1989).

For the Familiar Melodies test, six melodies wereselected with the restrictions that each melody be: (a)probably within the accessible cultural repertoire of so-called "familiar" melodies; (b) in a major mode; (c) in 4/4time; (d) four b ars in length; (e) typically re produ ced at a"moderate" tempo; and (f) ended on the tonic note. Themelodies selected were O h Susannah, Jo y to th e W orld,


4/20

Dissociation 319TABLE 1Music-theoretic levels of ton ality

Level Prob e-tone tests function of probe within contextA Tonic (1 " scale degree)B Othe r Tonic Tr iad (3"1 and 5* scale degrees)C Oth er Dia tonic (2nd , 4 th , 6 th and 7 th scale degrees)D Chro matic (nondiatonic)

Level Familiar and No vel Melodies test function of final noteA Tonic (1 " scale degree)B Othe r Tonic Tr iad (3 ri o r 5 th scale degree)C Oth er Dia tonic (2nd , 4 th , 6 th or 7* scale degree)D Chrom atic (nondiatonic note within 1 to 4semitones from p enultimate note)E Chro matic (nondiatonic note within 5 to 7semitones from p enultimate note)

Level Ton al/Ato nal Melodies testA Ton al/To nal (tonal diatonic melody)B Ton al /Aton al /Tona l (nondia tonic notes in middle of

level A melody)C At on al/T on al (nond iatonic notes in first half of levelA melody)D To nal/A tona l (nondiatonic notes in second half oflevel A melody)E At on al/A ton al (first half of level C and second half oflevel D)

Level Cho rd Progression testA Diatonic Progressions (beginning on the tonic and ending

with perfect cadence)B Close Mo dula tion (to perfect cadence in closely relatedkey)C Distant Mod ulation (to perfect cadence in distant key)D Randomly Ordered Diatonic ChordsE Randomly Ordered Chromatic Chords

Early One Morning, London Bridge, Frere Jacques, andGo od N ight Ladies. The m elodies contained an average of21 notes. The duration of each quarter note in eachmelody was .5 s.

Each m elod y ending was varied according t o a five-leveltonal-atonal continuum. The levels, from A, the mosttonal to E, the mo st aton al, are listed in Ta ble 1. Acc ord-ing to Table 1, therefore, six melodies ended w ith level A ,the tonic, and 24 variations on these melodies did no t endon the to nic. Twelve of the 24 variations maintained th eoriginal con tour of the tonic ending, and 12 violated th eoriginal conto ur. Participants we re asked to rate, o n a 10-point scale, how well the last note of the melody com-pleted the melody. See Appendix A for an example of afamiliar melody with five possible endings.For the Novel Melodies test, six melodies were con-structed similar in melodic structure t o th e melodies of theFamiliar Melodies test. There were two stylistic differ-ences. First, the rhythmic structure of the novel melodieswas somewhat simpler than the familiar m elodies. Second,for novel melodies, the melody ending for each of the firstthree levels of tonality (A, B, and C; see Table 1) wassounded equally often, on average, within the melody.The total duration of melod y notes corresponding to th eending note was, on average, the same 10.5 sixteenth-

note beats, or 1.33 s. (Level D and E endings, of course,never occurred in the melody.) For familiar melodies, onthe other hand, the total duration of melody notescorresponding to the ending n ote decreased across levels A,B, and C. In other respects, the novel melodies weresimilar to t he familiar melodies. See Ap pen dix A for anexample of a novel melody with five possible endings.Rating tonal structure of melodies. A sixth test involvedrating the tonal stru cture of unfamiliar m elodic sequences.Melodic changes, howev er, were no t l imited to th e f inalnote . The systematic addition of nonkey tones to a tonalsequence resulted in melodic sequences with increasinglyambiguous tonal centres. Previous studies have shown thatlisteners are reliably able to track the perceived degree ofsyntactic completeness of such melodies (Cuddy et al.,1981; Cuddy & Lyons, 1981).

For the Tonal/Atonal Melodies test, six tonal melodieswere constructed with the restr ictions that each melodybe : (a) in a major mo de; (b) in 4 /4 time ; and (c) four barslong. The duration of each quarter note was .5 s. Each ofthe six melodies was then used as a proto typ e and variedaccording to a five-level tonal-atonal continuum. Thelevels, A to E, are listed in Table 1. Participants were askedto rate, on a 10-point scale, how good or well-formed eachme lody so unde d. See Ap pen dix A for an examp le of fivelevels for one melody prot otyp e.Rating tonal structure of chord progressions. The seventhtonality test involved rating the perceived tonal structureof chord progressions. Studies have demonstrated thatvariations in the properties of chord sequences influencerecognition memory (Bharucha & Krumhansl, 1983;Krumhansl, Bharucha, & C astellano, 1982; Krum hansl &Castellano, 1983), prototypicality ratings (Smith &Melara, 1990), and perceptions of mod ulation (Cudd y &Th om pso n, 1992; Krum hansl & Kessler, 1982; Th om pso n& C uddy , 1989).

For the Chord Progressions test, 25 progressions of eightchords were constructed. The 25 progressions represe ntedfive levels of tonality, with five examples of each level.Ch ords used in the progressions were major, m inor, majorseventh, minor seventh, dominant seventh, augmented,diminished, and diminished seventh. The d uration of eachchord was 1.2 s. The levels of tonality, A to E, are listed inTable 1. Participants were asked to rate, on 10-point scale,ho w w ell the eight chords in the sequence followed oneano ther in an expected ma nner. See App endix A for anexample of each level of the chord progressions.Test of pitch memory. Memory for pitch is considered a"basic ingredient of musical ability" (Shuter-Dyson &Gabriel, 1981, p. 239). Memory for pitch may be assessedby requiring participants to judge wh ethe r a tone was o r


5/20

320 Steinke, Cuddy, and Holdenwas not part of a sequence of tones (Dewar, Cuddy, &Mewhort, 1977), to judge the relation between the firstand last tones of a sequence when other tones or silenceintervenes (e.g., De utsc h, 1970, 1972, 1978; Fran klan d &Cohen, 1996; Krumhansl, 1979), to note scalar andnonscalar changes in pairs of melodies (Bartlett &Dow ling, 1980; Do wling & B artlett , 1981), or to notechanges in short melodic fragments either tested inisolation or in the context of additional preceding andfollowing sequences of a tonal or atonal nature (Cuddy,Cohen, & Miller , 1979). The present study had partici-pants judge wh ethe r a tone presented in isolation was par tof a preceding sequence of tones.

Seventy-two trials were cons tructed, each c onsisting ofa sequence of tones . The d uration of each tone was .6 s.Each sequence was followed by a pause of .9 s, followedby a test tone of .6 s. The first eight trials consisted of oneton e followed b y a test tone, the next eight consisted oftw o tones followed b y a test tone, the next eight consistedof three tones followed by a test tone, and so on up toeight trials of nine tones followed by a test tone.The Pitch Memory test was constructed with a deliber-ate effort to avoid or violate ton al rules. It was intend ed t oassess memory for tonally unrelated pitches and, as such,to provide a musical counterpart for the Digit Span test(below) which assessed memory for unrelated digits.Several steps were taken. First, sequences of tones wererandom ly selected from the 12 chrom atic tones with in on eoctave. Next, sequences which predominantly containednotes from a single major o r m ino r key, contained majoror minor triads, or contained scalar sequences were dis-carded. Third, the first author played and listened to theremaining sequences; those sequences that conveyed a mu-sical impression of tonality to the author were discarded.Finally, a key-finding algorithm (Krumhansl & Schmu-ckler, cited in Krum hansl, 1990a) was applied post-ho c t oassess the tona l strength of the pitch distribution of t he 72sequences. Correlations were obtained between thedistribution of pitches in each sequence and the standard-ized tona l hierarchy for each of the 24 major and m inorkeys. Th e standardized ton al hierarchies for C major andC min or were re ported in Krum hansl and Kessler (1982)and th e set of probe-ton e values are given in K rumhansl

(1990a, p. 30). Values for each of the other keys wereobtained by o rienting the set to each of the different ton icnotes. For each sequence the highest correlation soobtained was selected to represent th e tonal strength of thedistribution. T he average of these correlations was .54; theaverage for each sequence length rang ed from .45 to .63,with no relation between length of sequence and size ofcorrelation. A correlation of .66 is required for signifi-cance at the .01 level (one-tailed test).Fo r four ran dom ly chosen sequences with in each groupof eight trials, the test tone was one of the preceding

tones; except for th e one- and two -tone sequences, the testtone was never the first or last note of the sequence. Forthe four remaining tr ials, the test tone was a tone w ithinthe con tou r boundaries of the preceding sequence but no toccurring in the sequence. A single random order withineach sequence length was constructed in an attempt tomodel the test on the procedures for the Digit Spansubtest of the WAIS-R (Wechsler, 1981).Participants were asked to respond 'Yes' if the test ton efollowing the sequence of tones was heard within thepreceding sequence, and 'No' if the test tone was notheard within the preceding sequence.

Stage 2 Nonmusic Tests and Evaluation ofFactorsNonmusic tests. Nonmusic tests were standardized psycho-logical tests specifically designed t o assess cognitive sk ills.The tests are listed in Table 2. They are widely availableand in c om mo n use in a variety of clinical and expe rimen-tal situations. The tests were selected to assess bothcognitive abstraction and n onabstraction abilities (colum n1 of Table 2) . As well, they were selected to provide bothaudito ry and nonau ditory co ntexts and both linguistic andnonlinguistic contexts (columns 2 and 3 of Table 2,respectively).The first three tests listed were developed to assessabstraction abilities. The Wisconsin Card Sorting Test(Heaton, 1981) was first introduced by Berg (1948) as anobjective test of abstraction and "shift of set." Participantsare required to sort cards of various forms, colours, andnumbers, according to shifting criterion principles.Abstraction ability is required to discern the correct

sorting principles based on information presented on thecards and information given by the examiner as towh ether each sort was correct or incorrect .The Abstraction subtest of the Shipley Insti tute ofLiving Scale is described as requiring the participant to"induce some principle common to a given series ofcomponents and then to demonstrate [his or her] under-standing of this principle by continuing the series"(Shipley, 1953, p . 752). Th e com pon ents in the subtestinclude letters, numbe rs, and word s.The Similarities subtest of the WAIS-R (Wechsler, 1981)consists of 14 items w hich assess logical abstract reason ing

or concept formation. The items require test-takers torecognize the relation between two objects or ideas.Th e three nona bstraction tests were tests of vocabulary,attention and memo ry. The V ocabulary subtest from theShipley Institu te of Living Scale is a meas ure of v ocabu laryknowledge. In addition, the Total score, a combin ation ofthe Vocabulary and Abstraction subtest scores, may beused to assess general intellectual functioning and t o detectcognitive impairme nt (Heaton, 1981). Th e To tal score canalso be used to obtain a reliable estimate of the WAIS-Roverall IQ (Zachary, 1986).


6/20

Dissociation 321TABLE 2Characteristics of Nonmusic Tests

Test

Wisconsin Card SortingAbstraction (Shipley)Similarities (WAIS-R)Vocabulary (Shipley)Di gi t Span (WAIS-R)Figural Memory (WMS-R)

AbstractionYe sYe sYe sN oN oN o

Nature of testAuditory

N oN oYesN oYe sN o

LinguisticN oYe sYesYesYe sN o

Summary statisticsMean76.733.821.933.516.87.7

SD23.2

5.03.13.93.71.7

Range'6-7001 2 - 4 01 1 - 2 72 3 - 4 08 - 2 64 - 7 0

* The italic score is the maximum possible score. The m aximum score for the Similarities test is 28.

The Digit Span subtest of the Wechsler Adult Intelli-gence Scale-Revised (WAIS-K.) was designed as a meas ure ofattention/concentration/freedom from distractibility andof immediate auditory m em ory (Wechsler, 1981; Zim-merman & Woo-Sam, 1973). Digit Span appears to be avalid measure of short- term aud itory m em ory and atten-tion, but is not considered a valid indicator of o the r typ esof mem ory skills (Zimmerm an & W oo-Sam, 1973).In contrast to the Digit Span test, the Figural Memorysubtest of the Wechsler Memory Scale-Revised (Wechsler,1987) is a nonverbal measure of short-term memory thattests ability to remember nonrepresentational designs.Evaluation of factors. Principal component analysis,followed by model testing analyses, was conducted on thefull data set (performance for each participant on eightmusic tests and six nonmusic tests). Various possibleoutcomes were evaluated. One was that if the abstractionof the tonal hierarchy required for th e tonality tests sharedresources with other processes of abstraction, the factorstructure should then isolate abstraction abilities (musicand nonmusic) from other abilities. Descriptions of tonalorganization include the general cognitive principles ofhierarchical ordering, categorization, classification, andprototypicality (Krumhansl, 1990a). Thus it is possiblethat general mechanisms are shared.

Examples of other outcomes evaluated were that thefactor structure would isolate all music tests fromnonmusic tests, auditory contexts from nonauditorycontexts, and/or l inguistic contexts from nonlinguisticcontexts (see Carroll, 1993; Ga rdne r, 1983; Shuter-Dyson& Gabriel, 1981; Sternberg & Pow ell , 1982; Waterhouse,1988). Yet another possible outcome was that all testswould primarily engage general intelligence or wouldreflect test-taking ability. In that case, no factor structurebeyond a single factor should the n emerge.METHODParticipantsOne hundred adults served as voluntary participants inthis study, 41 males and 59 females. They were recruited

both from the university community (from campusposters and a participant pool) and the community at large(through newspaper advertisements). All were able tospeak and read written English, and all claimed normalhearing.The age range of participants was 18-40 years

(mean = 26.8, SD = 6.2 years). The range of years of for-mal education was 7-22 years (mean = 15.6, SD - 2.7).Nine teen participants had 12 years or fewer of formal edu-cation, 66 had 13-16 years, and 15 had 18 or mo re years.Sixty-one participants had little or no music training,defined as no classroom or private music lessons afterelementary school and/or one year or less of secondaryschool band, and no curren t engagement in music instruc-tion or performance activit ies. Twenty-two participantshad moderate training, defined as classroom or privatemusic lessons during elementary school and/or two ormo re years of classroom or private lessons du ring second-

ary school plus current engagement in instruction orperformance activities (including choir singing, or playingan d/o r singing in a band o r other ensemble as a hob by).Seventeen participants were highly trained, defined asachievement of a university degree or college diploma inmusic, or present engagement in music performance orinstruc tion at a semi-professional or professional level.Music Test ProceduresStimuli for melodic sequences were synthesized musicaltimbres, created by a Yamaha TX81Z synthesizer. Thesynthesizer was controlled by an Atari 1040ST computerrunning "Notator" music processing software (Lengeling,Adam , & Schup p, 1990). A n exception was the Probe-toneMelody test for which the synthesizer was controlled bya Zenith Z-248 computer running "DX-Score" software(Gross, 1981). The synthesizer settings were factory presett imbres, and differed among tests to provide variety.Synthesizer sett ings were: Probe-tone Melody W oodPiano (A15); Familiar and No vel M elodies Pan Floot(B12); Tonal/Atonal Melodies Flute (Bll) ; and PitchM e m o r y New Electro (A12).

Stimuli for probe tones and harmonic sequences


7/20

322 Steinke, Cuddy, and Holden(cadences and chord progressions) were "circular" tones(Shepard, 1964) and "circular" chords (Krumhansl, Bha-rucha, & Kessler, 1982), respectively. They were createdon a Yamaha TX802 synthesizer controlled by an Atari1040ST computer running "Notator" music processingsoftware. Circular tones and chords consisted of 6 and 15sine-wave components, respectively, with rise and decaytimes of 20 ms each. The components were distr ibutedover a six-octave range under an amplitude envelope thatapproached hearing threshold at the high and low ends ofthe range. This procedure results in tones and chords thatsound organ-like and do not have a well-defined pitchheight. Th e purpose of this meth od of construction is toincrease the likelihood that listener judgments will bemade on the basis of tone or chord function within thetonal scale rather than o n pitch height.

All trials were recorded o n Son y UX-S60c audiocassetteswith an Alpin e AL-35 tape rec order. T he o rder of trialswas randomized, and, except for the Pitch Memory test ,three different rand om orders were recorded for each test.Trials were separated on the tape by a silent gap of 4.5 s. (Probe-tone Cadence tests) or 3 s (all other tests). Inaddition, for all tests other th an th e probe -tone tests, eachtrial was assigned at rand om to o ne of the 12 major k eys.Practice trials were also recorded. Fo r th e probe -tone tests,practice trials were sampled from the test trials. For theremaining tests, practice trials consisted of materialssimilar, but not identical to, the test trials.

Music sequences were reprod uced thr oug h the speakersof a Phillips AW 69 0/0 7 portable tape player at a comfort-able loudness level, as determined by each participant(abou t 55 to 70 dB SPL).For each music test , participants provided writ tenresponses. The rating scales were always oriented so that"10" was the high end of the scale, " 1 " the low. Partici-pants were told that there were no t ime l imits on theirratings for each trial of each test; the y we re instructed touse the pause butto n o n the tape player if necessary, or toindicate to the experimenter that more t ime was neededthan that which was provided by the si lences betweentrials. N o feedback w as given following p ractice trials onthis test or any subsequent music test , but instructionswere clarified whenever necessary.Nonmusic Test ProceduresEach participant was tested on each test listed in Table 2.Administration followed published test protocols. For thenonmusic tests, the Shipley Institute of Living Scalespecified a ten-m inute tim e limit for each subtest. N on e ofthe othe r non mu sic tests had t ime l imits, and pacing wasdetermined by each participant.General Testing ProceduresAll participants were tested in a quiet room. They were

asked to read a writ ten description of the study and toread and sign a consent form.Data were collected from each participant in thefollowing order. D emo graphic data, including age, gender,years of formal education completed, level of musictraining, and self-perceived level of musicality, werecollected f irst . The music and nonmusic tests were pre-sented next in an alternating fashion, beginning with amusic test . The order of presentation of both the musictests and the non mus ic tests was independently random-ized for each participant. Each participant was randomlyassigned to one of the three random orderings of stimulifor the music tests with th e exception of the Pitch M em-ory test which w as constructed in only one order. Ord erof presentation of Probe-tone Major and M inor Cadencetests was counterbalanced across participants.Each participants was verbally debriefed at the conclu-sion of the testing. Each testing session lasted approxi-mately two hours .

STAGE 1 RESULTS OF THE MUSIC TESTSFo r the Probe-tone Major and Minor Cadence and Probe-tone Melody tests, mean ratings for each of the 12 probetones were computed for each participant. For the Famil-iar, Novel, and T ona l/Ato nal M elodies tests, and for theCh ord Progressions test, mea n ratings for each of th e fivelevels on the tona l/aton al con tinuu m w ere computed foreach participant. Fo r the Pitch M em ory test , the num berof correct responses (out of eight) for each of the ninesequence lengths was calculated for each participant, aswell as the total number of correct responses out of apossible total of 72.Probe-tone tests. Mean ratings for the probe-tone tests aregiven in Figure 1 for the entire sample of 100 participantsand for each level of music training. For the Probe-toneMajor Cadence, Mino r Cadence, and M elody tests, overallmean ratings were highest for the tonic note. The thirdand fifth scale tone s were rated next m ost hig hly, followedby the remaining diatonic notes. The chromatic noteswere all rated lowest. Similar results were obtained foreach of the three levels of music training.

Analyses of variance for the Major Cadence revealedsignificant main effects for probe tones, f(ll,1067) = 81.93, MSe - 2.26, p < .001, and for the followingcontrasts: ton ic vs 3rd + 5th, f( l, 97) = 21.74, MSe - 2.83,p < .001; tonic tr iad vs othe r diatonic tones, /( I ,97) = 204.94, M Se = 4.05, p < .001; and diatonic vschroma tic ton es, /(1, 97 ) - 311.69, MSe = 2.80, p < .001.The music training by probe tone interaction was signifi-cant, F(22, 1067) - 9.40, MS, - 2.26, p < .001. Highlytrained participants tended to rate the diatonic probeshigher and the chromatic probes lower than the lessmusically trained participants.


8/20

Dissociation 323

- All Participants- U.W Music Training- Modern. Mutt: Training

Probe-Tone Major Caden ce

I5.c

C C I D C M E F F C G A Af B

Probe-Tone Minor Caden ce

C C D D E F F G G A Al B

Probe-Tone Melody

C Ci D Wt E F F I. C A A 0

Probe Tone

Figure 1. Mean rating for each probe tone for the Probe-tone MajorCadence, the Probe-tone Minor Cadence and the Probe-tone Melody testfor all participants and for each level of m usic training.

Results of analyses of variance for Probe-tone MinorCadence and Probe-tone Melody tests were similar to theresults described above. (Full details are available inSteinke, 1992.) The correlations between the overall meanratings in Figure 1 and the standardized to nal hierarchiesreported in Krumhansl (1990a) were .93 (Probe-toneMajor Cadence and standardized C-major hierarchy), .98(Probe-tone Minor Cadence and standardized C-minorhierarchy), and .96 (Probe-tone Melody and standardizedC-major hierarch y). All correlations are significant bey ondthe .001 level (one-tailed t-test), df - 10.Familiar Melodies and Novel Melodies tests. Mean ratings forFamiliar and Novel Melodies tests are shown in Figure 2(top two panels). Overall, listeners rated the Level Aendings, consist ing of the tonic note, the highest . Theother four types of endings (3rd or 5th, other diatonic,close chromatic, distant chromatic) were rated progres-sively lower with Level E endings rated the lowest. Thistrend was also evident at each level of mu sic training. N oparticipant reported that any melody from the FamiliarMelodies test was unfamiliar.

Analyses of variance revealed significant main effectsfor levels of melody ending for both Familiar Melodies,F{4, 388) = 738.81, MSe = 0.73, p < .001, and for Nove lM elodie s, i


9/20

324 Steinke, Cuddy, and HoldenFamiliar M elodies

^ Level A: Tonic Note Ending| Level B: 3rd or Sh Ending Level COther Diatonic Ending

Level D: Ocse Chromatic EndingLevel E-Distant Chromatic Ending

All Participants Lew Moderate HighMu sic Training Music Training Music Training

Novel M elodies^ Level A; Tonk Note Ending Q Level D: Close ChromaHc Ending Level B: 3rd or 5 * Ending Level C Other Diatonic Ending

Level E: Distant Chromatic Ending

All Participants Low Moderate HighMusic Training Mu sk Training Music Training

S To nal/ A tonal Melodiesg j Level A: Tonal/Tonal Q Level D: Tonal/Atonal Level B: Tonal/Atonal/Tonal | Level E: Atonal/Atonal Level C Atonal/Tonal

1 0 - ,9 -87 -6~5 -4 -3 -2 -1

All Participants Low Moderate HighMusdc Training Music Training Music Training

Chord Progressions^ Level A: DiatonicProgression| Level B: Close Modulation Level C: Distant Modulation

LevdD: Random Diatonic ChordsLevel E: Random Chromatic Chords

All Participants Low Moderate HighMusic Training Music Training Music TrainingMusic Training Level

Figure 2. Mean rating for each of five levels of tonality for the Familiar Melodies, Novel Melodies, Tonal/Atonal Melodies and Chord Progressions testsfor ad participants and for each level of music training.

interaction between tonality level and music training levelwas also evident, /(8 , 388) = 7.68, MSt = 0.78, p < .001.Chord Progressions test. Mean ratings for the C h o r dProgressions test are presented in the lower r ight-handpanel of Figure 2. Listeners rated the Level A progressions(consisting of non-modulating ch ord sequences beginningon the tonic chord and ending with a perfect cadence) thehighest. Ratings becam e progressively lowe r for Levels B-D, with the lowest rating given for Level E progressions(random chromatic chords) . This pattern was evident ateach level of music training. Analyses of variance andcontrasts revealed significant differences among levels oftonali ty of chord progressions, f(4, 388) = 213.77,MSe 0.82, p < .001 . Co ntras ts withi n main effects co m-pared ratings of Level A endings to Levels B to E, Level Bto Levels C-E, Level C to Levels D-E, and L evel D to LevelE, and w ere all significant, p < .001. A significant interac-tion between tonality level and music training level was

found, /(8, 388) = 6.31, MSe - 0.82, p < .001. Similar tothe preceding results, higher music training resulted ingreater differentiation between levels of tonali ty.Pitch Memory test Results of the Pitch Memory test arepresented in Figure 3. The average number of correctidentifications out of eight for each of the nine sequencelengths decreased from 7.9 for 1-note sequences to 4.4 forthe 9-note sequences. Analysis of variance revealed signifi-cant effects of sequence length, F{8, 766) - 81.01, MSe =1.35, p < .0001, level of music training, f(2, 97) = 23.37,M Se = 3.12, p < .0001, and the interaction betweensequence length and music training, f(16, 766) = 2.46,MSe - 1.35, p < .001. Beyond sequence length 1, highlytrained participants identified more sequences correctlythan participants with little or no m usic training.

Th e data were inspected for evidence that accuracy ofidentification was related to th e size of the correlation bet-ween the distr ibution of pitches in the sequence and the


10/20

Dissociation 325TABLE 3Mean Spearman rank-order correlation of participant ratings for m usictonality tests with music-theoretic levels of tonalityTest Music training level of participants

Probe-tone Major CadenceProbe-tone Minor CadenceProbe-tone M elodyFamiliar MelodiesNovel MelodiesTonal/Atonal MelodiesChord Progressions

Allparticipants

.49

.47

.67

.75.80.54.55

Low.33.32.58.73.80.46.47

Moderate.63.63.74.78.81.63.64

High.66.66.81.80.86.67.69

standardized tonal hierarchy (see M etho d abo ve). No reli-able trends we re found an unsurprising result given thatthe test was designed to y ield a range of low correlations.STAGE 2 RESULTS OF NON MUS IC TESTS ANDEVALUATION OF FACTORSNonmusic TestsMeans, standard deviations, and ranges for all the non-music tests can be found on the right-hand side of Table 2.Wisconsin Card Sorting Test Scores on the W isconsin CardSorting Test represent the percentage of conceptual levelresponses. The results are similar to normative scores fornormal participants (Heaton, 1981).

Abstraction. Scores on the Abstraction subtest of theShipley Institute of Living Scale are number of correctresponses to twenty i tems, multiplied by two. Althoughthe n ormative data of the Shipley Institute of Living Scaleare based on a psychiatric popu lation, th e obtaine d meansin Table 2 are similar to results obtained on normalpopulations of student nurses and hospital employees(Zachary, 1986).Similarities (WAIS-R). Scores on the Similarities subtest ofthe WAIS-R represent the number of items answeredcorrectly, with each item scored as 0, 1, or 2, dependingon the quality of the response. These means reflect averagescores based on the norms of the WAIS-R (Wechsler, 1981).Vocabulary. Scores on the Vocabulary subtest of the Shi-pley Institute of Living Scale represent the total numbercorrect of 40 vocabulary items. The obtained means aresimilar to results obtained on normal populations ofstudent nurses and hospital employees (Zachary, 1986).Digit Span (WAIS-R). Scores on the Digit Span subtest of theWAIS-R repre sent tot al n um be r of sequences recalledcorrectly. These means reflect average scores based on thenorms of the WAIS-R (Wechsler, 1981).

P i t c h M e m o r yAll Participants - - Moderate Music TrainingLow Music Training - * - Hig h Music Training

I

1 2 3 4 5 6 7 8 9Sequence Length

Figure 3. Mean numb er of correct identifications for each sequence lengthof the Pitch Memo ry test for all participants and for each level of m usictraining.

Figural Memory (WMS-R). These scores represent the totalnum ber of figures correctly recalled, out of 10. T he meanscores are slightly higher than the mean raw score re-ported in the standardization sample (Wechsler, 1987).Principal Components A nalysisFor the principal components analysis, each participantwas assigned a single score for each mu sic test. T he scorefor tonality tests represented the corresp onden ce betwee nthe participant's ratings and the music-theoretic levels oftonality as defined in Table 1. For the probe-tone tests,rank-order correlations were calculated between obtainedratings and a quantified pred ictor for w hich levels A, B, C,and D were coded 4, 3, 2, and 1, respectively. For theremaining tests, rank-order correlations were calculatedbetween obtained ratings and a quantified predictor forwh ich levels A, B, c, D, and E were coded 5, 4, 3, 2, and 1respectively. The mean rank-order correlation for each testis given in Table 3 for all participants and for each level ofmusic training. All mean correlations are significantlydifferent from zero (one-tailed t-test, p < .001, df = 99overall, and df = 60, 21, and 16, respectively, as musictraining level increased.) The single score for the PitchMemory test was the total correct out of 72 (eight se-quences x nine sequence lengths).

All distributions of scores for the cognitive and musictests were checked for normality and outliers. Several tests


11/20

326TABLE 4Varimax rotated two factor solution of the principal componentsanalyses on the music and n onmusic tests for full and partial correlations

Test

Probe-tone Major CadenceProbe-tone Minor CadenceProbe-tone M elodyFamiliar MelodiesNovel MelodiesTonal/Atonal MelodiesChord ProgressionsPitch MemoryWisconsin Card SortAbstraction (Shipley)Sim ilarities (WAIS-R.)Vocabulary (Shipley)Dig it Span (WAIS-R)Figural Memory (WMS-R)Compon ent VarianceProportion of VarianceTotal variance accounted for

Factor loadingsFull correlation

Factor I Factor II.764.798.776.581.688.741.629.696

-.011.153.020.363.348

-.0014.341

.311

.080.117

.181-.335-.117.191.142.291.577.674.693.500.375.450

2.181.156

46.7%

Partialcorrelation'Factor I Factor II

.636.710

.689

.427

.589

.655

.448

.541-.011.095.013.283.135

-.0592.946

.21036.1%

-.024.028.130

-.422-.207.159.081.277.567.658.671.489.380.417

2.110.151

"Partial correlation rem oves the effect of m usic training.

were transformed to reduce skewness and kurtosis, andthree outlying data points were brought to within threestandard deviations of the mean. A square-root transfor-mation was used on th e data from th e Abstraction subtestof the Sh ipley Instit ute of L iving Scale, Similarities subtestof the WAIS-R, Probe-tone Melody test, Novel Melodiestest, Familiar Melodies test , Ton al/A tona l M elodies test,and Cho rd Progressions test. A logarithmic transform wasused on the data from the Wisconsin Card Sorting Test .Bivariate scatterplots were produced to check for bivariatenormality.

The principal components analyses obtained onesolution based on full correlations between the music andnonmusic tests and another solution based on partialcorrelations. Music training was the variable controlled toobtain partial correlations. Each participant was assignedto one of three discrete levels of music training: little ornone, moderate, or high (designated " 1 " , "2", and "3",respectively, and described in detail above). Correlationswere then obtained between the music and nonmusic testswith m usic training partialled ou t.

A two -factor solution w as chosen on th e basis of factorloadings, a scree plot, and a parallel analysis procedure(Horn, 1965; Zwick & Velicer, 1986) which estimatedeigenvalues for random-data correlation matrices (Long-man, Cota, Holden, & Fekken, 1989). Table 4 pre-sents

Steinke, Cuddy, and Holdenthe varimax rotated two-factor solution of the principalcomponents analysis of the music and cognitive testscalculated on the full correlation matrix and with musictraining partialled out. Factor loadings indicate that forboth full and partial correlation matrices, the musicvariables loaded on the first c om pon ent and the cognitivevariables loaded on the second com pone nt. H owe ver, thetotal variance accounted for was higher in the full correla-t ion solution than in the partial correlation solution th atcontro lled for the effects of music training .Model TestingModel testing analyses were carried out to assess thecongruence of the obtained factor loadings of the cognitiveand music variables with each of eight hypothesizedmodels. Analyses were conducted on bo th full a nd partialcorrelation matrices. Model testing involved: (a) extractionof as man y com ponen ts as indicated by a particular model;(b) orthogonal procrustean rotation of observed compo-nent loadings to a hypothesis matrix representing themode l; and (c) com putation of a coefficient of congruence(Harm an, 1976) between each rotated com pone nt 's load-ings and i ts corresponding m odel com pone nt 's loadings.

The first model tested (Model 1) was a basic "musicversus nonmusic" model. This model reflected the factorloadings obtained from the principal compo nents analysisdescribed above. Other models were assessed to furtherverify the validity of the factor structure that emergedfrom the principal comp onents analysis, and to address thepossibil i ty that a num ber of other m odels might providea better account of the data.Models 2, 3, and 4 addressed auditory, linguistic, andabstraction characteristics of the nonmusic tests, summa-rized in Table 1. For Model 2, "auditory versus non-auditory," all of the music tests plus the Similarities andDigit Span tests were loaded on the auditory factorbecause they require auditory processing. All other testswere loaded on the other factor. Model 3 considered thatthe variables would split on a "linguistic/nonlinguistic"dimension. Nonlinguistic tests did not require spoken orwritten language and included all music tests plus twononmusic tests, the Wisconsin Card Sorting Test andFigural Memory. Linguistic tests included Similarities,Digit Span, and the A bstraction and V ocabulary subtestsof the Shipley Institute of Living Scale. Model 4, "abstrac-tion versus nonabstraction," considered that bo th sense oftonality and nonm usic abstraction involve the same under-lying abstraction ability. This model therefore suggestedthat both the tonali ty and nonmusic abstraction testswou ld load on on e factor , while the nonabstraction tests(pitch memory and nonmusic nonabstraction) would loadon a second factor.Models 5,6 , and 7 explored still other p artitions of thedata "probe-tone versus non probe-tone versus n on-


12/20

Dissociation 327TABLE 5Congruences of observed and random orthogonal procrustean rotated factor loadings with hypothesized target models forfull and partial correlation solutions

Target model

1. MusicNonmusic2. Auditory

Nonauditory3. LinguisticNonlinguistic4. Abstraction

Nonabstraction5. Music non-probe-tone

Music probe-toneNonmusic

6. MusicAbstractionNonabstraction

7. Music non-probe-toneMusic probe-toneAbstractionNonabstraction

8. General factor

Observed dataFull correlation

0.970.900.920.750.860.750.800.510.740.590.890.960.750.520.720.590.750.520.92

CongruencesRandom data'

Partial1' correlation Full correlation0.970.890.900.730.860.750.810.500.770.710.900.970.690.430.760.710.730.500.90

0.650.610.680.570.670.620.690.580.620.710.700.710.680.660.720.720.730.750.49

Partialb correlation0.650.610.680.570.670.620.690.580.620.710.700.710.680.660.720.720.730.750.49

"(upper 95% confidence interval)'Partial correlation removes the effect of music training.

music" (Model 5), "music versus nonmusic abstractionversus non mu sic" (Model 6); and a four-factor solu tion,"probe-tone music versus nonprobe-tone music versusabstract ion nonmusic versus nonabstract ion nonmusic"(Model 7). Finally, Model 8, "general factor," consideredthat all tests would load on a single factor.

The congruence of each of these eight hypothesizedtarget models with the obtained rotated factor loadings ispresented in the left-hand columns of Table 5. Model 1,"music vs nonmusic," achieved the highest level of congru-ence for the full correlation and for the partial corre lationmatrices. Ne xt strongest supp ort em erged for Mo del 8, the"general factor" mo del. To d etermine w heth er the targetedrotations had capitalized on chance, the magnitude of thecongruences was evaluated by undertaking 100 paral lelanalyses on data matrices of equal order but comprisingrandom normal deviates. The upper 95% confidenceinterval for the congruences of correspon ding ran dom datasets, show n in the right-hand colu mn s of Table 5, indicatesthat the congruences for the observed data were no t likelyto have occurred by chance.SUMMARY O F STAGES 1 AND 2 ~The first stage of this study sought convergent evidence

for the validity of measures of sensitivity to tonal struc-ture, collected from the same sample of participants.Several music tests employing different method ologies andrespo nse dema nds w ere all successful in assessing partici-pants' sense of tonality evaluated against quantifiedpredictors derived from music theo ry. T he second stage ofthe study involved principal component analysis andmodel testing. It revealed that the tonality tests, alongwi th a test of pitch m em ory , loaded o n a different factorfrom tests of nonmusic cognit ive ski l ls , even when thecontribution of levels of music training was statisticallyremoved. According to the two-factor solut ion , the senseof tonality was not associated with the ability to abstractin several nonmusic tests, nor was dissociation foundbetween audi tory/nonaudi tory or be tween l inguis t ic/nonlinguistic factors. A general intelligence model wasthe next-best fitting model following the music/nonmusicmodel .

Stage 3 The Amusic ParticipantThe aim of Stage 3 was to examine data from a singleamusic participant, C.N., to determine whether the selec-tive sparing and deficit in C.N.'s abilities wer e conv ergen twith the music/nonmusic dissociat ion found for the


13/20

328 Steinke, Cuddy, and HoldenTABLE 6Test scores from C.N.; mean test scores and ranges from six matchedcontrolsTest

Music testsProbe-tone Melody

Familiar M elodiesNovel MelodiesTonal/Atonal MelodiesChord ProgressionsPitch Memory (/48)Nonmusic testsWisconsin Card Sorting (/100)Similarities (WAIS-R) (/30)Digit Span (WAB-R) (/26)Figural Me mo ry (WMS-R) (/10)

.53

.73

.84

.43

.5337

89.222.714.86.5

ParticipantsControls

(.3S-.78)(.65-.76)(.68-.94)(.24-.S7)(J2-.64)(33-42)

(84-93)(22-26)(13-19)(4-9)

C N .

.24

.23

.46

.15

.2429

962410

5No te. Scores for music tonality tests represent correlation of participantratings with music - theore tic levels of tonality. Ranges are presented inparentheses.

general sample. Mc Closkey (1993), while cautioning aboutthe extent to which generalizations can be made fromneurologically compromised to normal functioning,nevertheless values "bring[ing] to b ear data from mu ltiplesingle-patient studies in formulating and evaluatingtheories" (p. 728). Indeed, it has been argued that neuro-psychological data must be explored only w ith regard toan explicit or implicit model of the normal cognitivesystem (Peretz, 1993b).METHODParticipantC.N. was referred to us by Isabelle Peretz, University ofMontreal . C.N. was a 40 year-old woman with a pureamusic disorder follow ing successive brain surgeries to clipaneurisms in her right (in 1986) and left (in 1987) middlecerebral arteries. Patel , Peretz, Tramo, and Labreque(1998) provid e a lesion profile of C.N., including a CT scanimage that revealed bilateral temporal lobe lesions.Primary auditory cortex was spared.

C.N. was right-handed, French-speaking and had 15years of formal education. She had no musical training,but used to sing every day to her child.After surgery, C.N. scored within th e normal range o ntandardised tests of intelligence, speech comprehension,and speech expression, but complained, however, ofexclusively music-related symptoms. Peretz, Kolinsky,Tramo, Labreque, Hublet , Demeurisse, and Bellevil le(1994) demonstrated in C.N. an auditory dissociationbetween music and no nm usic st imuli perception ofunes, prosody, and voice recognition was impaired, but

perception of speech and environmental sounds waspreserved. Peretz and Kolinsky (1993) also demonstratedin C.N. a dissociation between the processing of melodicand rhythmic information. The amusic disorder thusappeared to be one of atonalia.Testing ProceduresIn one session, C.N. was tested on six of the music testsfrom Stage 1 and o ne no nm usic test from Stage 2. AFrench speaker, Isabelle Peretz, translated test instructionsand recorded C.N.'s responses. The order of presentationof music tests and the single nonmusic test was random-ized for C.N., and she was randomly assigned to one of thethree random orders of trials for each of the tonalitymeasures. Music sequences were reproduced by a AiwaXK-009 Excelia cassette player t hro ug h Epo s speak ers.

The single nonmusic test was the Wisconsin CardSorting Test. Scores from a previous administration (inFrench) of the Similarities and Digit Span subtests of theWAIS-R, and th e Figural M em ory subtest of the WechslerMemory Scale-Revised were also available (Peretz et al.,1994). Published test proto cols we re followed in adm inis-tering each of these tests. The Vocabulary and Abstractionsubtests of the Shipley Institute of Living Scale were notadministered to C.N. because they were not available inFrench.

Th e data for six con trol participants w ere selected fromthe data obtained in Stages 1 and 2. These participantswere matched to C.N. for age, sex, handedness, years offormal education, and level of music training. None,however, w as French-speaking.RESULTS AND DISCUSSIONCN .'s scores on the m usic and non mus ic tests, and scoresfor the matched controls (mean and range) are presentedin Table 6. Al tho ugh it is difficult t o inte rpret the resultsof any single test because of the problem of reliability ofa single score, it is clear tha t the re is an overall patte rn inthe data. For all music tests, C.N.'s scores were consis-tently lower than the lowest score obtained from controlparticipants. For the nonmusic tests, however, C.N. 'sscores were comparable to t he scores of the co ntro l parti-cipants. These nonmusic tests scores are also consistentwith normative data for the tests. Th e results suggest th atthe selective loss experienced b y C.N. is convergent withthe statistical solution obtained for the general sample.

General DiscussionTh ree general findings address the purpo ses of this s tudy .First, sensitivity to five levels of tonal structure wasdemonstrated for a variety of test measures and partici-pants' musical backgrounds. The data support Krum-hansl's (1990b) defence of th e reliability and va lidity of th eprobe-tone m etho d and, in addition, provide evidence that


14/20

Dissociation 329sensitivity to levels of tonality in each test converged withmusic-theoretic descriptions of levels of tonality. The dataillustrate a trend for more highly trained participants torate the more tonal st imuli higher and the less tonalstimuli lower tha n th e oth er participants. For each level ofmusic training, however, sensitivity to tona lity was foundfor each tonality test. This finding docu ments the sensitiv-ity of the musical novice to tonal structure (see alsoCuddy & Badertscher, 1987) thus adding informationabout a population under-represented in music science(Smith, 1997). Moreover, it suggests that participantsshared a com mo n representation of tonali ty independentof the level of music training and the type of musicalcontext.

Second, factor analyses and model testing of the datacollected from the general sample revealed that th e m usicvariables reflecting sensitivity to to nal struc ture and pitchmemory dissociated from variables reflecting nonmusiccognitive skills. Results implicating dissociation werefound both when the contribution of levels of musictraining was included and when it was statistically re-moved from the analyses. A general intelligence modelwas the next best-fitting model over other candidatemodels in which alternative partitions of the data wereevaluated.

Third, the performance of a neurologically compro-mised participant, C.N., revealed a pattern of selective losswith respect to controls that was consistent with the two-factor (music vs nonmusic) solution for the generalsample. Th e results verified and e xtended earlier findings(Peretz et al., 1994) to new tonality tests. C.N. demon-strated a consistent loss of a sense of ton ality w ith preser-vation of nonmusic cognitive abilities.

The music factor isolated in this study involves theprocessing of pitch and pitch relations. This result m ay beconsidered in the l ight of the proposal by Peretz andMorais (1989) that tonal encoding of musical pitch fulfilsmany of the properties of modularity as proposed byFodor (1983). As noted at the outset, music shares manycharacteristics with other cognitive processes, but theprocessing of tonality does not appear to share featureswith other domains (Patel & P eretz, 1997). Evidence fromthe present study sup ports the view that tasks involvingprocesses of pitch abstraction and categorization operatewithin n eurally specialized subsystems. W hile it has beenobserved that categorization and classification are basic toall intellectual activities (Estes, 1994; Repp, 1991), thedissociations observed in the present study suggest thatcategorization ability is task-specific, and d oes no t pro ceedfrom a more generalized abil i ty, as proposed by Ashby(1992), or Anderson (1983). Tonal encoding, therefore,may be a task-specific examp le of c ategorization.

One of the present results may further address the

nature of the proposed module. Pitch memory wasassociated with sensitivity to tonal structure, despite thefact that the Pitch M em ory test was constructed to avoidtonal conventions. Possibly, despite the efforts to con-struct nontonal materials, the Pitch Memory test never-theless engaged tonal know ledge. O n th e othe r hand , thePitch Memory test may have involved low-level categori-zation that assigned pitches along a continuous sensorydimension to discrete steps of the chromatic scale. In thelatter case, the mod ule isolated by the factor solution maybe a perceptual mechanism at the "front-end" of tonalprocessing. We discuss these two possibilities in turn.

The first possibility is that, despite the nontonalconstruction of the pitch sequences, the Pitch Memorytest did engage participants' tonal knowledge. Participantsmay have attempted to assimilate the nonton al informa-tion to a tonal schema, hearing the sequences as tonalmelodies with "wrong notes." Such a possibility, howev er,is not strongly supported by the pitch memory studiescited earlier . When tonal and nontonal st imulus condi-tions are compared, large differences in performanceaccuracy are consistently reported with significantlypoorer performance for nontonal materials. If tonalknowledge is engaged as a strategy to encode an d remem -ber no nto nal m aterials, it is not very effective.

Participants may, howev er, have differed in the degreeto w hich a tonal strategy was applied. Sub stantial individ-ual differences w ere reporte d b y Kru mh ansl, Sandell, andSargeant (1987) in a probe-tone study evaluating responsesto excerpts of serial (twelve-tone) music. Among thedifferences revealed by analyses of probe-tone ratings wasthe extent to w hich tonal (key) implications of the excerptwere present in the ratings. Thus, participants in thepresent study ma y have varied in their attem pts t o assimi-late the nonton al sequences to a tonal framew ork. W hatis no t yet clear, how ever, is that th is kind of assimilationhas a facilitating effect on performance.

The second possibil ity is that th e Pitch M em ory testsand the tonality tests shared low-level processes of catego-rization and a common sensit ivity to the distr ibution ofpitch categories. The common sensitivity reflects anattunement to the regulari t ies in the auditory environ-me nt (Bregman, 1990,1993). Krum hansl (1987,1990a) hasprovided statistical evidence of the close correspondencebetween the prominence of tones in the tonal hierarchyand the frequency w ith w hich these tones are sounded inmusic. O ram and C udd y (1995; see also Cu ddy , 1997)demo nstrated that for non tona l musical contexts, listenersuse the surface prop erties of the p itch distr ibution thefrequency with which tones are sounded to construct ahierarchical organization of pitch structure. Moreover,responsiveness to the surface properties supersededassimilation to the ton al hierarchy.


15/20

330 Steinke, Cuddy, and HoldenIt is impo rtant t o no te that this second possibility doesnot imply that all tests merely reflected rote memory forth e stimulus materials. N o t all findings for tonal ity testscan be accounted for in ter ms of the surface pro perties ofthe musical context. R ather, th e poin t is that high sensitiv-ity to pitch distribution, which would be advantageous ina pitch memory task, would also facilitate the acquisition

of the tonal knowledge and its application in the tonalitytests. The more precisely one is able to preserve thedistr ibutional pitch properties of the tonal contexts, themo re effectively tona l knowledge may be internalized an dapplied. Dam age to this mechanism wo uld lead to imp air-men t of tona l processing, as in the case of C.N.Finally, we briefly discuss the finding that under thetwo-factor solution the ton ality tests dissociated from theonm usic abstraction tests, despite the sharing of descrip-ive characteristics. Descriptions of tona l processing ofteninclude properties of abstraction such as prototypes,categorization, classification, and expectancy. Perhaps the

description of shared characteristics is incorrect, orisleading. An alternative interpretation, however, isor the two-factor mu sic/non mu sic m odel was the

each other b ut also with all nonm usic cognitive skillsd. A general intelligence factor m ay acco unt for ho w

Considerations of tw o of the questions arising from ou rest first, a pitc h proc essing m odu le of a special-

en an approp riately parsed perceptual environ men t, we

ition is the inte rpol ation , first of modularize d

The l imitations of the present study preclude conclu-ve statements abou t Fod or 's (1996) notion s. The stud yof othe r musical doma ins such as rhy thm , dynam ics

al sample, tho ugh d rawn from a popu lation mo re

healthy, young adult , and educated.Th e idea offered here for further testing, nevertheless,is that the processing of discrete musical pitches engages alow-level, dom ain-specific, process one that at tunes topitch regularities. Such a proposal is compatible withDeliege (1995; see also Melen & Deliege, 1995) whosuggests that a hierarchy of modules may be involved.

Deliege (1995) argues that a higher-level m odu le carryingout a process of "cue abstraction" is integral to the pro-cesses of classification and com parison necessary for m usiclistening, and suggests that tonal encoding of pitch, likerhythmic organization, may be examples of cues ab-stracted by modules situated within such a higher-levelmodule .Th e domain-specific process we propose is no t depend-ent on music training and may not reflect the culturalattributes of the Western, or for that matter, any particu-lar tonal system. It therefore can be engaged pansty-listically at early stages of musical apprehension. As well,

however, th e processing of mu sical pitch involves generalproperties of thought including those that reflect interac-tions with the cultural/ l inguistic environment. Thesegeneral properties of thought share resources with multi-ple oth er skills and abilities.The data for Stages 1 and 2 were prese nted in a thesis sub mittedin partial fulfilment of the MA degree by W.R. Steinke, under thesupervision of L.L. Cud dy. Research was supported by a researchgrant to L.L .C. and a postgraduate fellowship to W.R .S. from theNatu ral Sciences and Engineering Research Coun cil of Canada.We thank I. Peretz, University of Montreal, for support,encouragement, and the opportunity to test the late C.N., -whosepatience and co-operation contributed much.

We acknowledge continued support of C. L. Krumh ansl andmany productive discussions. Preliminary results were presentedat meetings of the Acoustical Society of America (1993), theSociety for Music Perception and Cognition (1993), and theInternational Conference for Music Perception and Cognition(1994). Correspondence may be sent to L.L. C uddy, Dep artme ntof Psychology, Queen's University, Kingston, Ontario, K7L3 N 6 .ReferencesAbe, J., & Hoshino, E. (1990). Schema driven properties inmelody cognition: Experiments on final tone extrapolation

by music experts. Psychomusicology, 9,161-172.Anderson, J.R. (1983). The architecture of cognition. Cambridge,

MA: Harvard University Press.Ashby, F.G. (1992). Multidimensional models of categorization.

In F.G. Ashby (Ed.), Multidimensional models of perceptionand cognition (pp. 449-483). Hillsda le, NJ: Er lbau m.

Bardett, J.C., & Dow ling, W J. (1980). Recognition of transposedmelodies: A key-distance effect in developm ental pe rspective.Journal of Experimental Psychology: Human Perception and


16/20

Dissociation 331Performance, 6, 501-515.

Berg, E.A. (1948). A simple objective techniq ue for measuringflexibility in thin kin g. Journal of General Psychology, 39 ,15-22.

Besson, M. (1997). Electrophysiological studies of m usic process-ing. In I. Deliege & J. Sloboda (Eds.), Perception and cognitionof music (pp. 217-250). Hove, East Sussex, UK: Taylor &Francis.Bharucha, J.J., & Krum hansl, C.L. (1983). The representation ofharmonic structure in music: Hierarchies of stability as afunction of context. Cognition, 13, 63-102.

Bigand, E. (1993). Contribution s of m usic to research on hum anauditory cognition. In S. McAdams & E. Bigand (Eds.),Thinking in sound: The cognitive psychology of human audition(pp. 231-277). Oxfo rd: Clarend on Press/O xford Univ ersityPress.

Boltz, M. (1989). Perceiving the end: Effects of tonal relation-ships on melodic completion. Journal of Experimental Psychol-ogy: Human Perception and Performance, 15, 749-761.

Bregman, A.S. (1990). Auditory scene analysis: The perceptualorganization of sound. Cam bridge, MA: Th e MIT Press.

Bregman, A.S. (1993). Auditory scene analysis: Hearing incomplex environments. In S. McAdams & E. Bigand (Eds.),Thinking in sound: The cognitive psychology of human audition(pp. 10-36). Oxfor d: O xford Unive rsity Press.

Carroll, J.B. (1993). Hum an cognitive abilities: A survey offactor-analytic studies. Cambridge, UK: Cambridge U nivers i tyPress.

Croonen, W.L.M. (1994). Two ways of defining tonal strengthand implications for re cognition of tone series. Music Percep-tion, 13 ,109-119.

Cuddy, L.L. (1991). Melodic patterns and tonal structure:Converging evidence. Psychomusicology, 10, 107-126.

Cuddy, L.L. (1993). Melody com prehension and to nal structure.In T.J. Tighe & W.J. Dowling (Eds.), Psychology and music:The understanding of melody and rhythm (pp. 19-38). H illsdale,NJ: Erlbaum .

Cuddy , L.L. (1997). Ton al relations. In I. Deliege & J. Sloboda(Eds.), Perception and cognition of music {pp. 329-352). Hove,East Sussex, UK: Taylor & Francis.

Cuddy, L.L., & Badertscher, B. (1987). Recovery of the tonalhierarchy: Some com parisons across age and levels of musicalexperience. Perception & Psychophysics, 41 , 609-620.

Cuddy, L.L., Cohen , A.J., & M ewh ort, D.J.K. (1981). P erceptionof structure in shor t melodic sequences. Journal of Experimen-tal Psychology: Human Perception and Performance, 7, 869-883.

Cud dy, L.L., Coh en, A .J., & M iller, J. (1979). Melody recogni-tion: The exp erimental application of musical rules. CanadianJournal of Psychology, 33,148-157.

Cuddy, L.L., & Lyons, H.I. (1981). Musical pattern recognition:A comparison of listening to and studying tonal structure andtonal am biguities. Psychomusicology, 1(2), 15-33.

Cuddy, L.L. , & Thompson, W. F. (1992). Asymmetry ofperceived key movement in chorale sequences: Converging

evidence from a probe-tone analysis. Psychological Research/Psychologische Forschung, 54 , 51-59.

Davidson, L., & Scripp, L. (1988). Young children's musicalrepresentations: Windows on music cognition. In J.A.Sloboda (Ed.), Generative processes in music (pp. 195-230).Oxford: C larendon Press.

Deliege, I. (1995). The two steps of the categorization process inmusic listening: An approach of the cue extraction mecha-nism as mod ular system. In R . Steinberg (Ed.), Music and themind machine (pp. 63-73). Berlin He idelberg: Springer-Verlag.

Deutsch, D . (1970). Ton es an d num bers: Specificity of interfer-ence in short-term m emo ry. Science, 168, 1604-1605.

Deutsch, D. (1972). Effect of repetition of standard and compari-son tones on recognition memory for pitch. Journal ofExperimental Psychology, 93, 156-162.

Deutsch, D. (1978). Delayed pitch comparisons and the principleof proximity. Perception & Psychophysics, 23, 227-230.

Dewar, K.M., Cuddy, L.L., & Mewhort, D.J.K. (1977). Re cogni-tion memory for single tones with and without context.Journal of Experimental Psychology: Human Learning andMemory, 3, 60-67.

Dow ling, W.J. (1991). Ton al strength and melody recognitionafter long and short delays. Perception & Psychophysics, 50,305-313.

Dowling, W.J., & Bartlett, J.C. (1981). The importance ofinterval information in long-term memory for melodies.Psychomusicology, 1(1), 30-49.

Dowling, W.J . , & Harwood, D.L. (1986). Music cognition.Tor onto: Academic Press.

Estes, W.K. (1994). Classification and cognition. Oxford: OxfordUniversity Press.

Fod or, J. (1983). The modularity of mind Cambridge, MA: MITPress.

Fodor, J. (1996). It's the thought that counts [Review of the bookThe prehistory of the mind]. London Review of Books, 18 (23),22-23.

Frankland, B.W., & Cohen, A.J. (1996). Using the Krumhansland Schmuckler key-finding algorithm to quantify the effectsof tonality in the interpolated-tone pitch-comparison task.Musk Perception, 14, 57-83.

Frances, R. (1988). The perception of m usic. (W.J. Dowling,Trans.). Hillsdale, NJ: Erlbaum. (First French edition pub-lished in 1958)

Gardner, H. (1983). Frames of mind: The theory of multipleintelligences. NY: Basic Books I nc.

Gross, R. (1981). DX-Score [Computer program]. Rochester, NewYork: Eastman School of Music.

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

Harman, H.H. (1976). Modem factor analysis (3rd edition,revised). Chicago: Unive rsity of Chicago Press.

Heaton, R.K. (1981). A manual for the Wisconsin Card SortingTest. Odessa, FL: Psychological Assessment Resources.

H o rn , J.L. (1965). A rationale and test for th e nu mb er of factors


17/20

332 Steinke, Cuddy, and Holdenin factor analysis. Psychometrika, 30,179-185.

Howe, M.J.A. (1990). The origins of exceptionalabilities. Oxford-Basil Blackwell Ltd.

Jackendoff, R. (1987). Consciousnessan d the computational mind.Camb ridge, MA: MIT Press.

Johnston, J.C. (1985). The perceptual salience of m elody andmelodic structures. Unpublished B.A. thesis, Queen's Univer-s i ty , Kingston, On tario , Canada.Jones, M.R. (1981). Music as a stimulus for psychologicalmotion: Part I . Some determinants of expectancies.Psychomusicology, 1(2), 34-51.

Jones, M.R. (1982). Music as a stimulus for psychologicalmotion: Part II. An expectancy model. Psychomusicology, 2 (l),1-13.

Jones, M.R. (1991). Preface. Psychomusicology, 10 , 71-72.Judd, T. (1988). Th e varieties of musical talent. In L.K. Ob ler &

D . Fein (Eds.), The exceptional brain: Neuropsychology of talentand specia l abilities (pp. 127-155). N ew Yor k: Guilford Press.

Krumhansl, C.L. (1979). The psychological representation ofmusical pitch in a tonal context. Cognitive Psychology, 11,346-374.

Krumhansl, C.L. (1987). Tonal and harmonic hierarchies. In J.Sundberg (Ed.), Harmony and tonality (pp 13-32). Stock holm ,Sweden: Royal Swedish Academy.

Krumhansl, C.L. (1990a). Cognitive foundations of musical pitch.New York: Oxford Univers i ty Press .

Krumhansl, C.L. (1990b). Tonal hierarchies and rare intervals inmusic cognition. Music Perception, 7, 309-324.

Krum hansl, C.L., Bha rucha, J.J., & CasteUano, M.A. (1982). Keydistance effects on perceived harmonic structure in music.Perception & Psychophysics, 32, 96-108.

Krum hansl, C.L., Bha rucha, J.J., & Kessler, E J . (1982). Perceivedharmonic structure of chords in three related musical keys.Journal of Experimental Psychology: Human Perception andPerformance, 8, 24-36.

rumhansl, C.L., & CasteUano, M.A. (1983). Dynamic processesin music percept ion. Memory & Cognition, 11, 325-334.

rum hansl, C.L., & Kessler, EJ . (1982). Tracing the dynam icchanges in perceived tonal organization in a spatial represen-tat ion of musical keys. Psychological Review, 89, 334-368.

rumhansl, C.L., Sandell, G., & Sergeant, D. (1987). Theperception of tone hierarchies and mirror forms intwelve-tone serial music. Music Perception, 5, 31-78.

R. N. (1979). Quantification of thehierarchy of tonal functions within a diatonic context.Journal of Experimental Psychology: H uman Perception andPerformance, 5, 579-594.

G., Ada m, C , & Schupp, R. (1990). C-LAB: NotatorSL/CreatorSL (Version 3. l ) [Computer p rog ram] . Hamburg ,Germany : C-LAB Sof tware GmbH.

Music Perception, 5,315-349.

Jackendoff, R. (1983). A generative theory of tonalmusic. Cam bridge, MA: MIT Press.

Longman, R.S. , Cota, A.A., Holden, R.R., & Fekken, G.C.(1989). A regression equation for the parallel analysis crite-rion in principal components analysis: Mean and 95thpercentile eigenvalues. Multivariate Behavioral Research, 24,59-69.

Lynch, M.P., Short, L.B., & Chua, R. (1995). Contributions ofexperience to the development of musical processing ininfancy. Developmental Psychobiology, 28 , 377-398.Marin , O.S.M. (1982). Neurological aspects of music perceptionand performance. In D . Deutsch (Ed.), Th epsychology of music(pp. 453-477). N ew Yor k: A cademic Press.

McCloskey, M. (1993). Theory and evidence in cognitiveneuropsychology: A "radical" response to Robertson, Knight,Rafal, and Shimamura (1993). Journal of Experimental Psychol-ogy: Learning, Memory, and Cognition, 19, 718-734.

Melen, M ., & Deliege, I. (1995). Extrac tion of cues or u nde rlyin gharmonic structure: Which guides recognition of familiarmelodies? European Journal ofCognitive Psychology, 7, 81-106.

Meyer, L.B. (1956). Emotion and meaning in music. Chicago:Unive rsity of Chicago Press.Or am , N ., & Cudd y, L.L. (1995). Responsiveness of Westernadults to pitch-distributional information in melodic se-quences. Psychological Research/Psychologische Forschung, 57 ,103-118.

Patel, A., & Peretz, I. (1997). Is music autonomous from lan-guage? A neuropsychological appraisal. In I. Deliege & J.Sloboda (Eds.), Perceptionan d cognition of music (pp. 191-215).Ho ve, E ast Sussex, UK: Tay lor & Francis.

Patel, A., Peretz, I., Tra mo , M ., & Labreque , R. (1998). Process-ing prosodic and musical patterns: A neuropsychologicalinvestigation. Brain & Languag e, 61, 123-144.

Peretz, I. (1993a). Auditory atonalia for melodies. CognitiveNeuropsychology, 10, 21-56.Peretz , I. (1993b). Au dito ry agnosia: A functional analysis. In S.

Me Adams & E. Bigand (Eds.), Thinking in sound: The cogni-tive psychology of human audition (pp. 199-230). Oxford:Clarendon Press /Oxford Univers i ty Press .

Pe retz , I., & Ko linsky , R. (1993). Boun daries of separab ilitybetween melody and rhythm in music d iscriminat ion: Aneuropsychological perspective. Quarterly Journal ofExperi-mental Psychology, 46, 301-325.

Peretz, I ., Kolinsky, R. , Tram o, M., Labrecque, R., Hublet , C ,Demeurisse, G., & Belleville, S. (1994). Functional dissocia-tions following bilateral lesions of auditory cortex. Brain,117,1283-1301.

Peretz , I., & M orais, J. (1989). Music and mo dula rity. Contempo-rary Music Review, 4, 279-294.

Piston, W. (1987). Harmony (Revised and expanded by M.DeVoto ) . Ne w York : Nor ton .

Polk, M., & Kertesz, A. (1993). Music and language in degenera-tive disease of the brain. Brain & Cognition, 22, 98-117.

Repp, B.H. (1991). Some cognitive and perceptual aspects ofspeech and music. In J. Sundberg, L. Nord & R. Carlson(Eds.), Music, language, speech and brain (pp. 257-268). Lon-


18/20

Dissociation 333don: M acmillan.

Samson, S., & Zatorre, RJ. (1994). Contribution of the righttemporal lobe to musical timbre discrimination.Neuropsychologia, 32 , 231-240.

Shepard, R.N . (1964). Circularity in judgments of relative pitch.Journal of the Acoustical Society of America, 36, 2346-2353.

Shipley, W.C. (1953). Shipley-Institute of Living Scale formeasuring intellectual impairment. In A. Weider (Ed.),Contributions toward medical psychology: Theory andpsychodiagnostic methods (Vol. 2, pp. 751-756). New York:The Ronald Press Company.

Shuter-Dyson, R., & Gabriel, C. (1981). Th e psychology of musicalability (2nd edition). London: M ethuen.

Sloboda, J.A. (1985). The musical mind: The cognitive psychologyof music. Oxford: Clarendon Press.

Smith, J.D. (1997). The place of musical novices in music science.Music Perception, 4, 227-262.

Smith, J.D., & Melara, R.J. (1990). Aesthetic preference andsyntactic prototypicality in m usic: Tis the gift to be simple.Cognition, 34, 279-298.Spearman, C. (1904). "General intelligence," objectively deter-mined and measured. American Journal of Psychology, 15,201-293.

Spearman, C. (1927). The abilities of man: Their nature andmeasurement. New York: Macmillan and Company Ltd.[Reprinted: New York: AMS Publishers, 1981].

Steinke, W.R. (1992). Musical abstraction and non musical abstrac-tion abilities in musically trained and un trained adults. Unpub-lished M.A . Thesis, Queen's Un iversity, Kingston, Ontario,Canada.

Steinke, W.R., Cudd y, L.L., & Jacobson, L.S. (1996). Melody andrhythm processing followingright-hemisphere stroke: A casestudy. In B. Pennycook & E. Costa-Giomi (Eds.), Proceedingsof the 4th International Conference on Music Perception an dCognition (pp. 323-325). Montreal: McGill University.

Sternberg, R.J., & P owe ll, J.S. (1982). Theories of intelligence. InRJ. Sternberg (Ed.), Handbook of human intelligence (pp.

975-1006). Ne w York: Cambridge University Press.Thompson, W.F., & Cuddy, L.L. (1989). Sensitivity to key

change in chorale sequences: A com parison of single voicesand four-voice harm ony. Music Perception, 7,151-168.

Trehub, S.E., & Trainor, L.J. (1993). Listening strategies ininfancy: The roots of mu sic and language develop me nt. In S.McAdams & E. Bigand (Eds.), Thinking in sound (pp.278-327). Oxford: Clarendon Press.Vernon, P.E. (1950). The structure of human abilities. London:Methuen.

Warren, R.M. (1993). Perception of acoustic sequences: Globalintegration versus temporal resolution. In S. McAdams & E.Bigand (Eds.), Thinking in sound {pp. 37-68). Oxford: C laren-don Press.

Waterhouse, L. (1988). Speculations on the substrate of specialtalents. In L.K. O bler & D . Fein (Eds.), The exceptional brain:Neuropsychology of talent and special abilities (pp. 493-512).New York: Guilford Press.

Wechsler, D. (1981). Manual for the W echsler Adult IntelligenceScale-Revised. New York: The Psychological Corporation.Wechsler, D. (1987). Wechsler Memory Scale-Revised Manual N ewYork: The Psychological Corporation.

Zachary, R.A. (1986). Shipley Institute of Living Scale-Revisedmanual. Los A ngeles: Western Psychological Services.

Zatorre, R.J., Evans, A. C. , & Meyer, E . (1994). Neural m echa-nisms underlying melodic perception and m emo ry for pitch.Journal ofNeurosdence, 14, 1908-1919.

Zatorre, R.J., Halpern, A .R., P erry, D.W ., M eyer, E., & Evans,A.C. (1

Documents

Steinke, W. R., Cuddy, L. L., & Holden, R. R. (1997). Dissociation of Musical Tonality and Pitch Memory From Nonmusical Cognitive Abilities. Canadian Journal of Experimental Psychology,