Music Perception © 1998 by THE REGENTS OF THE Summer 1998, Vol. 15, No. 4, 357-389 UNIVERSITY OF CALIFORNIA Musical Abilities in Individuals With Williams Syndrome DANIEL J. LEVITIN Interval Research Corporation and Stanford University URSULA BELLUGI The Salk Institute for Biological Studies ‘We report evidence for relatively preserved musical rhythm processing in individuals with Williams syndrome, supporting the theory that musical ability constitutes an independent intelligence. Williams syndrome oc- curs in 1 out of 20,000 births and is associated with a defect in elastin production, impaired cognitive function, and poor spatial, quantitative, and reasoning abilities, coupled with excellent face processing and rela- tively strong language abilities in adolescents and adults. Previously, in- formal qualitative observations have revealed an unusual degree of mu- sicality and engagement with musical stimuli in many individuals with Williams syndrome. In the present study, rhythm production was assessed for eight subjects with Williams syndrome and eight subjects in a com- parison group by using an echo clapping task. Despite serious deficits in other cognitive domains and generally poor coordination, the subjects with Williams syndrome achieved accuracy scores equivalent to those of subjects in the comparison group and demonstrated equivalent abilities in meter change and beat maintenance. Most interestingly, when the sub- jects with Williams syndrome made errors in rhythm production, their errors were far more likely than comparison subjects’ errors to form rhyth- mically compatible musical elaborations to the test items; that is, responses of subjects with Williams syndrome, when incorrect, tended to be cre- ative extensions of the reference rhythm. F or much of its history, experimental psychology has viewed intelligence as somewhat monolithic, and mental retardation as reflecting more or less uniform impairment across the various domains of cognitive function- ing. The study of distinct, well-defined, and atypical populations affords a unique opportunity to investigate specific aspects of cognition. In particu- lar, the study of populations with genotypic abnormalities (including Down Address correspondence to Daniel J. Levitin, Center for Computer Research in Music and Acoustics (CCRMA), Department of Music, Stanford University, Stanford, CA, 94305. (e-mail: levitin@ccrma.stanford.edu) 357358 Daniel J. Levitin & Ursula Bellugi syndrome, autism, and Williams syndrome) is useful in providing evidence for the debate over the independence of specific mental functions as inves- tigators seek to establish the degree to which various cognitive abilities are correlated with, or can be decoupled from, one another. Williams syndrome is a rare developmental disorder of genetic origin characterized by mental retardation and an unusual pattern of cognitive abilities and deficits. For more than a decade, one of us (Bellugi) and col- leagues at the Salk Institute have pursued a comprehensive and systematic program of research on Williams syndrome using a wide variety of cogni- tive neuroscientific approaches (Bellugi, Klima, & Wang, 1996; Bellugi, Lai, & Korenberg, in press; Bellugi, Lai & Wang, 1997; Bihrle, Bellugi, Delis, & Marks, 1989; Galaburda, Wang, Bellugi & Rossen, 1994; Jernigan & Bellugi, 1994; Reilly, Klima & Bellugi, 1990; Wang & Bellugi, 1994; Wang, Hesselink, Jernigan, Doherty, & Bellugi, 1992). The unusual pattern of abilities and disabilities found in Williams syn- drome challenges traditional ideas about neural organization and architec- ture and promises to inform the debate over multiple intelligences (Gardner, 1983; Gazzaniga, 1988). The emerging picture of Williams syndrome is that despite severe deficits across a broad class of cognitive domains (e.g., quantitative, reasoning, spatial skills), individuals with Williams syndrome exhibit disproportionately preserved language, face processing, and per- haps also musical abilities. To date, the anecdotal claims of unusual musi- cal ability in individuals with Williams syndrome have not been systemati- cally tested, and that is the focus of our ongoing research program, the initial findings of which constitute the present preliminary report. Background on Williams Syndrome Williams syndrome was discovered independently by Fanconi (1952) and the British cardiologist Williams (Williams, Barratt-Boyes, & Lowe, 1961). The syndrome occurs in approximately 1 in 20,000 births and is character- ized by the deletion of one copy of a small set of genes on Chromosome 7 (7q11.23), including the genes for elastin, LIM1kinase, Frizzled, WSCR1, and Syntaxin1A, etc. (Korenberg et al., 1997). PHYSICAL MANIFESTATIONS The physical manifestations of Williams syndrome include a specific heart defect (narrowing of the aorta), a defect in the production of elastin, hypercalcemia, and curvature of the spine. Williams syndrome is further characterized by a specific physiognomy, including “elfin” or “pixie-like” facial features: full lips, high and prominent cheekbones, a broad forehead,Musical Abilities in Individuals With Williams Syndrome 359 flat nasal bridge, wide mouth, upturned nose, stellate pattern in the iris, and almond-shaped eyes. Scattered and informal references in the Euro- pean medical and popular literature for the last several hundred years to “elfin” syndrome might have been descriptions of individuals with Will- iams syndrome before its identification this century (Lenhoff, Wang, Greenberg, & Bellugi, 1997). A few cases of reproduction involving indi- viduals with Williams syndrome have been reported, and the heritability of the disease appears to be 0.5. Williams syndrome is also characterized by unusually poor spatial coor- dination and lower than normal digit independence between the third and fourth fingers. Many individuals with Williams syndrome have difficulty tying their shoes, hanging clothes on a hanger, making change, or walking up and down stairs. COGNITIVE MANIFESTATIONS IQs of individuals with Williams syndrome range from 40 to 100 (mean, ~60) , accompanied by commensurate impairments in spatial, quantitative, and reasoning abilities. Severe deficits have been documented for individu- als with Williams syndrome on conceptual reasoning tasks, with most such individuals failing Piagetian tests (Piaget, 1963) for conservation of num- ber, space, substance, weight, and quantity (Bellugi et al., 1996). There have been reports of hyperacusis (Nigam & Samuel, 1994; Klein, Armstrong, Greer & Brown, 1990), but these are difficult to interpret. Loud noises such as thunder cause gross distress in some children with Williams syndrome, and many such children are also extremely sensitive to broad- band sounds that are not loud, such as those made by vacuum cleaners, fans, helicopters, and other motors (Levitin & Bellugi, 1997a, 1997b). Sounds such as these, which normal individuals can often screen out, can be very distracting to many individuals with Williams syndrome, who seem to be unable to tune them out. Many individuals with Williams syndrome are also unusually attracted to these types of sounds and actively seek them out. (According to parental reports, one child with Williams syndrome has a passion for vacuum cleaners and wants nothing else for Christmas; the child now owns 18 of them. Another is so fascinated with the sound of lawn mowers that he begged his parents for one and now mows all the lawns in the neighborhood. Another child expresses a fascination with the sound of leaf blowers. Some parents claim their children can identify the specific models of automobiles and vacuum cleaners solely on the basis of the sound of their motors.) From these descriptions, we prefer to use the term “soundscape sensitivity” (cf. Schafer, 1969) to characterize the un- usual interest in timbre among some individuals with Williams syndrome and their reported abilities to classify sounds on the basis of timbre. Fur-360 Daniel J. Levitin 8 Ursula Bellugi ther details about the nature and origin of these auditory sensitivities re- main a mystery and these topics are the subject of current investigation in our laboratory. The most striking cognitive feature of individuals with Williams syn- drome is that they exhibit better face processing and larger vocabularies than would be expected given their other cognitive deficits. Typical adoles- cents and adults with Williams syndrome in fact command a large and unusual vocabulary and employ complex syntactic structures (Bellugi, Wang, & Jernigan, 1994). As noted in several papers, this neuropsychological profile provides a striking contrast to Down syndrome, in which language is compromised relative to general cognitive functioning (Bellugi et al., 1996; Mervis & Bertrand, 1997; J. F. Miller, 1987). For example, on a fluency task (“name as many animals as you can in 60 seconds”), it is not unusual for a child with Williams syndrome to offer such exotic exemplars as weasel, newt, ibex, yak, or saber-toothed tiger. When asked to tell a story illustrated by a sequence of drawings, they tell coherent, imaginative stories and frequently employ paralinguistic devices such as whispering or exaggerated prosody, and semantic “attention-get- ters” such as “lo-and-behold” or “all-of-a-sudden” or “guess what hap- pened next!” (Reilly et al., 1990). Williams syndrome is further characterized by hypersociability (Bellugi, Losh, Reilly, & Anderson, 1998). Most children with Williams syndrome are eager to talk with people, including strangers, and in fact parents say that it is difficult to teach them not to talk to strangers. One young woman with Williams syndrome told us, “There are no strangers, there are only friends.” Individuals with Williams syndrome typically strike up conversa- tions easily with anyone, are generally polite, and have learned a number of complex social mores to the extent that they may be able to carry on a conversation for a long time without the other participants suspecting that they are mentally retarded. This forms a stark contrast to both children with Down syndrome and autistic children, who are impaired in both lan- guage and social skills. One important finding is that individuals with Williams syndrome have a striking deficit in global versus local processing of visual stimuli that is the opposite of the deficit that researchers have found in persons with Down syndrome (Bihrle et al., 1989; Bellugi et al., 1997). When presented with the Wechsler Intelligence Scale for Children—-Revised (WISC-R) block de- - sign task, individuals with Williams syndrome typically have difficulty with the overall configuration, attending primarily to the local aspects of the figure, whereas individuals with Down syndrome show the reverse pattern (Figure 1). This pattern of processing is not just revealed in tasks requiring the arrangement of three-dimensional figures, but with pen and paper tasks as well. When asked to copy Navon-style hierarchical local/global prece-Musical Abilities in Individuals With Williams Syndrome 361 WISC-R Block Design 11 yrs. 11 yrs. a4 16 yrs. 18 yrs. DNS Fig. 1. Processing differences between subjects with Williams syndrome (WMS) and sub- jects with Down syndrome (DNS) suggesting local/global (hierarchical) processing deficits on the Wechsler Intelligence Scale for Children-Revised (WISC-R) block design task. dence figures with two levels of structure, individuals with Williams syn- drome typically respond with attention to local features while ignoring global features (Figure 2). For example, when shown a large letter “D” composed of smaller letter “Y’s,” the typical child with Down syndrome will draw a canonical large letter “D” using smooth lines, revealing an insensitivity to local structure (Figure 2a). A typical child with Williams syndrome may draw an assortment of letter “Y’s” sprinkled across the page, without any apparent global structure. Interestingly, though perhaps not surprisingly, the letters that the child with Williams syndrome draws are of the appropriate size. Similarly, when asked to draw an arrow com- posed of horizontal line segments, the child with Down syndrome draws a continuous arrow whereas the child with Williams syndrome draws an assortment of line segments of various lengths and without any particular global pattern (Figure 2b); analogous results are seen for a global “dia- mond” shape composed of local “plus” signs (Figure 2c). This pattern of fractionated attention to detail and failure to organize into global structures does not occur only for reproduced drawings but also for drawings made from their imagination or memory (Figure 3). We note that the bicycle drawn by the child with Down syndrome is simple but recognizable, with both good closure and form. In contrast, the bicycle drawn by an age- and IQ-matched child with Williams syndrome is highly362 Daniel J. Levitin & Ursula Bellugi i 7 a ererry tr ‘ ryt | Modet [D| fas Figure 2b Figure 2c Fig. 2. Processing differences between subjects with Williams syndrome and subjects with Down syndrome showing local/global (hierarchical) processing deficits on pen and paper copying tasks of (a) a global letter “D” composed of small letter “Y’s,” (b) a global arrow composed of local dashes, and (c) a global “diamond” shape composed of local “plus” signs. (a) and (b) reprinted with permission from Bihrle et al. (1989).Musical Abilities in Individuals With Williams Syndrome 363 pedals a person 4 wheel Pedals Fig. 3. Examples of drawings of a bicycle by a subject with Williams syndrome (WS) and a subject with Down syndrome (DS) matched for chronological age and IQ (labels are pro- vided to indicate subjects’ oral descriptions of the drawings). The drawing by the subject with Williams syndrome reveals attention to fragmented detail and the drawing by the subject with Down syndrome illustrates attention to gross configuration. fractionated, with the bike above, the person depicted upside down, the pedals off to the side, and the chain floating in the air above the person. BRAIN MORPHOLOGY Although the just-mentioned visual processing deficit observed in indi- viduals with Williams syndrome is on the surface similar to that seen in patients with damage of the right parietal lobe (Delis, Kiefner, & Fridlund, 1988; Delis, Robertson, & Efron, 1986), it does not mean that subjects with Williams syndrome have frank damage of the right side of the brain. Magnetic resonance imaging studies have not shown subjects with Will- iams syndrome to possess focal right hemispheric lesions (Jernigan & Bellugi, 1994). In fact, the evidence emerging is that the brains of individuals with Williams syndrome may be subject to very different principles of neural organization than the normal brain (Bellugi, Mills, Jernigan, Hickok, 8 Galaburda, in press). Subjects with Williams syndrome show reduced cerebral volumes rela- tive to control subjects, and slightly larger cerebral volume than subjects with Down syndrome, although the difference between subjects with Wil- liams syndrome and subjects with Down syndrome was not statistically significant (Jernigan & Bellugi, 1990). There are proportional differences between Williams syndrome and Down syndrome: brains of individuals364 Daniel J. Levitin & Ursula Bellugi with Williams syndrome are characterized by preservation of anterior re- gions, limbic system, and cerebellar areas, specifically the neocerebellum. In fact, cerebellar size of subjects with Williams syndrome was actually similar to that of normal control subjects in overall volume, with signifi- cantly increased areas of neocerebellar vermal lobules but decreased paleocerebellar vermal modules (Jernigan & Bellugi, 1994). Cytoarchitectonic analysis has revealed exaggerated horizontal organiza- tion of neurons within layers, especially within area 17; increased cell pack- ing density throughout the brain; and abnormally clustered and oriented neurons (Galaburda et al., 1994). Schlaug, Jancke, Huang, and Steinmetz (1995) discovered a leftward asymmetry in the planum temporale that was larger in musicians than nonmusicians and larger still in musicians with absolute pitch. Hickok, Bellugi, and Jones (1995) found that the planum temporale in three brains of individuals with Williams syndrome had a greater asymmetry than did the planum temporale of Schlaug’s musicians and somewhat less asymme- try than the planum temporale of Schlaug’s absolute pitch subjects. Most importantly, subjects with Williams syndrome did not differ from normal subjects in total surface area of the planum temporale despite the overall reduction of cerebral volume in individuals with Williams syndrome, sug- gesting a disproportionate growth of the posterior supratemporal region in individuals with Williams syndrome (Hickok, Neville, et al., 1995). Music and Williams Syndrome The disproportionately good language ability among individuals with Williams syndrome suggests that language may not “piggyback” on gen- eral mental function and intelligence, but may truly represent an indepen- dent faculty. We wondered to what extent musical abilities are also dispro- portionately good in individuals with Williams syndrome, suggesting that music may also comprise an independent cortical faculty or faculties. Of course, music is not a monolithic ability, any more than language is, and there are no doubt separate subsystems for the processing of pitch, melody, contour, timbre, and rhythm (Platel et al., 1997), and the specific pattern of musical abilities/deficits in individuals with Williams syndrome is unknown. In general, it is known that various musical subskills appear to be differen- tially affected by subnormal mental capacity. For example, musical tasks requiring memory or reasoning tend to be more difficult for those with subnormal IQs than are tasks that require simple discrimination (Gfeller, 1992; McLeish & Higgs, 1982). In normal populations, evidence for the separability of musical abilities from more general cognitive abilities is mixed. Isolated case studies appearMusical Abilities in Individuals With Williams Syndrome 365 in the literature describing patients who have lost musical function because of lesions (amusia) while language and other cognitive abilities remain pre- served, or vice versa (Basso, 1993; Benton, 1977; Edgren, 1895; Fasanaro, Spitaleri, & Valiani, 1990; Marin, 1982; Patel & Peretz, 1997; Patel, Peretz; Tramo, & Lebreque, 1998; Sacks, 1985; Stanzione, Grossi, & Roberto, 1990). The dissociation of musical function has been studied extensively by Peretz and Zatorre and their colleagues (Peretz, 1990, 1993, 1996; Peretz & Hébert, 1995; Peretz & Kolinsky, 1993; Peretz, Kolinsky, Tramo, & Labrecque, 1994; Samson & Zatorre, 1988, 1992, 1994; Zatorre, 1989; Zatorre, Evans, Meyer, & Gjedde, 1992; Zatorre & Halpern, 1993; Zatorre. Halpern, Perry, Meyer, & Evans, 1996). Luria (1966) found patients with severe deficits in perception and repro- duction of temporal patters (arhythmia) when damage occurred to left au- ditory secondary association areas, with melodic and timbral processing remaining preserved. Milner (1962) found deficits in tone and timbre per- ception and reproduction after right temporal lobectomy, with relatively preserved rhythm. Unfortunately, studies of brain-damaged patients can be difficult to interpret; head injuries are seldom localized to particular func- tional subsystems, and the loss of a particular ability can be due to the severing of important connections rather than damage to a specific module per se (Platel et al., 1997). Also frustrating attempts to draw general infer- ences from lesion work is the fact that no two case studies describe identi- cal impairments. For these reasons, the study of distinct populations with developmental or genotypic abnormalities can be a valuable research tool. The only formal study of music and Williams syndrome thus far was conducted by A. Don (1997), who tested 19 individuals with Williams syn- drome using the tonal and rhythm subtests of Gordon’s Primary Measures of Music Audiation (1986). “Audiation” is defined as the ability to hear music in one’s head, akin to visual imagery. Don’s subjects demonstrated relatively preserved audiation ability, at levels commensurate with their relatively preserved language. Although the musical abilities of individuals with Williams syndrome have not yet been studied further, it is perhaps useful to briefly review what is known about music among individuals with Williams syndrome, autistic individuals, and individuals with the various unspecified forms of musical savantism. First, the strong affinity for music in children with Down syndrome is well documented (Shuttleworth, 1900) and parallels anecdotal accounts of a similar affinity in children with Wil- liams syndrome. In perception tasks, Stratford and Ching (1983) found that children with Down syndrome showed no difference in rhythm dis- crimination from mental age-matched control subjects. However, in pro- duction tasks, Freeman (1986) found that individuals with Down syndrome tend toward delayed rhythmic response to musical selections and demon- strate a failure to maintain beat.366 Daniel J. Levitin & Ursula Bellugi The literature on so-called musical “savants” tends to lump together individuals with mental retardation of very different etiologies, and so it is difficult to draw general conclusions from these accounts. But the domi- nant picture is that so-called musical savants excel in melodic and har- monic tasks to the exclusion of rhythmic ones (Miller, 1989). For example, O’Connell’s (1974) case study of an autistic child found exceptional har- monic and melodic skill and an almost complete absence of rhythmic abil- ity. For these reasons, and because individuals with Down syndrome ex- hibit specific rhythmic impairments, a reasonable prediction might be that individuals with Williams syndrome may also show rhythmic impairments, the focus of the present study. The existence of musical savants has been cited as evidence for various theories of “multiple intelligences,” the most prominent being that pro- posed by Gardner (1983). Gardner argues that musical intelligence is a separate cognitive domain, and he identified eight criteria for an intelli- gence: (1) the potential isolation by brain damage, (2) the existence of idi- ots savant and prodigies, (3) identifiable core operations (for our present purposes these might be sensitivity to pitch and rhythm relations), (4) a developmental history that leads to expert performance, (5) an evolution- ary history, (6) support from experimental psychological tasks, (7) support from psychometric findings, and (8) the ability to encode the information with which an intelligence deals in a symbol system (Gardner, 1983, pp. 62-67). Jackendoff (1987) and Peretz and Morais (1989) have also made claims for the independence of musical ability. To date, empirical studies of the separability of music cognition provide mixed reports. In one study, low correlations (~0.04) were found between the Musical Aptitude Profile (Gordon, 1965/1995) and intelligence tests (Bixler, 1968), but in a meta-analysis of several studies, Shuter-Dyson and Gabriel (1981) report correlations of musical ability and other intellectual abilities as high as 0.50. Related to the multiple intelligences debate are two competing theories regarding developmental stages and artistic competence/artistic apprecia- tion. Parsons (1987) takes a Piagetian view and proposes that all children go through five stages of increasing artistic-cognitive sophistication, from “favoritism” (in which the child takes pleasure in many works of art with little cognitive discrimination) to “autonomy” (in which the child is able to adopt a more mature, and reflective attitude toward the cultural value of particular works). These artistic-cognitive stages match up in principled ways to the traditional cognitive Piagetian stages. Gardner (1973) predict- ably takes an opposing view that the types of cognitive operations outlined by Piaget are not required for artistic activity, writing that “the groupings, groups and operations described by Piaget do not seem essential for mas- tery of understanding of human language, music, or plastic arts” (p. 45). (See also Bamberger, 1991; Gardner, 1983; Gardner & Perkins, 1988.)Musical Abilities in Individuals With Williams Syndrome 367 Thus, Gardner’s theory predicts that musical abilities are possible in re- tarded populations, but the theories of Parsons and Piaget predict that the particular types of musical abilities should be limited by the cognitive stages achieved by the individuals. In particular, Piaget (Piaget, 1970; Piaget & Inhelder, 1969) claimed the child’s ability to organize temporal sequences of events develops in stages, with class inclusion, a stable durational unit, and seriation of events appearing by 5-8 years of age. Separate from these cognitive stages of development, Hargreaves (1996) proposes five phases (he deliberately avoids the loaded word “stages”) of musical development: sensorimotor, figural, schematic, rule systems, and professional, but the predictions of his model are less well defined for the case of retarded and developmentally disabled populations, and the rela- tionship (if any) between the Hargreavesian and Piagetian modal ages of onset is not clear. Against this conceptual background, we set out both to observe the gen- eral musical abilities of individuals with Williams syndrome in a natural setting (qualitative observations) and to test their rhythmic abilities more formally (quantitative observations). If the anecdotal reports are true and individuals with Williams syndrome show preserved abilities for general musical tasks (despite gross impairment on other cognitive domains), this could provide new and converging evidence for the independence of music cognition and perception from general cognitive abilities. Beyond the assessment of general musical ability in individuals with Wil- liams syndrome, we were particularly interested in the rhythmic abilities of such individuals because (as described earlier) two important theories of musical development make competing predictions. Piagetian theory states that conservation is required for activities involving temporal order, seriation, and rhythmic organization, regardless of the domain. Thus, we would pre- dict that rhythmic abilities ought to be impaired in children who have not yet reached the conservation stage. This claim is supported by Serafine’s (1979) finding that metric conservation in normal children was highly cor- related with the standard (nonmusical) Piagetian conservation tasks (see also Pflederer, 1964; Pflederer & Sechrest, 1968). The vast majority of in- dividuals with Williams syndrome fail standard Piagetian conservation tasks (Bellugi & Wang, in press; Bellugi et al., 1994). Furthermore, Miller (1989) found no evidence for rhythmic abilities in musical savants, arguing that their musical skills were focused in melodic and harmonic domains. On the other hand, Gardner argues that artistic endeavors in general constitute special sets of domains, and so it would seem by implication, rhythmic conservation (as a subset of musical skill) could constitute a separate set of competencies and should not necessarily be affected by failure to reach nonmusical conservation. Related to this claim is the report of Nordhoff and Robbins (1971) of a 6-year old autistic boy who demonstrated seriation of complex motor events only when such events were set to music. The boy368 Daniel J. Levitin & Ursula Bellugi had poor visual-motor coordination, but learned to tie his shoes when the activity was set to a song; it seemed as though he could organize the tempo- ral demands of the procedure through the song. Similarly, there have been clinical reports that listening to music facilitates smooth movement in pa- tients with Parkinson’s disease. Method SUBJECTS An unusual opportunity enabled us to take initial steps to observe and probe musical abilities in individuals with Williams syndrome. This was the Williams Syndrome Music and Art Camp at Belvoir Terrace in the Berkshire Mountains of Massachusetts in the sum- mer of 1996, During the first 8 weeks of the summer season, The Music and Arts Camp is a private music camp for normal girls. During the past few years, the camp has worked together with the Williams Syndrome Association and Foundation to arrange for individu- als with Williams syndrome to attend camp for 1 week during the summer. During that week, the Williams Camp offers a program of swimming, ensemble playing, arts and crafts, modern dance, musical theater, singing, and music lessons in drum, voice, wind instru- ments, or guitar. The major goal of the camp is to foster the natural musical abilities com- mon to many individuals with Williams syndrome.' Many of the individuals with Williams syndrome who attend the camp have had no prior musical training. Eight individuals with Williams syndrome participated in our informal studies, selected at random from the 40 children attending the camp that summer. The age of the subjects with Williams syndrome ranged from 9 to 20 (mean, 13.4 years, SD 3.6 years); two subjects were female and six were male. Afterwards, in order to provide an initial comparison, we recruited eight young normal subjects from Palo Alto, California, in June 1997. This con- trol group included two girls and six boys, with chronological age ranging from 5 to 7 years, with two subjects for each age category. For the initial comparison group, we re- cruited educationally sophisticated young normal children who had taken at least 3 years of musical instruction, were from one of the highest rated school districts in California, and were currently involved in private music lessons. MATERIALS Test stimuli consisted of a graduated series of one- and two-measure rhythmic patterns, clapped by the experimenter. We refer to this as an “echo clapping” or “rhythm repetition task,” A complete list of the rhythms given is included as Appendix 1 and included such familiar patterns as a strictly isochronous sequence (Standard 1) and the pattern known in America as “shave-and-a-haircut” (Standard 8; the “two bits” portion was not given). Al- though no child received all examples, each child received a block of trials that began with simple patterns in straight { time and progressed to more difficult patterns (those with higher numbers in the Appendix). Patterns included both “straight time,” and “swing time,” straight-eighths, dotted-eighth-sixteenth, and triplet patterns, and these different rhythmic “feels” were presented in a mixed list design. The patterns were based on those used in the 1. This camp has come to attention through several articles and television programs such as the segment““A Very Special Brain” on 60 Minutes (CBS Television, 1997); part of the Scientific American article “Williams Syndrome and the Brain,” (Lenhoff et al., 1997); and the segment “Don’t Be Shy, Mr. Sacks” from the series “The Mind Traveller” (BBC 2 television, 5 November 1996).Musical Abilities in Individuals With Williams Syndrome 369 Gordon Musical Aptitude Profile (Gordon, 1965/1995) and by Bruscia (1981) in a similar paradigm, and they provided a wide variety of temporal ratios. Tempos for the examples ranged from J = 80 to 120. Previous research has shown that rhythmic patterns with temporal ratios of 2:1 are the easiest to perceive and replicate (Dowling & Harwood, 1986; Drake, 1993; Fraisse, 1982). Moreover, there is a strong tendency to “quantize” or assimilate other ratios to the 2:1 ratio; “music teachers ... seem to have to guard continually against the drift of dotted- eighth sixteenth rhythms (3:1 ratios) into quarter-eighth triplet rhythms (2:1 ratio) in their pupils’ playing” (Dowling & Harwood, 1986, p. 186). Thus, our set contained exemplars that we had a priori reason to expect would be simple (Standards 1-9) as well as exemplars that we expected to be moderately to quite difficult (Standards 21-38). A “split-halves” estimate of the test’s reliability was performed (Nunnally, 1978) and the correlation be- tween odd-numbered and even-numbered items was 0.98. PROCEDURE Each subject was tested individually. Subjects with Williams syndrome were tested out- doors at the camp at a redwood picnic table; comparison group subjects were tested in the backyard of their homes. Both the experimenter’s and the subjects’ productions were re- corded on a Sony TDC-D8 Digital Audio (DAT) recorder at 44.1 kHz sampling rate, for later analysis by independent judges who were professional musicians. The experimental participants were asked if they would like to play a “rhythm game” and were advised that they could stop playing the game at any time. The experimenter then gave the following instruction: “I’m going to clap a rhythm. When I’m done, I'd like you to copy it and clap it back to me.” The experimenter then began the first trial and proceeded until 20 trials had been given or the subject asked to stop. Although they were not instructed to do so, all the subjects tended to look the experimenter in the eye (rather than watching the experimenter’s hands) during the demonstration of the reference rhythm. ANALYSIS In Bruscia’s (1981) study, independent coders were employed to make the binary dis- crimination of performances as being either a correct or incorrect repetition of the stimulus sequence. Gfeller (1992) notes that this does not fully address the continuum of difficulties that can occur in rhythmic production, We employed two jazz musicians (one a profes- sional drummer, the other a semiprofessional vocalist and trumpetist) as independent cod- ers, blinded to the hypotheses and group membership, and asked them to mark each ex- ample as correct or incorrect. In response to Gfeller’s point, the coders were asked to make any qualitative notes they cared to about each example. We defined as “incorrect” any response in which (1) the subject clearly played a different rhythm than that presented; (2) the tempo or rhythmic performance varied by more than 1/48 of a beat from the exemplar. For the binary correct/incorrect judgment, interrater agreement was 0.95. Results Without being explicitly told to do so, in nearly every trial, both groups of subjects clapped back the demonstrations immediately in perfect time, without missing a beat, as if their response formed part of the same rhyth- mic sequence. That is, when the experimenter was finished giving the ex- emplar, tlie subjects came right in on the next beat without pausing. (The experimenter held his hands together in the closed position after he had finished clapping the last beat of an example and the subjects apparently370 Daniel J. Levitin & Ursula Bellugi took this as a cue that it was their turn to clap.) All of the subjects thus appeared to interpret the examples as forming part of a larger musical set; they seemed to assume that there was an implied time signature and tempo, and they responded to the “first measure” of music played by the experi- menter in time for the downbeat (or in some cases pickups) to the “second measure.” Moreover, the subjects revealed a remarkable ability to track changes in rhythmic pulse, including changes to swing time, straight eighths, triplets, sixteenths, syncopations, and so on. In some cases, the experimenter began the next trial without pausing after the subjects’ response, giving the experimental session the flavor of a jazz “jam” session of “trading ones,” the technical term used to describe musicians who alternate playing mea- sures of a musical phrase. The fact that subjects came in immediately sug- gests that they may not have rehearsed the examples (at least not in real time) and that we might be tapping into the subjects’ echoic memory for this task, the closest way we know of to measure their direct perception of the rhythms without the confound of working memory or long-term memory. Looking strictly at the percent correct scores, the subjects with Williams syndrome performed equivalently to the normal comparison group, with each group making errors on roughly 31% of the trials (Figure 4). In fact, the subjects with Williams syndrome performed slightly better, although this difference was not statistically significant. Our finding is in contrast to that of Kaplan (1977), who found that nonspecific developmentally dis- abled children of mental age 6-8 years performed far worse in echo tap- ping tasks than did comparison subjects. For both groups of subjects, the incorrect responses tended to be correlated with rhythmic difficulty, but this correlation was not statistically significant. 100% 90% 80% 70% J ee I Normals 10% aad sn O williams 30% 20% 10% 0% Percent Responses Incorrect Fig, 4. Percent correct responses in echo clapping task as a function of group membership. Subjects with Williams syndrome have the same proportion of incorrect trials as normal subjects do, and exemplar difficulty was not correlated with performance.Musical Abilities in Individuals With Williams Syndrome 371 During the data coding, each of the data coders, blind to experimental group and hypothesis, volunteered the observation that some of the trials of the subjects with Williams syndrome (although not of the comparison group) appeared to be “wrong in interesting ways.” We asked the coders to elaborate, and each volunteered that on many of the trials in which the subjects with Williams syndrome did not reproduce the presented rhythm perfectly, the rhythm the subjects did clap bore a clear musical relationship to the referent. Data coder No. 1 explained that “it sounds like a call-and- response; as though the subject is creating a musically logical completion to the rhythm provided.” To put it another way, on some of the trials, the subjects with Williams syndrome seemed to be making music out of their responses, rather than slavishly mimicking the experimenter, rather like an improvisatory or “jam” session. We asked both coders independently to go back and listen again to the responses they had previously coded as “incor- rect” for both groups of subjects and to divide the incorrect responses into those that they felt constituted “elaborations or creative completions” ver- sus those that appeared to be clearly wrong. The coders made their ratings with no knowledge of group membership and independently of one an- other. They agreed upon 95% of their ratings, and disagreements were resolved by discussion. Considering just the trials on which the responses were coded as “incor- rect,” or not precise copies of what was presented, subjects with Williams syndrome made “creative completions” (i.e., rhythmically compatible re- sponses) on 45% of the trials, as compared with the normal comparison group, which made creative completions for only 15% of the trials (Figure 5). Each subject with Williams syndrome made creative completions rang- ing from 17% to 80% of the trials (mean creative completions per subject = 50.6%; SD = 22.4%). Figure 6 shows examples of responses that were considered by the cod- ers to constitute creative completions. The coders were not given any ex- plicit rules on which to base their judgments of rhythmic compatibility and instead apparently relied on their experiences and judgment as professional musicians. The coders noted that subjects with Williams syndrome tended to make nonrepetitions that followed rules of musical grammar that in- cluded, for example, rhythmic resolution. These rhythmic elaborations that were appropriate to the musical context occurred significantly more often in subjects with Williams syndrome than in normal mental-age-matched peers. Further, it appears that such “errors” occur not only for difficult items, but across levels of item difficulty. Note that the responses of sub- jects with Williams syndrome include embellishments that fit with the pre- sented rhythmic pattern (Figure 6, left), and that the subject carried over the syncopation into the response (Figure 6, right), preserving the global structure of the target rhythm.372 Daniel J. Levitin & Ursula Bellugi 100% 90% a6 RX gS x B Normals Williams 6 xe Q Percent Responses y or a QNo RELL 10% 0% Fig. 5. Within incorrect trials, subjects with Williams syndrome were three times as likely as normal subjects to offer a “creative completion” to the rhythmic exemplar. In searching post facto for a rule system on which the coders based their judgments of creativity, the most reasonable explanation appears to be that they relied on the rhythmic equivalent of the Gestalt principle of good con- tinuation (Wertheimer, 1923/1938) as further elaborated by Lerdahl and Jackendoff’s (1983) grouping preference rules. Often, college students studying “ear-training” (solfeggio) will scramble the elements of a rhythmic phrase (R. Ashley, personal communication). Consider, for example, Standard 8, the “shave-and-a-haircut” rhythm (with- out the “two-bits”). One could characterize this pattern as consisting of a rhythmic motif, further decomposable into four submotivic elements: a quarter note, two eighth-notes, a quarter note, and another quarter note: d))JJ. A scrambled version of this rhythm would retain the submotivic elements, but present them in the wrong order, such as this: JJ DJJ. It is Presented Rhythm Presented Rhythm 3 WMS Rhythmic Elaboration WMS Rhythmic Elaboration 3 Note: This WMS response includes two Note: This WMS response carries over the triplets that fit with the presented syncopations and preserves the global rhythmic pattern. structure of the original. Fig. 6. Examples of “incorrect” answers given by subjects with Williams syndrome in the echo clapping task. Both examples were judged by two independent coders as constituting, “creative completions.”Musical Abilities in Individuals With Williams Syndrome 373 reasonable to ask if some of the incorrect responses (of both groups) are attributable to this type of scrambling, and in particular, if the creative completions were merely motivic inversions of the standard. This could occur if individuals with Williams syndrome show a deficit in local versus global processing for rhythmic elements, analogous to the local/global defi- cits found in visual processing. There is no a priori reason to expect that Williams syndrome should manifest a local/global deficit in auditory pro- cessing: in the speech domain, individuals with Williams syndrome do not scramble words within a sentence, or sentences within paragraphs; and in the music domain, individuals with Williams syndrome do not scramble measures and phrases of a musical piece they have learned. Nevertheless, to assess this possibility, we analyzed the incorrect productions of subjects with Williams syndrome for evidence that their incorrect responses con- tained the right local elements, but in the wrong order. For example, if the exemplar was J))JJ and the response was JJ.) this could be taken as evidence for the preservation of local motivic elements at the expense of global. We examined all the productions of both groups to test this hypoth- esis, and no such patterns were found. Discussion On nearly every trial, all of the subjects in both experimental groups demonstrated conservation of musical time. Even when a subject produced a rhythm incorrectly, her erroneous production conformed to the implied bar lines of a musical measure; that is, even if a subject clapped a greater or lesser number of musical notes, in most cases, the subject correctly altered the length of the notes in order to take up the same amount of metrical time as the exemplar. This is striking because most individuals with Williams syndrome fail the Piagetian conservation tasks. In the present study, we found an ability to reproduce complex rhythmic patterns, conserve musi- cal time, and to maintain musical beat structures across rapid changes in meter, We interpret this as evidence that the Piagetian logical cognitive op- erations are orthogonal to artistic operations and as evidence for the Gardnerian notion of multiple intelligences. At the motoric level, we observed in the subjects with Williams syn- drome the typical profile of impairments, including poor spatial coordina- tion and poor eye-hand coordination. Yet in the clapping task reported here, they clapped with apparent ease and lack of difficulty, converging evidence for the notion that musical behavior is independent of other ac- tivities. Rauscher, Shaw, and Ky (1993) proposed that spatial coordina- tion, kinesthetic abilities, and mathematical abilities are correlated with musical abilities; we interpret our findings as evidence against this belief.374 Daniel J. Levitin & Ursula Bellugi Further, our finding of “creative” wrong answers on the echo clapping task by subjects with Williams syndrome supports the recent theory by Hermelin (personal communication) that inventiveness and creativity are independent of general intelligence, based on her work with autistic indi- viduals. The subjects with Williams syndrome show evidence for a quality we propose to call rhythmicity, or rhythmic musicality. We used young normal children with some musical background and train- ing as the comparison group, although obviously other comparisons could also be made. We note that if we had used chronological age-matched sub- jects, it is likely that they would have performed the echo clapping task at or near ceiling. We base this assertion on the norms for the Gordon Musi- cal Aptitude Profile (Gordon, 1965/1995), a test that has been validated on thousands of students in the past 30 years. (Our echo clapping task used the simpler rhythmic items from the Musical Aptitude Profile, items that students between 12 and 16 usually perform without error). Although at the time we performed the experiments we did not know that there would be interesting data to explore in the “wrong answers,” in retrospect using a younger comparison group was a good choice. With chronological age- matched subjects performing at ceiling, there would have been too few opportunities to examine errors for creative completions. Moreover, if nor- mal subjects were making creative completions on the Gordon test, after 30 years and 10,000 validation trials, we assume someone would have reported this in the literature by now. Most importantly, however, we rec- ognize that the work we report here is only a first step and that much work remains to be done. As FE. Heuser (personal communication) has pointed out, rhythmic ability tends to stabilize around the age of 12, and it would be interesting to test rhythmic ability of individuals with Williams syn- drome in the context of multiple comparison groups. We hope that such worthwhile research will be performed in the near future, either in our laboratory or by colleagues. Freeman (1986) hypothesized that the tendency for children with Down syndrome to delay beats (while still demonstrating relatively accurate tempo perception) could be due to difficulties with muscle tone, modality strength, neural and muscular conduction latency, and coordination skill. To use Schmidt’s (1982) term, the deficit among individuals with Down syndrome might occur in “response translation,” that is, turning a cognitive response into a motor action plan. By all available evidence, subjects with Williams syndrome suffer the same general motor difficulties that subjects with Down syndrome do, and yet in the present study, our sample of individuals with Williams syndrome did not seem to have trouble with response translation for a musically based task. An alternative account for the difference in rhythmic abilities between the two groups is that subjects with Williams syndrome appear to have a strong sense of phrasing and knowing whereMusical Abilities in Individuals With Williams Syndrome 375 the downbeat (or next “point of arrival”) is in a metrical sequence, a sense that subjects with Down syndrome may lack. In other words, many indi- viduals with Williams syndrome have a strong sense of metric expectation. This could be linked to the differences such subjects exhibit in cerebellar volume indicated earlier. The echo clapping task we used probably drew on the subjects’ echoic memory, not working memory, as it is unlikely that they had time to re- hearse the pattern before clapping it. Subjects with Williams syndrome have relatively poor verbal working memory capacity (as assessed by digit span), compared with subjects with Down syndrome, but are more impaired than subjects with Down syndrome on spatial working memory (Wang & Bellugi, 1994). Whether subjects with Williams syndrome would show relatively good working memory capacity in a rhythmic or melodic task is a subject for future investigation. We note that researchers have found that phono- logical memory appears to be a relative strength in individuals with Will- iams syndrome (Vicari, Carlesimo, Brizzoloara, & Pezzini, 1996). Révész (1953/1954) proposed the term “musicality” to refer to “the ability to enjoy music aesthetically.” The creative, rhythmically compatible comple- tions we observed could be interpreted as a type of musicality and this is consistent with our subjective observations that subjects with Williams syn- drome appear to be more engaged with music than normal subjects. In particular, we propose the term rhythmicity as a subcategory of Révész’s musicality index, to capture the idea that some individuals have a high degree of rhythmic engagement. All subjects in our Williams syndrome group used creative rhythmic completions, and there were no significant differ- ences between subjects on their propensity for creative completions. How can this behavior be explained? Perhaps when subjects with Williams syn- drome were unable to replicate a rhythm (for whatever reason) they tended to err on the side of rhythmic musicality (rhythmicity), taking the larger musical context into account. IMPRESSIONS The first author observed that there is variability among this population in their behaviors and abilities, just as there would be in any normal popu- lation. Yet, it appeared that the subjects with Williams syndrome had an unusually high degree of engagement with music. Music seemed to be not just a very deep and rich part of their lives, but one that was omnipresent; most of them spent a great proportion of the day singing to themselves or playing instruments, even while walking to the mess hall. They appeared to do this with a good deal less self-consciousness than normal music camp attendees, or normal individuals. When one camper encountered another camper or group of campers involved in a musical activity, no matter how376 Daniel J. Levitin & Ursula Bellugi informal, the newcomer would either join in immediately or begin swaying appreciatively in time to the music. We feel the need to emphasize our observation that this consuming involvement with music is unusual in nor- mal populations; one of us (DJL) spent a number of years working as a professional musician and rarely encountered this type of total immersion among musicians—even though the type of connectedness to music the subjects with Williams syndrome naturally displayed is often considered a goal among many musicians. As previously noted, by Révész’s (1953/1954) definition, many of the individuals with Williams syndrome we observed appear to be very much engaged with music and showed the external be- havior of being in a “flow” state, a state of concentration and deep enjoy- ment that allows absolute absorption in an activity (Csikszentmihalyi, 1990). All the more striking was the inability of many of the individuals with Williams syndrome to carry out tasks that required low level hand-eye co- ordination or visual-spatial skills. For example, many subjects had diffi- culty walking and handled eating utensils with an uncomfortable awkward- ness. Yet, when presented with the musical instrument on which they specialize—be it a clarinet, piano, drum set, or guitar—some were able to execute musical passages requiring a much finer degree of control and co- ordination than that required to successfully walk a flight of stairs or cut their food (Levitin & Bellugi, 1997a, 1997b). There is a decided lack of technical perfection in all these musical performances, but this is more than compensated for by the subjective quality of musicality they bring to their performances. SPONTANEOUS SONG IMPROVISATION IN ONE SUBJECT One participant (“B”) was known among the other campers as a prolific creator of songs (we avoid the term “songwriter” because B does not actu- ally write anything down). To examine B’s ability, the experimenter asked him to write a song about a breakfast cereal, specifically “Kellogg’s Rice Krispies,” a topic which was a complete non sequitur during the experi- mental session. To the Experimenter’s surprise, B spontaneously composed a song about Rice Krispies on the piano, complete with verse, chorus, and rhyming lyrics (Appendix 2).* Without further observations, it is not clear how much this apparently new composition differed from others that B has written, but a quick review of some of B’s repertoire revealed both a distinct compositional style and a marked difference among the various tunes. The experimenter then asked B to write a song about his favorite animal. B wrote a new song about dogs, playing and composing it in real time (lyrics reprinted in Appendix 2). Some of the rhymes were clever and 2. These songs can be downloaded from the Web: http://www-ccrma.stanford.edu/~levitin/ WMS.htmlMusical Abilities in Individuals With Williams Syndrome 377 employed assonance, clearly a sophisticated form of rhyming. The first author described B’s vocal quality as being very natural, having great pres- ence, and characterized by a complete lack of inhibition. There are some structurally interesting aspects of B’s songs worth not- ing. In the Dog song, the first two verses share the following parallel struc- ture: the first three lines begin with the phrase “a [the] dog,” and the fourth line switches to a pronoun (she/I). Note that in the second verse, the narra- tor shifts the viewpoint from the third person to the first person right be- fore beginning the next major portion of the song. He does this in the Rice Krispies song as well, so it seems as though it is a compositional predispo- sition. Note that in the chorus, most of the pronouns are first person. A similar “three and one” structure holds for the third verse: “She is so beau- tiful/She is so nice looking/She stands so tall” followed by the odd line out for the fourth line. And again, following the fourth verse the narrator shifts from third person to first person just before the next major section, putting the narrator in focus again. An interesting question is whether B can tell the difference between the parts of the song that are good and those that aren’t, and whether B can actually edit his ideas into a coherent piece. In other words, is B capable of composing music? The songs contained standard structural elements, such as repeating motivic ideas and motivic development. Although they lacked an identifiable chorus, it is not unusual even for accomplished songwriters to explore various musical ideas on a first run through, only later develop- ing the standard verse-chorus structure out of the raw ideas. In evaluating B’s lyrics, it is important to understand the way that many professional songwriters write. Typically the songwriter will experiment with different phonological sounds set to a melody; the sound of words at this point of the writing process is very important, even if the words them- selves do not make sense. Once the song has a melodic and phonological shape, the songwriter typically goes back to edit the words so that they make more semantic sense, while still retaining as much of their phonology as possible. For example, Paul McCartney’s original lyric for his song Yes- terday was “scrambled eggs” and Paul Simon’s original lyric for his song Kodachrome was “goin’ home” (Levitin, 1998). Based only on our data, we don’t know how or if B might have edited or refined his lyrics. We raise this issue only to alert the careful reader to the fact that certain semantic or syntactic anomalies in B’s lyrics are typical among songwriters’ first at- tempts at creating a song. All of B’s piano chords are played in “root” position (that is, they all use nearly identical fingerings, save for the changes from major to minor) and this might be a consequence of the poor muscle and limb coordination from which most individuals with Williams syndrome suffer. B later sang us some original songs he had written on the guitar. One of the music378 Daniel J. Levitin & Ursula Bellugi instructors tuned the guitar to an “open” tuning (such that in the unfretted position the strings are tuned to the pitches of a major triad) and B played “barre” chords exclusively, in the style of guitarist Richie Havens. RECOGNITION MEMORY AND CLASSIFICATION OF TIMBRE: CASES OF “HYPERTIMBRIA?” According to the parents we interviewed, many children with Williams syndrome are fascinated by certain sounds, and it appears that they can make accurate identifications along the sensory continuum of timbre. While we were talking to one subject (A), a teacher started to play a Japanese wooden flute in the adjacent room. “A” became very, very excited upon hearing the sound through the wall. “That’s a new instrument! It sounds like a flute, but not like any flute I’ve heard before. It must be made of wood!” Consider that first “A” had to recognize that the instrument she heard was some sort of flute, then that it must not be made of metal, and she had to infer what a wooden flute might sound like. As mentioned above, one child could reportedly distinguish among a dozen different brands and models of vacuum cleaners and leaf blowers based exclusively on the sound of their motors; another child could report- edly identify the makes of dozens of cars as they rounded the corner out- side of his house based solely on the sound of the engine. These behaviors suggest to us that many individuals with Williams syndrome may have something akin to absolute pitch. Whereas a subject with absolute pitch is able to code, recall, and label discrete stimuli along the pitch continuum (Levitin, 1996, in press) , perhaps individuals with Williams syndrome can perform the analogous feat in the timbre domain. This is the topic of re- search currently underway in our laboratory. The picture emerging is that many individuals with Williams syndrome notice subtle variations in timbre more than others; this is meant not to imply that they have finer discrimination abilities than others, just that they appear to be more interested in developing abilities in this domain. And their easy classification and identification of objects based on timbre suggests a highly developed type of recognition memory in this regard. Perhaps these behaviors might best be described by the term “hypertimbria.” MUSICALITY The parents we interviewed all reported that their children have an in- tense connection to music away from camp as well. Many parents reported that their children knew hundreds, if not thousands of songs from popular music records; although this is not unusual among “normal” American teen-agers, it is worth noting as a contrast to the deficits in other cognitiveMusical Abilities in Individuals With Williams Syndrome 379 domains among individuals with Williams syndrome. Individuals with Williams syndrome also express a strong connection to music. “Music is my favorite way of thinking,” one subject reported, and another stated that “Music is like soup for the soul, it feels so good” (Lenhoff et al., 1997). We heard many stories about infants (12 weeks old) who could match pitch along with a parent playing the piano, or toddlers (24 months old) who could sit down at the piano and play back their older siblings’ piano lessons. One adult with Williams syndrome adores opera, and in a fluency test we administered to “name all the operas that you can,” he performed terrifically. Such anecdotal accounts demand controlled experimental veri- fication, but the similarities among them—and the sheer number of them— lead us to believe that individuals with Williams syndrome do have a much greater degree of musical involvement and “musicality” than normal indi- viduals, autistic individuals, or children with Down syndrome. Our labo- ratory is currently undertaking a comprehensive program of research to quantify and better characterize the musical profile of individuals with Williams syndrome in detail. General Discussion Our eight subjects with Williams syndrome were selected from among 40 such individuals who were attending a music and arts camp, and these 40 subjects in turn are a nonrandom sample of the more than 3000 indi- viduals with Williams syndrome diagnosed to date in the United States. Thus, although we are reluctant to generalize our findings to all individu- als with Williams syndrome, that was not the main point of our inquiry. The existence of relatively preserved musical ability—for both production and reception of musical ideas—forms an intriguing contrast to the pattern of cognitive deficits found in individuals with Williams syndrome. Whether continuing studies reveal that musical ability is found often or only occa- sionally in subjects with Williams syndrome, the types of behaviors we report form strong evidence for the notion that musical ability is indepen- dent of, and uncorrelated (to a large degree) with, overall cognitive func- tion. One interesting meta-perspective on the musical abilities we observed relates to recent arguments about the evolutionary basis for musical ability (Huron, 1998; Pinker, 1997). One hypothesis is that through humans’ evo- lutionary history, music served an important social role for bonding, form- ing alliances, and communication. It is interesting to consider as evidence the contrast between individuals with Williams syndrome, who show high musicality and high sociability, and autistic individuals, who typically show low musicality and low sociability. This is just the type of dissociation one380 Daniel J. Levitin & Ursula Bellugi would expect to find if music and sociability were related to one another as evolutionary adaptations. Recall that one of Gardner’s (1983) evidences for an intelligence is its evolutionary plausibility, so this suggests further support for the idea of music as an independent intelligence. Conclusions We have described new evidence for the cognitive independence of rhyth- mic ability in individuals with Williams syndrome. Despite substantial im- pairments in many cognitive domains, individuals with Williams syndrome appear to possess unusual strength in language and music abilities. Although Williams syndrome is characterized by a general inability to properly per- form the Piagetian tasks for conservation of number, space, substance, weight, and quantity, the subjects with Williams syndrome in the present study demonstrated relatively preserved abilities for echo clapping, an ac- tivity that requires the perception of temporal order, seriation, and rhyth- mic organization. And, despite the fact that our subjects exhibited the gen- eral motor and spatial impairments typical of Williams syndrome, once they became engaged in a musical task (the clapping game), they were ca- pable of clapping their hands to specific rhythms with remarkable preci- sion. In our qualitative observations, we found they were also able to play musical instruments requiring fine muscle control and dexterity. When sub- jects with Williams syndrome made errors on the echo clapping task, they were much more likely to clap a rhythm that was musically coherent with the referent, demonstrating a sense of creativity and musicality we did not find in the comparison group. We report evidence that rhythmic abilities are preserved in individuals with Williams syndrome, and moreover, that they have an interesting pro- pensity for creative rhythmic productions. This opens the possibility that others individuals with impaired cognition due to different causes (indi- viduals with Down syndrome, autistic individuals, and unspecified musical savants) might also have relatively preserved rhythmic abilities. And, al- though we did not formally test the timbral, melodic, or harmonic abilities of individuals with Williams syndrome, there is reason to believe that these components of musical ability are also relatively preserved. The profile of abilities we found in subjects with Williams syndrome supports Gardner’s theory that musical ability constitutes a separate fac- ulty and is to a large degree uncorrelated with other aspects of cognitive functioning. Our findings are consistent with the work of Jackendoff (1987) and Peretz and Morais (1989), who argue for.the modularity of musical ability. Our findings are inconsistent with previous findings that musical ability in retarded persons tends to be primarily melodic and not rhythmicMusical Abilities in Individuals With Williams Syndrome 381 and that rhythmic ability is highly correlated with Piagetian conservation tasks. Our finding of “creative” wrong answers by subjects with Williams syn- drome on the echo clapping task supports Hermelin’s idea that inventive- ness and creativity are independent of general intelligence. The subjects with Williams syndrome tested herein showed evidence for a quality we propose to call rhythmicity, or rhythmic musicality. Finally, we note that the study of distinct, well-defined, and atypical populations such as individuals with Williams syndrome affords a unique opportunity to investigate specific cognitive abilities (such as music recep- tion and production). The emerging evidence that the underlying neural architecture in individuals with Williams syndrome is very different from that of normal individuals makes these investigations even more compel- ling and relevant to theories of the biological basis of behavior? References Bamberger, J. (1991). The mind behind the musical ear: How children develop musical intelligence. Cambridge, MA: Harvard University Press. Basso, A. (1993). Amusia. In H. Spinnler & F. Boller (Eds.), Handbook of neuropsychology (Vol. 8, pp. 391-410). Amsterdam: Elsevier, Bellugi, U., Klima, E. S., & Wang, P. P. (1996). Cognitive and neural development: Clues from genetically based syndromes. In D. Magnusson (Ed.), The life-span development of individuals: A synthesis of biological and psychological perspectives. (Proceedings of the Nobel Symposium, Stockholm, Sweden, June 19-22, 1994; pp. 223-243). New York: Cambridge University Press. Bellugi, U., Lai, Z. C., & Korenberg, J. (in press). Genes, brains and behavior: What genetic disorders reveal about behavior. In E. Bizzi, P. Calissano, & V. Volterra (Eds.), The brain of Homo sapiens: Vol. 4. Frontiere della biologia. Rome, Italy: Istituto della Enciclopedia Italiana. Bellugi, U., Lai, Z. C., & Wang, P. (1997). Language, communication and neural systems in Williams syndrome. In T. Dolan (Ed.), Special issue: Communication processes in chil- dren with developmental disabilities. Mental Retardation and Developmental Disabili- ties Research Reviews, 3(4), 334-342. Bellugi, U., Losh, M., Reilly, J. & Anderson, D. (1998). Excessive use of linguistically en- coded affect: Stories from young children with Williams syndrome (Tech. Rep. CND- 3. This research was supported by Interval Research Corporation and by grants from the National Institutes of Health (HD33113, NS22343, DC01289) to Dr. Ursula Bellugi at the Salk Institute for Biological Studies. Preliminary versions of this report were presented at the third triennial meeting of the European Society for the Cognitive Sciences of Music (ESCOM), Uppsala, Sweden, in June 1997, and at the meeting of the Society for Music Perception and Cognition (SMPC), Cambridge, MA, in August 1997. We thank Nancy Goldberg and the Williams Syndrome Music and Arts Camp at Belvoir Terrace for their hospitality. We also thank Edward S. Klima, Michael Chiles, Ric Ashley, Jamshed Bharucha, Frank Heuser, David Huron, Helen Shwe, John Sloboda, and the members of the Stanford Music Pereeption & Cognition Group (Roger Shepard, Jeff Zacks, Paul von Hippel, Natalia Engavatov, and Karen Wang) for many helpful discussions and comments on earlier drafts of this paper. Eric Jensen and James Sagebiel assisted in coding and analyzing the data.382 Daniel J. Levitin & Ursula Bellugi 9801). San Diego: University of California, San Diego, Center for Research in Language, Project in Cognitive and Neural Development. Bellugi, U., Mills, D., Jernigan, T., Hickok, G., & Galaburda, A. (In press). Linking cogni- tion, brain structure and brain function in Williams syndrome. In H. Tager-Flusberg {Ed.), Neurodevelopmental disorders: Contributions to a new framework from the cog- nitive neurosciences. Cambridge, MA: MIT Press. Bellugi, U., & Wang, P. P. (in press). Williams syndrome: From cognition to brain to gene. Encyclopedia of Neuroscience. Amsterdam: Elsevier Science. Bellugi, U., Wang, P. P., & Jernigan, T L. (1994). Williams syndrome: An unusual neuropsychological profile. In $. Broman & J. Grafman (Eds.), Atypical cognitive defi- cits in developmental disorders: Implications for brain function (pp. 23-56). Hillsdale, NJ: Lawrence Erlbaum Associates. Benton, A. L. (1977). The amusias. In M. Critchley & R. A. Henson (Eds.), Music and the brain (pp. 378-397). Springfield, IL: Thomas. Bihrle, A. M., Bellugi, U., Delis, D., & Marks, S. (1989). Seeing either the forest or the trees: Dissociation in visuospatial processing. Brain and Cognition, 11, 37-49. Bixler, J. (1968). Musical aptitude in the educable mentally retarded child. Journal of Music Therapy, 5(2), 41-43. Bruscia, K. E, (1981). Auditory short-term memory and attentional control of mentally retarded persons. American Journal of Mental Deficiency, 85(4), 435-437. Csikszentmihalyi, M. (1990). Flow. New York: Harper & Row. Delis, D. C., Kiefner, M. G., & Fridlund, A. J. (1988). Visuospatial dysfunction following unilateral brain damage: Dissociations in hierarchical and hemispatial analysis. Journal of Clinical and Experimental Neuropsychology, 10, 421-431. Delis, D. C., Robertson, L. C., & Efron, R. (1986). Hemispheric specialization of memory for visual hierarchical stimuli. Newropsychologia, 24, 205-214. Don, A. (1997). Auditory pattern perception in children with Williams syndrome (WMS). Unpublished doctoral dissertation, University of Windsor, Ontario. Dowling, W. J., & Harwood, D. L. (1986). Music cognition. San Diego: Academic Press. Drake, C. (1993). Reproduction of musical rhythms by children, adult musicians, and adult nonmusicians. Perception & Psychophysics, 53(1), 25-33. Edgren, J. G. (1895). Amusie (musikalische Aphasie). Deutsche Zietschrift der Nervenheilkunde, 6, 1-64. Fanconi, G. (1952). Textbook of pediatrics (W. R. F. Collis, Ed.; E. Kawerau, Co-Ed., Trans.). London: Wm. Heinemann. Fasanaro, A. M., Spitaleri, D. L. A., & Valiani, R. (1990). Dissociation in musical reading: ‘A musician affected by alexia without agraphia. Music Perception, 7(3), 259-272. Fraisse, P. (1982). Rhythm and tempo. In D. Deutsch (Ed.), The psychology of music, (pp. 149-180). New York: Academic Press. Freeman, I. A. (1986). Rhythmic beat perception in a Down’s syndrome population: A computerized measure of beat accuracy and beat interval response. Unpublished doc- toral dissertation, University of North Carolina, Greensboro. Galaburda, A. M., Wang, P. P., Bellugi, U., & Rossen, M. (1994), Cytoarchitectonic anoma- lies in a genetically based disorder: Williams syndrome. NeuroReport, 5(March 21), 753-757. Gardner, H. (1973). The arts and human development. New York: Wiley. Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. New York: Basic Books. Gardner, H., & Perkins, D. (Eds.). (1988). Art, mind, and education: research from Project Zero. Urbana: University of Illinois Press. Gazzaniga, M. S. (1988). Mind matters. Boston: Houghton-Mifflin. Gfeller, K. (1992). Research regarding students with disabilities. In R. Colwell (Ed.), Hand- book of research on music teaching and learning (pp. 615-632). New York: Schirmer/ Macmillan. Gordon, E. E. (1965; re’ Boston: Houghton Mi 1995). Musical Aptitude Profile [Audio recording and book]. in.Musical Abilities in Individuals With Williams Syndrome 383 Gordon, E. E. (1986). Primary measures of music audiation [Book and sound recording]. Chicago: G.LA. Publications. Hargreaves, D. (1996). The development of artistic and musical competence. In I. Deliége & J. Sloboda (Eds.), Musical beginnings: Origins and development of musical compe- tence (pp. 145-170). Oxford: Oxford University Press. Hickok, G., Bellugi, U., & Jones, W. (1995). Letters to Science: Asymmetrical ability. Sci- ence, 270, 219-220. Hickok, G., Neville, H., Mills, D., Jones, W., Rossen, M., & Bellugi, U. (1995). Electro- physiological and quantitative MR analysis of the cortical auditory system in Williams syndrome. Cognitive Neuroscience Society Abstracts, 2, 66, #40. Huron, D. (1998, January 22). Is music-making an evolutionary adaptation? Paper pre- sented at the meeting of the CCRMA Hearing Seminar, Stanford, CA. Imberty, M. (1996). Linguistic and musical development in preschool and school-age chil- dren In I. Deliége & J. Sloboda (Eds.), Musical beginnings: Origins and development of musical competence, (pp. 191-213). Oxford: Oxford University Press. Jackendoff, R. (1987). Consciousness and the computational mind. Cambridge, MA: MIT Press. Jernigan, T. L., & Bellugi, U. (1990). Anomalous brain morphology on magnetic resonance images in Williams syndrome and Down syndrome. Archives of Neurology, 47, 529- 533. Jernigan, T. L., & Bellugi, U. (1994). Neuroanatomical distinctions between Williams and Down syndromes. In $. Broman & J. Grafman (Eds.), Atypical cognitive deficits in de- velopmental disorders: Implications for brain function (pp. 57-66). Hillsdale, NJ: Lawrence Erlbaum Associates. Kaplan, P. R. (1977). A criterion-referenced comparison of rhythmic responsiveness in nor- mal and educable mentally retarded children. Unpublished doctoral dissertation, Uni- versity of Michigan, Ann Arbor. Klein, A. J., Armstrong. B. L., Greer. M. K., & Brown, E.R. (1990). Hyperacusis and otitis media in individuals with Williams syndrome. Journal of Speech and Hearing Disorders, 55(2):339-344. Korenberg, J. R., Chen, X-N., Lai, Z., Yimlamai, D., Bishighini, R., & Bellugi, U. (1997). Williams syndrome: The search for the genetic origins of cognition. American Journal of Human Genetics, 61(4), 103-106. Lenhoff, H. M., Wang, P. P., Greenberg, F, & Bellugi, U. (1997). Williams syndrome and the brain. Scientific American, 277, 68-73. Lerdahl, F., & Jackendoff, R. (1983). A generative theory of tonal music. Cambridge, MA: MIT Press. Levitin, D. J. (1996). Mechanisms of memory for musical attributes. Unpublished Ph.D. Dissertation, University of Oregon, Eugene. Levitin, D. J. (1998). Still creative after all these years: A conversation with Paul Simon. Grammy Magazine, 16(1, February), 16-19, 46. Levitin, D. J. (in press). Memory for musical attributes. In P. R. Cook (Ed.), Lectures in psychoacoustics. Cambridge, MA: MIT. Press. Levitin, D., & Bellugi, U. (1997a, June). Musical abilities in individuals with Williams syn- drome. Paper presented at the meeting of the European Society for the Cognitive Sci- ences of Music (ESCOM), Uppsala, Sweden. Levitin, D., & Bellugi, U. (1997b, August). Musical abilities in individuals with Williams syndrome. Paper presented at the meeting of the Society for Music Perception and Cog- nition, Cambridge, MA. Luria, A. R. (1966). Higher cortical functions in man. New York: Basic Books. Marin, O. $. M. (1982). Neurological aspects of music perception and performance. In D. Deutsch (Ed.), The psychology of music (pp. 453-478). New York: Academic Press. McLeish, J.,-& Higgs, G. (1982). Musical ability and mental subnormality: An experimen- tal investigation. British Journal of Educational Psychology, 52, 370-373. Mervis, C. B., & Bertrand, J. (1997). In L. B. Adamson & M. A. Romski (Eds.), Communi- cation and language acquisition: Discoveries from atypical development (pp. 75-106). Baltimore, MD: P. H. Brookes.384 Daniel J. Levitin & Ursula Bellugi Miller, J. E. (1987). Language and communication characteristics of children with Down syndrome. In S. M. Pueschel (Ed.), New perspectives on Down syndrome (pp. 39-50). Baltimore, MD: P. H. Brooks. Miller, L. K. (1989). Musical savants: Exceptional skill in the mentally retarded. Hillsdale, NJ: Lawrence Erlbaum Associates. Milner, B. (1962). Laterality effects in audition, In V. Mounteastle (Ed.), Interhemispheric effects and cerebral dominance (pp. 177-195). Baltimore, MD: Johns Hopkins Press. Nigam, A., & Samuel, P. R. (1994). Hyperacusis and Williams syndrome. Journal of Laryngology and Otology, 108(6), 494-496. Nordhoff, P., & Robbins, C. (1971). Therapy in music for handicapped children. London: Gollancz. Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill. O’Connell, T. (1974). The musical life of an autistic boy. Journal of Autism and Childhood Schizophrenia, 3, 223-229. Parsons, M. J. (1987). How we understand art. Cambridge: Cambridge University Press. Patel, A. D., & Peretz, I. (1997). Is music autonomous from language? A neuropsychological appraisal. In I. Delige & J. Sloboda (Eds.), Perception and cognition of music. East Sussex, U.K. Psychology Press/Taylor & Francis. Patel, A. D., Peretz, I., Tramo, M. 8& Lebreque, R. (1998). Processing prosodic and musical patterns: A neuropsychological investigation. Brain and Language, 61(1), 123-144. Peretz, I. (1990). Processing of local and global musical information by unilateral brain- damaged patients. Brain, 113, 1185-1205. Peretz, I. (1993). Auditory atonalia for melodies. Cognitive Neuropsychology, 10(1), 21- 56. Peretz, I. (1996). “Can we lose memory for music? A case of music agnosia in a nonmusician.” Journal of Cognitive Neuroscience, 8(6), 481-496. Peretz, I., & Hébert, S. (1995). Music processing after brain damage: The case of rhythm without melody. In R. Steinberg (Ed.), Music and the mind machine: The psychophysiol- ogy and psychopathology of the sense of music. (pp. 127-137). Berlin: Springer-Verlag. Peretz, I., & Kolinsky, R. (1993). Boundaries of separability between melody and rhythm in music discrimination: a neuropsychological perspective. Quarterly Journal of Experi- mental Psychology, 46A, 301-325. Peretz, I., Kolinsky, R., Tramo, M., & Labrecque, R. (1994). Functional dissociations fol- lowing bilateral lesions of auditory cortex. Brain, 117(6), 1283-1301. Peretz, I., & Morais, J. (1989). Specificity for music. In F. Boller & J. Grafman (Eds.), Handbook of neuropsychology (pp. 373-390.) Amsterdam: Elsevier. Pflederer, M. (1964). The responses of children to musical tasks embodying Piaget’s prin- ciple of conservation. Journal of Research in Music Education, 12(4), 251-268. Pflederer, M., & Sechrest, L. (1968). Conservation-type responses of children to musical stimuli, Council for Research in Music Education Bulletin, 13, 19-36. Piaget, J. (1963). The origins of intelligence in children. New York: Norton. Piaget, J. (1970). The child’s conception of time. New York: Basic Books. Piaget, J., & Inhelder, B. (1969). The psychology of the child. London: Routledge & Kegan Paul. Pinker, S. (1997). How the mind works. New York: W.W. Norton. Platel, H., Price, C., Baron, J.-C., Wise, R., Lambert, J., Frackowiak, R. S. J., Lechevalier, B., & Eustache, EF, (1997). The structural components of music perception: A functional anatomical study. Brain, 120, 229-243. Rauscher, F. H., Shaw, G. L., & Ky, K. N. (1993). Music and spatial task performance. Nature, 365(14 October), 611. Reilly, J., Klima, E. S., & Bellugi, U. (1990). Once more with feeling: Affect and language in atypical populations. Development and psychopathology, 2(4), 367-391. Révész, G. (1953/1954): Introduction to the psychology of music (G. I. C. de Courcy, Trans.). Norman: University of Oklahoma Press. Sacks, O. W. (1985). The man who mistook his wife for a hat and other clinical tales. New York: Summit.Musical Abilities in Individuals With Williams Syndrome 385 Samson, S., & Zatorre, R. J. (1988). Melodic and harmonic discrimination following uni- lateral cerebral excision. Brain and Cognition, 7, 348-360. Samson, S., & Zatorre, R. J. (1992). Learning and retention of melodic and verbal informa- tion after unilateral temporal lobectomy. Neuropsychologia, 30(9), 815-826. Samson, $., & Zatorre, R. J. (1994). Contribution of the right temporal lobe to musical timbre discrimination. Neuropsychologia, 32(2), 231-240. Schafer, R. M. (1969). The new soundscape. New York; Associated Music Publishers. Schlaug, G., Jancke, L., Huang, Y., & Steinmetz, H. (1995). In vivo evidence of structural brain asymmetry in musicians. Science, 267(3 February), 699-701. Schmidt, R. A. (1982). Motor control and learning: A behavioral emphasis. Champaign, IL: Human Kinetics Press. Serafine, M. L. (1979). A measure of meter conservation in music, based on Piaget's theory. Genetic Psychology Monographs, 99, 185-229. Shuter-Dyson, R., & Gabriel, C. (1981). The psychology of musical ability. London: Methuen. Shuttleworth, G. C. (1900). Mentally deficient children: Their treatment and training. Lon- don: P, Blakiston’s. Stanzione, M., Grossi, D., & Roberto, L. (1990). Note-by-note music reading: A musician with letter-by-letter reading. Music Perception, 7(3), 273-284. Stratford, B., & Ching, E. Y. (1983). Rhythm and time in the perception of Down’s Syn- drome children. Journal of Mental Deficiency Research, 27, 23-38. Vicari, $., Carlesimo, G., Brizzoloara, D., & Pezzini, G. (1996). Short term memory in children with Williams syndrome: a reduced contribution of lexical-semantic knowledge to word span. Neuropsychologia, 34(9), 919-925. Wang, P. P., & Bellugi, U. (1994), Evidence from two genetic syndromes for a dissociation between verbal and visual-spatial short-term memory. Journal of clinical and experi- mental neuropsychology, 16(2), 317-322. Wang, P. P., Hesselink, J. R., Jernigan, T. L., Doherty, S., & Bellugi, U. (1992). Specific neurobehavioral profile of Williams’ syndrome is associated with’ neucerebellar hemi- spheric preservation. Neurology, 42(10), 1999-2002. Wertheimer, M. (1938). Laws of organization in perceptual forms. In W. D. Ellis [Ed.], A source book of Gestalt psychology. London: Routledge & Kegan Paul. (Original work published 1923) Williams, J. C. P., Barratt-Boyes, B. G., 8 Lowe, J. B. (1961). Supravalvular aortic stenosis. Circulation, 24, 1311. Zatorre, R. J. (1989). Intact absolute pitch ability after left temporal lobectomy. Cortex, 25, 567-580. Zatorte, R. J., Evans, A. C., Meyer, E., & Gjedde, A. (1992). Lateralization of phonetic and pitch discrimination in speech processing. Science, 256(8 May), 846-849. Zatorre, R. J., & Halpern, A. R. (1993). Effect of unilateral temporal-lobe excision on perception and imagery of songs. Neuropsychologia, 31(3), 221-232. Zatorre, R. J., Halpern, A. R., Perry, D. W., Meyer, E., & Evans, A. C. (1996). Hearing in the mind’s ear: a PET investigation of musical imagery and perception. Journal of Cog- nitive Neuroscience, 8(1), 29-46. Appendix 1 Williams Rhythm Study Standards Listed by Difficulty Standard 2386 Daniel J. Levitin & Ursula Bellugi Standard 10. Standard oe f Bets Standard 12 Standard 13 Standard 14 ‘Standard 1S Standard 16 Statdard 174 ‘Statdard 18 Suidard 19Musical Abilities in Individuals With Williams Syndrome p_Siandart20 SS ° SS ‘Standard 22 é iG — Standard 23 Standard 24 Standard 35 387388 Daniel J. Levitin & Ursula Bellugi Appendix 2 Lyrics of Spontaneous Songs Written by Subject “B,” August 1996 Dog Song A dog starts to bark A dog starts to get mean A dog starts to sing. And then she starts to scream The dog starts to sigh Then the dog starts to cry Then the dog starts to growl Then I start tonight But then all a sudden it just happened to me I thought I had a dog who had some fleas But then I didn’t know for certain But she was the one I was thinking of ‘My dog’s name is Claudine And she’s the one that I love She is so beautiful She is so nice looking She stands so tall And do whatever is to again She thinks I’m scared She thinks I don’t care But whatever happened to me that night It’s already gone But then I sit down to watch TV And then I turn out the light And I say, “that dog is staying out tonight” Tcan’t wish this nothin I don’t even care Why all of a sudden would she be standin there If Thad to stare But then one day the police came along ‘And tried to take my dog away But then I said, “Hmm” That sure is a price to pay But then she came back ‘And then I started to go slack Because then she finally started to growl at me Like something elseMusical Abilities in Individuals With Williams Syndrome 389 The dog is so handsome Ican’t understand a word she says She put some mud in the floor And my momma starts to scream at her I said, “Oh what a bust” Then I started to boast T begin to scream My eyes filled with tears Then I started to shout Then I started to sound sincere But then one day the came down to say I'm sorry because she had something bad to say You better take me back home Rice Krispies Song Rice Krispies sound alright Until somebody came along, and it was dark at night And then all of a sudden I heard them say “You sure do like rice Krispies” “Cause you finally know that you've got to eat I said, “well yeah, I do” But then somebody told me That they were gonna steal them from me “Cause I didn’t know what to do with it Then I put some milk in the bowl And I said, “That’s not gonna do it” So when ever you're eating Rice Krispies Be sure you put some milk in it And then just enjoy yourself And then just eat it all up And then you are feeling mighty well