Commentary

Robert Walker's "Can we understand the music of another culture? espouses much that passes for common wisdom among some music scholars. Regrettably, mixed with the grains of truth are intemperate claims that sow much misinformation.

First, let me begin by affirming what I think are important points raised in Walker's commentary. It is true that a good deal of research and theory in music psychology still tends to emphasise pre-Debussian musical language. Early, contemporary, and non-Western musics have received comparatively little attention -- either due to chauvinism, disinterest, unfamiliarity, or impracticality. I also agree with Walker's assessment that idiomatic possibilities (such as technology) have had a considerable effect on cultural expressions and values. Moreover, in attempting to understand the music of another culture, it is clearly important to attempt to understand the cultural milieu from which the music arises. This is surely the first lesson that is taught to students of ethnomusicology. Finally, I also agree with the view that the principal concern of music scholarship should be the investigation of what music participants actually do, rather than a normative description of what music ought to be. (I would hasten to add, however, that there remains an essential, though secondary, role in music scholarship for critical aesthetics).

With regard to Walker's comments concerning nature/nurture, his views are regrettable in their lack of balance, and in their failure to address the extant experimental literatures. Disciplines like biology, psychology, medicine, and education have long had to wrestle with problems of whether particular behaviours are learned or innate. As these disciplines have matured, the simple polarisations into "nature" and "nurture" camps have typically dissolved. Ultimately, researchers have collaborated on the more sober research programmes of deciphering the proportion of nature/nurture for specific types of behaviours.

In the case of music, Walker claims already to have the answer: musical behavior is overwhelmingly acculturation with perhaps a few per cent physiologically or developmentally innate. I, too, tend to suspect that cultural context accounts for the lion's share of musical behaviour. What I find unfortunate is Walker's certainty, and his willingness to present speculation as conclusion:

"There is no empirical evidence from biology, from anthropology, nor from psychology that humans are predetermined towards any particular behaviours in cultural activities such as music and language."

There is no room here to recite the litany of studies pertaining to innate language behaviours. However, a simple example might be cited to whet the reader's interest -- that of studies of deaf children raised by hearing parents in isolation from other deaf children.

These children learn sign language from parents whose signing is ungrammatical and incompetent. Nevertheless, by the age of four, these children regularly and spontaneously sign grammatical inflections, tenses, and other constructions that they have never observed. Brain imaging and other techniques show that the sign-language skills of congenitally deaf children activate the same cortical regions as for spoken language in non-deaf humans (Pinker, 1994).

Although Pinker's review is extensive, it is by no means exhaustive. Other examples supporting innate or cross-cultural bases include work in infant-directed speech. In all cultures studied to date, infant-directed speech is higher in overall pitch compared to adult-directed speech. In addition, pitch contours are more highly exaggerated, repetition is more common, and the rate of speaking is slower when addressing children. When trying to arouse or capture an infant's attention, adult speakers tend to speak with high, ascending pitch contours (Stern, Spiker and MacKain, 1982). However, when trying to lull or soothe an infant, low, descending pitch contours are more common (Papousek and Papousek, 1981). Moreover, infants prefer such "motherese" inflections to normal vocal tone (Ferald, 1985).

The relationships between infant-directed speech and infant-directed song has been chronicled in the pages of Psychology of Music. Unyk, Trehub, Trainor and Schellenberg (1992) carried out a study in which 28 recorded lullabies from a wide variety of cultures were matched with 28 non-lullaby songs from the same cultures. In analysing the structure of the lullabies, Unyk et al. found the average pitch of the lullabies to be higher, and the number of changes of direction in pitch contour to be fewer. Without any other information, Western adult listeners judged the non-Western lullabies as simpler than the culturally-matched non-lullaby songs. In addition, compared with the non-lullaby songs, the lullabies tended to employ a greater proportion of descending pitch intervals -- consistent with the descending pitch contours prevalent in "soothing speech". The results suggest that infant-directed song shares many similarities with infant-directed speech, and that these features cannot be attributed to acculturation alone.

Walker claims that there is:

"no biological evidence for the a prioriness of such musical intervals as the octave, the 5th, 3rd, etc., nor any sonic configuration. So-called similarities of musical elements across cultures are probably spurious . . . such apparent similarities in musical elements may be satistically unavoidable."

In my own research, I have frequently encountered "sonic configurations" which implicate innate or physiological foundations. For example, in a study of over 26,000 sonorities from works by 17 Western composers spanning five centuries, Huron and Sellmer (1992) showed that the spacing of pitches within sonorities produce spectra whose engergies are linearly dispersed along the basilar membrane. We were also able to show that two popular competing hypotheses were inconsistent with the data. Actual musical practice of these 17 composers is more consistent with cochlear physiology than with the recommendations offered in music theory texts. Nor can one claim that the spacing of notes within musical sonorities is of little musical consequence.

Although there are many distinct phenomena contributing to what is generally called "consonance," one such phenomenon (so-called "sensory dissonance") is strongly linked to a physiological basis (see Greenwood, 1991 for an extensive review). Recently, I amalgamated sensory dissonance data from three classic perceptual studies (including data from non-Western listeners). Based on these data I used a computer program to rank all possible equally-tempered pitch-class sets according to the opportunities for generating harmonic dyads with low sensory dissonance. It was found that the pitch-class sets that provide the least dissonant interval-class inventories are the major diatonic scale, the harmonic and melodic minor scales, and equally tempered equivalents of the Japanese Ritsu mode, the common pentatonic scale, and the common "blues" scale. Of the hundreds of other possible sets that can be drawn from the 12 equally tempered pitch chromas, these sets provide harmonic interval inventories exhibiting the least sensory dissonance (Huron, 1994). Recall that the only models that have successfully accounted for the original sensory dissonance data are neurophysiological models of the cochlea.

Nor is the influence of sensory dissonance limited to scales alone. In a study of Bach's polyphonic writing, I found that the vast majority of variance in the distribution of harmonic intervals could be attributed to the twin goals of the avoidance of tonal fusion and the minimising of sensory dissonance (Huron, 1991). Evidence of the influence of cochlear mechanics is rampant throughout Bach's oeuvre.

In citing the above examples, in no way is it my intention somehow to prove that much of language or music is innate. This is neither my intention nor my hypothesis. (Indeed, in other recent research -- such as Simpson and Huron (1994) -- learning phenomena have figured prominently in my work.) In the first instance all of the above studies are in need of careful reveiw and replication. My purpose in citing this handful of studies has been simply to show that the questions of musical (and linguistic) origins remain wide open, and that it is entirely premature to declare that nothing of any musical consequence is physiologically determined or truly cross-cultural. As I wrote in Huron (1993) regarding the perceptual foundations of voice-leading:

"The question of nature versus nurture is important in music since it has implications for how we view music and what we can expect of music's creative future. If perceptual principles . . . are primarily learned, then it suggests that human perceptual capacities have a remarkable fluidity and adaptability. Musicians might take heart in such knowledge, since it implies that there is great potential for new creative directions in musical culture. Conversely, if such principles are found to be physiologically determined, then it suggests that human perceptual capacities are constrained by fixed biological limits. Musicians might be disheartened by such knowledge, since it implies that musical cultures are destined to traverse within a finite domain of auditory possibilities."

"There are several points to keep in mind regarding the nature/nurture question. First, beliefs concerning this question are always likely to be more strongly held than is warranted by the available experimental research. Second, both extremes of belief have possible negative repercussions. Those who believe that perceptual mechanisms are largely fixed and immutable are in danger of under estimating human capabilities -- and therefore of impeding the development of musical culture. Conversely, those who believe that perceptual mechanisms are highly adaptable are in danger of over estimating human adaptability -- and therefore of encouraging activities that are irrelvant to human needs."

"Although questions concerning the adaptibility of human perceptuion may invite opinions, these questions more properly invite further experimental investigation." (pp. 78-79.)

All musical claims are hypotheses waiting to be tested. For example, when Walker claims that "[t]o the Pygmy there is no such thing as musical ability;" an empirically-inclined ethnomusicologist might test this claim by asking individual Pygmies questions such as: "Is there someone in your village who has an especially good singing voice?" or "Is there someone who doesn't always clap at the proper time?". Standardised questions posed to people of different cultures might help establish whether, and to what degree, Pygmies differ in their views of musical ability. We ought to be sceptical of Walker's claim that "The notion of musical ability is itself a modern invention". How plausible is it that ancient Egyptian, Chinese, or Thai cultures had no concept of musical ability?

In claiming that there is no biological or psychological evidence that humans are pre-determined towards any particular musical or linguistic behaviours, perhaps Walker is trying to stimulate debate by being intentionally provocative. Moreover, perhaps such debate needs to be generated, for reasons of which I am unaware. However, presenting idle speculation as distilled wisdom does a disservice to those who see music psychology as bringing some leadership to the sober and systematic study of music and musical experience.

References

Huron, D. (1991). Tonal consonance versus tonal fusion in polyphonic sonorities. Music Perception, 9 (2), 135-154. Abstract.

Huron, D. (1993). A derivation of the rules of voice-leading from perceptual principles. Manuscript submitted for publication. Abstract.

Huron, D. (1994). Interval-class content in equally tempered pitch-class sets: Common scales exhibit optimum tonal consonance. Music perception, 11 (3), 289-305. Abstract.

Huron, D. and Sellmer, P. (1992). Critical bands and the spelling of vertical sonorities. Music Perception, 10, 129-149. Abstract.

Mehler, J. and Dupoux, E. (1994). What Infants Know; The New Cognitive Science of Early Development Cambridge, MA: Blackwell Publishers.

Pinker, S. (1994). The Language Instinct. New York: William Morrow and Company.

Simpson, J. and Huron, D. (1994). Absolute pitch as a learned phenomenon: Evidence consistent with the Hick-Hyman Law. Music Perception, 12 (2), 267-270. Abstract.