Erenlai - Displaying items by tag: cantometrics
Wednesday, 23 March 2011 10:09

Music as a Marker of Human Migrations

Debate on the question of how and why music varies cross-culturally was recently reawakened by the provocative claim that traces of the ancient migration of anatomically modern humans out of Africa can be heard in contemporary songs (Grauer 2006). Grauer‟s claim drew on data from the landmark Cantometrics Project (Lomax 1968), which remains the only global scientific study of human song. At the time, Lomax‟s causal interpretation of the correlation between culture and music – for example, male dominance causing nasal singing – was highly criticized even by other members of the Cantometrics Project (e.g., Erickson 1976).

While Grauer‟s recent migratory interpretation avoids Lomax‟s pitfall, many of the original criticisms of the Cantometrics Project resurfaced in skepticism about music‟s time-depth as a migration marker (e.g., Stock 2006). Could the acoustic surface of music really reflect ancient connections between cultures? If so, are these reflected in performance features (“singing”) or in the structural features (“song”) traditionally emphasized in Western musicology?

Lomax himself was highly critical of the use of Western musical notation in ethnomusicology, which he saw as emphasizing surface structural features at the expense of deeper performance features. He spent his life developing a performance-oriented approach that was concerned “not with songs abstracted from the stream of vocalizing we encountered on the tapes, but with the stream itself, with „singing‟” (Lomax 1980). Nevertheless, the Cantometric classification scheme that Lomax and Grauer (1968) developed contained roughly equal numbers of features devoted to “songs” and “singing”.

Our own view differs from both Lomax‟s and his critics‟ in that we propose that the structural features of song should have the greatest time-depth to track migrations, especially when applied to group performance in choral songs. Our reasoning is that structural features such as melody, texture and form require greater consensus among singers than the more idiosyncratic variation that goes into performance, such as timbre or ornamentation. Hence, features like scales and rhythms should be more stable over time than features like nasality or rubato.

These claims are testable. As a case-study to examine music‟s time-depth in the context of human migrations , we have examined the traditional choral music of the aboriginal tribes of Taiwan, who have been well-studied in terms of music, genetics, and migrations. (Loh 1982; Trejaut et al. 2005; Diamond 2000). Our primary aim, therefore, was to use existing information about the relative patterns of genetic and musical similarity among the Taiwanese aboriginal tribes to empirically test for the first time whether song structure or singing style has the time-depth required for studying human migrations. Our basic method was to compare music – a marker of unknown time-depth – against the best available marker with a well-established time-depth, namely mitochondrial DNA (Oppenheimer 2004).



Of the 14 officially recognized tribes of Taiwan, eight had a sufficient number of both genetic and musical samples to permit comparative analysis: Amis, Bunun, Paiwan, Puyuma, Rukai, Saisiyat, Tao (Yami), and Tsou.


Genetics: Partial mtDNA sequences for 531 individuals from these eight tribes were taken from the dataset of Trejaut et al. (2005).

Music: YW and SB obtained 364 traditional songs from these eight tribes from commercial and archival ethnomusicological recordings. Restricting our sample to adult, choral songs left 222 songs for analysis. Sample sizes were: Amis=56, Bunun=31, Paiwan=28, Puyuma=32, Rukai=33, Saisiyat=14, Tao=13, Tsou=15.


Distances between samples: Pairwise distances between individual a) genetic, and b) musical samples were calculated based on the number of pair-wise differences between a) mtDNA nucleotide sequences, and b) Cantometric classifications. This is the simplest possible distance measurement, as it makes no evolutionary assumptions about how those differences arose. We reserve more complicated methods that incorporate models of musical and genetic evolution for future studies.

Cantometric classification of the songs was done by VG. Two separate musical distance-matrices were calculated: one using the 15 song-structure characters from Cantometrics, the other using the 14 singing-style characters (see Figure 1 for details about these features). Eight Cantometric characters related to instruments alone were excluded from this analysis.

Distances between populations: For both genetics and music, the 28 possible pairwise distances among the 8 tribes were calculated using an Analysis of Molecular Variance (AMOVA) framework (Excoffier, Smouse, and Quattro 1992). These distances were measured using a statistic called FST, which represents the proportion of variability among individual samples that is due to among-population differences. Thus, it explicitly incorporates within-population heterogeneity, avoiding the assumptions of within-

population homogeneity that plagued Lomax‟s original statistical methodology (e.g., Henry 1968; Leroi and Swire 2006).

Figure 1. Organization of the 15 song-structure (red) and 14 singing-style (blue) Cantometric classification features used in this analysis. Note that our method focuses on the vocal component of the music and therefore ignores 8 classification features related to instruments.

Correlations: The statistical significance of the correlations between musical and genetic distances was tested using the permutation-based Mantel test (Mantel 1967) using 10,000 permutations, with the threshold for significance set at p < 0.05 (one-tailed). This test controls for the fact that the 28 pairwise distances among the eight tribes are not independent of one another.


Correlations between genetic and musical distances were highly significant (see Figure 2), suggesting that patterns of genetic similarity among the 8 tribes were matched by corresponding patterns of musical similarity. This observation makes a strong case for music having an ancient time-depth in analyses of human migrations.

To examine the “song” vs. “singing” comparison, the two panels of Figure 2 show the correlations between genetics and either song structure (Panel A) or singing style (panel B). Both correlations were significant. However, features of song structure accounted for twice as much variance in genetic distance as did features of singing style (song structure: r2=0.27, singing style: r2=0.13).

Figure 2. Scatterplots of the 28 pairwise genetic and musical distances among 8 Taiwanese aboriginal tribes. Genetic distances (y-axis) are based on an Analysis of Molecular Variance (AMOVA) of 531 mitochondrial DNA haplotypes. Analagous musical distances (x-axis) were calculated from 222 traditional choral songs using Cantometric characters of either A) song structure or B) singing style (i.e., performance). Statistical significance of distance-matrix correlations is based on Mantel‟s (1967) test.


Our main finding was that musical similarities among the 8 tribes were significantly correlated with genetic similarities. This provides the first empirical support for Grauer‟s (2006) claim that music has the time-depth required for use as a marker in studying prehistoric human migrations. Consistent with our predictions, the correlations with genetics were stronger when calculated using features of song structure compared to singing style, contrary to Lomax. However, the differences between these features were not nearly as striking as we had predicted. The simplest interpretation is that both singing and songs are useful as migration markers, which makes the overall case for using music as a marker even more persuasive. It allows for a pluralism of musical features that Lomax discounted, most especially with regard to structural features.

Our findings in Taiwan lend strong provisional support for music‟s time-depth in the case of a relatively recent (~6,000 years ago) migration. Whether music‟s time-depth reaches as far back as Grauer‟s Out-of-Africa claim, however, remains an open empirical question.



Diamond J. (2000). Taiwan‟s gift to the world. Nature, 403, pp. 709-710.
Erickson E.E. (1976). Tradition and evolution in song style: A reanalysis of Cantometric data. Cross-Cultural Research, 11, pp. 277-308.
Excoffier L., Smouse P.E., and Quattro J.M. (1992). Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction data. Genetics, 131, pp. 479-491.
Grauer V. (2006). Echoes of our forgotten ancestors. The World of Music, 48, pp. 5-59.
Henry E.O. (1976). The variety of music in a North Indian village: Reassessing Cantometrics. Ethnomusicology, 20, pp. 49-66.
Leroi A.M. and Swire J. (2006). The recovery of the past. The World of Music, 48, pp. 43-54.
Loh I. (1982). The tribal music of Taiwan: With special reference to the Ami and Puyuma tribes. Ph.D. dissertation: University of California Los Angeles
Lomax A. (1980). Factors of musical style. In S. Diamond (ed.), Theory & practice: Essays presented to Gene Weltfish (pp. 29-58). The Hague: Mouton.
Lomax A. (ed.) (1968). Folk song style and culture. New Brunswick: American Association for the Advancement of Science.
Lomax A. and Grauer V. (1968). The Cantometric coding book. In A. Lomax (ed.), Folk song style and culture (pp. 34-74). New Brunswick: American Association for the Advancement of Science.
Mantel N. (1967). The detection of disease clustering and a generalized regression approach. Cancer Research, 27, pp. 209-220.
Oppenheimer S. (2004) The "express train from Taiwan to Polynesia": On the congruence of proxy lines of evidence. World Archaeology, 36, pp. 591-600.
Stock J.P.J. (2006). Clues from our present peers? A response to Victor Grauer. The World of Music, 48, pp. 73-91.
Trejaut J.A., Kivisild T., Loo J.H., Lee C.L., et al. (2005). Traces of archaic mitochondrial lineages persist in Austronesian-speaking Formosan populations. PLoS Biology, 3, pp. 1362-1372.


Patrick Savage(1), Tom Rzeszutek(1), Victor Grauer(2), Ying-fen Wang(3), Jean Trejaut(4), Marie Lin(4), and Steven Brown(1)

(1) Department of Psychology, Neuroscience & Behaviour, McMaster University, Canada
(2) Independent scholar, Pittsburgh, USA
(3) Graduate Institute of Musicology, National Taiwan University, Taiwan
(4) Transfusion Medicine Laboratory, Mackay Memorial Hospital, Taiwan

Photo: Cathy Chuang

Help us!

Help us keep the content of eRenlai free: take five minutes to make a donation


Join our FB Group

Browse by Date

« June 2019 »
Mon Tue Wed Thu Fri Sat Sun
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30

We have 3172 guests and no members online