Thursday, May 31, 2007

19. More Branches

Before proceeding, I need to correct an error. The "haplogroups" exhibiting P/B style are A1-A4, but not A5, which apparently comes later and is more characteristic of mainstream Bantu singing, i.e., neither hocketed nor interlocked but generally in "call and response" form. I see that this error has crept into the maps as well. This probably reflects some ambivalence on my part regarding the presence of call and response vocalizing outside Africa. This is a practice which may or may not have been part of the stylistic mix the "Out of Africa" migrants would have carried with them. Call and response is now found in many parts of the world, but it's history isn't clear, since it's less distinctive than either P/B or Breathlessness and might have been independently invented both inside and outside of Africa.

I want now to consider another very important style family, or in this case superfamily: B3, "Social Unison." Social unison is a type of group vocalizing where all parts share the same rhythm, as in a typical Christian hymn. They may be singing either polyphonically (i.e. harmonizing) or in unison. It's easy to take this type of relatively simple group coordination for granted as a reasonable or logical way for any group to organize itself when singing together. It is, however, rare in subsaharan Africa. For that reason, it's unlikely that the Out of Africa migrants or their descendents were singing in social unison at any time prior to the bottleneck. So when, and how -- and also why -- could such a practice have developed?

An interesting clue is provided by the Australian Aborigines, whose history on that continent goes back at least 50,000, and very possibly over 60,000, years, which could not have been too long after the initial migration out of Africa. All indications, including the genetic research, point to their descent from the original African migrants, yet they have a very contrastive and distinctive physical morphology and a totally contrastive musical style. A clue to their origins could be the existence of so-called "Australoid" peoples in southern India, many of whom, such as the Vedda of Sri Lanka, bear a strong morphological resemblance to Australians. Geneticists Alan Redd and Mark Stoneking have reported genetic results that "link Aboriginal Australian populations with populations from the subcontinent of India," whereas their findings "do not support a close relationship between Aboriginal Australian and PNG (Papua New Guinea) populations..." as many assumed would be the case. (Redd and Stoneking, "Peopling of the Sahul: mtDNA Variations in Aboriginal Australian and Papua New Guinean Populations," in American Journal of Human Genetics, 65:808, 1999).

While the genetic picture remains complicated and controversial, with no one clear interpretation dominating, it is possible, nevertheless, to speculate, as I did in my essay, that the Australians may have originated in southern India as one small band of Toba or tsunami survivors, some of whom remained in India while others eventually made their way to Australia. In other words, it looks as though their musical style could have resulted from the same sort of disastrous culture loss that produced the very different style families B1 and B2, associated with the same genetic and cultural "bottleneck." Only in this case, the break with the original P/B style seems to have been more complete, as there is little the two would seem to have in common.

Style family B3 is subdivided into two main branches, B3a, "Unison Iterative One Beat" and B3b, "Polyphonic Iterative." What both have in common is the "iterative" aspect, i.e., a tendency to reiterate the same note, often several times in a row, and especially at phrase endings. Since this type of melody is so distinctive and unusual, and since it is so often associated with social unison, it makes sense to treat all instances as stemming from the same root. In this case we would appear to have no one surviving style that exemplifies the root style and must therefore conjecture that it may have been lost.

B3 subdivides into a polyphonic and a unison branch, conveniently separating Polynesian iterative social unison, predominantly polyphonic and open throated, from what I've called "Unison Iterative One-Beat," a relatively harsh, tense-voiced unison style, often supported by a very simple one-beat accompaniment on drums or idiophones such as clappers or rattles. As I mentioned earlier, in section 11 of this blog, this is a style apparently held in common by two geographically very distant groups, Australian aborigines and native Americans. The two styles, labeled B3a1 and B3a2, have a great deal in common, as already discussed, strongly suggesting that despite the truly enormous geographical distance, they might well stem from a single root -- a root that would have formed very early after the bottleneck, most likely in India.

To understand just how similar the two styles can be, let's pause to compare them. Here, first, is an example from western Australia, followed by a Flathead Indian Powow Dance (Both examples are accessible via Smithsonian Folkways: Note the very striking similarities: tense voices, sung in unison, frequent note iteration, especially at phrase endings, gradually downward melodic trend, and one-beat rhythmic accompaniment. The most significant difference is a tendency for most Australian melodies to employ narrow intervals, while most Amerindian songs favor wide intervals. A great many other similar examples could be provided for comparison, especially among Amerindian groups from the Prairie and Plains regions.

Consulting the third of the little maps, labeled "Post-bottleneck founder effects," we see B3 splitting off in three directions, one toward Australia, another toward Melanesia (where both the unison and polyphonic variants of B3 can be found) and a third pointed northeast, in the direction of Bering Strait. In the following map, "Continued Migrations," we see B3a2 in Australia, and B3a1 in North America, both stemming from a hyphothetical root, B3a, which may have at one point migrated, as in the map, to somewhere in or around China, but is apparently no longer to be found anywhere in Asia. The polyphonic variant, B3b is shown safely ensconsed in Melanesia, from whence it has made its way all points east to the islands of Polynesia.

Wednesday, May 30, 2007

18. The Bottleneck Continued

To better understand the bottleneck theory, let’s take a closer look at the set of evolutionary maps first presented in section 12. The map at the upper left, labeled “Out of Africa,” represents the earliest stages of the “Out of Africa” migration, which, according to most proponents of this model, proceeded eastward along the south Asiatic coast, all the way from Yemen, along the coasts of India, Southeast Asia and Indonesia (all one landmass at that time) to the Sahul (New Guinea and Australia, then one single landmass) and northward along the East Asiatic coast.

Musically, we see style families A1 through A5 represented in Africa. All appear to have been present among the original “Out of Africa” band, so the red arrows we see progressing along the Asiatic coast are intended to represent all five versions of P/B style. (Note the offshoot in northern India, labeled A3 & 4 – we’ll deal with that presently.) This map represents the situation after our migration “wave” reached the southeast coast of Asia, prior to the bottleneck event. Various colonies would presumably have been left all along the Asiatic coast.

Proceeding to the next map, labeled “Bottleneck Event,” we see a hypothetical “Tsunami” represented, which could have had a devastating effect on all the colonies along the coast of the Indian Ocean. Since the migrants may well have had a coastal culture, “beachcombing” for shellfish and other ocean-based food sources, a tsunami might well have destroyed entire populations. It’s also possible that the eruption of Mt. Toba was the cause of the bottleneck. In any case, a great deal of evidence, both genetic and musical, points to a major bottleneck of some sort, caused, in all likelihood, by some enormous natural disaster.

It’s important to note that any human colonies located beyond the Indian Ocean, on the east coast of Indonesia, the Sahul or the Asiatic mainland would not have been affected by either an Indian Ocean tsunami or the Toba explosion. (Since the prevailing winds would have been westerly, the ash from Toba would not have affected most areas to its east.) And it is indeed in these areas that we find, even today, most of the remnants of P/B styles A1 through A5, as indicated on the phylogenetic tree.

As for the rest, the people who’d been living along the Indian Ocean coast? Could they have survived? And if so, what would their lives have been like? We have no way of knowing, of course, but from the evidence we now have, both genetic and musical, the ensuing population bottleneck might well have been responsible for most of the diversity, genetic, phenotypic, cultural, and musical, that we see today.

Moving to the next map, “Postbottleneck Founder Effects,” we see the distribution of style families B1, B2 and B3, in their early stages, all presumably rooted in major changes, biological, environmental, social and cultural, that could have left their mark on whatever individuals would have survived. The discontinuities we have been discussing in the musical styles of so many indigenous peoples today could thus be explained by the cultural discontinuities produced by this one single major disaster.

B1 is represented in red, like the A superfamily, because, of all the post-bottleneck styles, it is still identifiable as African. This is what I have called “canonic-echoic” style, a form of canonically overlapping imitation very possibly derived from A4, “Canonic Interlock,” as indicated by the scored line linking A4 to B1. The styles can be quite similar, as is evident when we examine the list of individual traits (haplotypes) in each. The major difference is the presence of LC, for “loosely coordinated,” in B1. We find LC only very rarely in subsaharan Africa, where almost all forms of music are very tightly coordinated indeed. There are in fact no real traces of B1 anywhere in Africa, strongly suggesting that it must be a post-bottleneck derivative.

We have already discussed B2, “Breathless Solo,” at some length. What the map implies is that this style could have had its origin on the Indian subcontinent, rather than its current home, northern Eurasia, as a result of the same bottleneck event we’ve been discussing. Could it be the result of some later event, possibly some other disaster, or the effects of the ice age? Possibly, but it would have had to be rooted very early on in human history to have made its way into the many different, remote, places where it’s found today. And we must remember that Paleolithic Europe was dominated by a very similar, reindeer oriented culture, as far south as Spain, so groups such as the Saami, of northern Scandinavia, could be marginalized descendents of originally mainstream Paleolithic Europeans, the westernmost wing of bottleneck survivors who moved off in all directions from an Indian homeland. If their “breathless” singing style originated after one such wave had headed north to Siberia, then it would be difficult to explain how the same style could have made its way to Spain prior to moving north to Scandinavia, where the reindeer people would have been forced north during the Neolithic. Sorry about that long and confusing and admittedly speculative sentence, but it does tell you something about the scale of my thinking on this topic, which is wide ranging indeed.

While B2 seems in many ways almost the opposite of any of the African A styles, there are some very interesting points in common, as indicated by the scored line linking B2 with A2. The two styles are both characterized by yodel (especially in the well known Saami “joik” songs, but also reflected in the heavy glottalization found throughout this style area), continuous vocalizing (albeit interrupted by gasps for breath in B2), wide intervals, and an emphasis on “nonsense” vocables. It is thus possible to see how B2 might have emerged directly from P/B, but in a post-disaster setting where communal life might have broken down and individuals could have been forced into isolation for long periods.

I tend to see B2 as the root of other important styles that may have come later, through a gradual process that better fits our usual notion of “evolution.” The tree diagram shows a derivative of B2 branching off to a substyle called “Phrased Solo,” i.e., a development from the “breathless,” apparently unphrased singsong of Paleosiberian singing, to a type of melody organized according to what we can now call “phrases,” i.e., syntactic elements delineated by breaths. At least two branches survive today, one being B2A1, “Elaborate Solo,” a type of highly embellished solo vocalizing characteristic of many Asiatic groups. The Eastern branch of this style, especially in China, Korea and Japan is characterized by considerable glottal activity, possibly a clue to its origins in B2. The other branch, B2A2, is characterized by much less embellishment, lack of glottal emphasis, and strophic (or verse) form of the type most closely associated with the European lyric song and ballad. This is a style that appears to have its origin in Central Asia, where strophic songs are still an important part of the bardic repertoire.

Monday, May 28, 2007

17. The Bottleneck

We may now return to the phylogenetic tree introduced in section 12. Note the break between style family (or “haplogroup”) A4 and the families just above it, to the right, B1, B2 and B3, all rooted at the same place, the thick horizontal line representing the “Bottleneck.” Note also that there is no solid line connecting any of the A families with any of the B families. What is represented here is a break in continuity.
What could this mean? If "modern" humans originated in Africa, with one small group migrating from that continent to populate Asia, Europe, Oceania and the Americas with their descendents, then it’s only logical to assume their music would have migrated with them. According to this “Out of Africa” model, we would expect that either the original African style family would persist, or gradually evolve, or both. However, as should be clear from our discussion of the “old problem” faced by the early comparative musicologists, as well as my original tree diagram, the expected continuities simply aren’t there.

For example, let’s consider style family B2, “Breathless Solo.” (Equivalent in my original tree to B, “Breathless” style.) What is this style and why is it important? By “breathless” I refer to an unusual and highly characteristic feature of this type of vocalizing, in which a solo singer produces a continuous stream of notes as a sort of musical “run-on” sentence. Breathing often appears arbitrary in this style, i.e., not coordinated with the melodic structure. In many cases, the singer appears to be attempting to continue for as long as possible and then audibly gasps for breath with no apparent regard for where he or she is in the melody. We called this style “breathless” because it doesn’t seem to take the singer’s need to take a breath into account. This is radically different from what we would ordinarily expect, since the coordination of the breath with important syntactic junctures is a very much taken for granted aspect of both the musical and linguistic phrase in the great majority of cultures. Other distinctive features of this style are the use of nonsense vocables, wide intervals and a voice quality characterized by heavy glottalization and nasality.

“Breathless Solo” appears to be the dominant style for a very widespread family of so-called “Paleosiberian” reindeer-oriented hunting societies stretching across the length of northern Eurasia, from Scandinavia (the Saami Laplanders) through vast stretches of northern Russia and Siberia (Samoyed, Evenk, Yukaghir, Kamchatka, etc.) to the northernmost island of Japan, Hokkaido (the Ainu). As with the Pygmies and Bushmen of Africa, the very widespread distribution of such a distinctive musical style (not to mention lifestyle) over such a vast and difficult terrain, strongly suggests a common cultural root dating back, in all likelihood, to the upper paleolithic.

So how do we get from the mellifluous, flowing, highly integrated, interlocking, polyphonic style of “haplogroups” A2 through A4 to the rather singsong, runon, monophonic vocalizing of the Paleosiberians represented by B2? If the replacement theory is correct, the Paleosiberians must share a common genetic root with the Africans. Could something have happened during the migration out of Africa through the southern coast of Asia that might have suddenly caused such a drastic change? Yes, or at least that’s the theory.

That “something” would have been what I’ve been calling the “bottleneck,” i.e., a drastic reduction in population size that would have wiped out much if not all the genetic variation in a given population, with the possibility of an equally drastic alteration of its culture. This isn’t my idea, by the way. There are several references in the genetic literature to a major population bottleneck thought to have occurred at some point between 30,000 and 130,000 years ago. Geneticist Steven Oppenheimer, in his book The Real Eve, is more specific, attributing such a bottleneck to the historically verified super-explosion of Mt. Toba, in Sumatra, sometime around 72,000 years ago. Oppenheimer has written of the “paradox of the Indian genetic picture, in which the genetic trail of the [Out of Africa] beachcombers can be detected, but the bulk of Indian subgroups . . . are unique to the continent, especially among the tribes of the southeast. This is what we would expect for a recovery from a great disaster” (p. 193).

Oppenheimer’s theory zeroes in on the Toba eruption as the disaster in question since it was so vast and occurred when, according to his timeline, the Out of Africa migrants could have left colonies all along the south Asiatic coast, precisely in the path of a lethal ash cloud that could have, in his words, precipitated a “nuclear winter,” wiping out all or almost all living things in its path.

Oppenheimer’s theory has been questioned, since many archaeologists doubt that the Out of Africa exodus could have occurred prior to 72,000 years ago. However, there is another possibility for a disaster that could have been just as devastating, a Tsunami of the type we experienced two years ago, which killed almost a quarter of a million people along the various coasts of the Indian Ocean. A similar event, centered due south of the Indian subcontinent, at some time between 80,000 and 50,000 years ago could have wiped out all or almost all the Out of Africa migrants living along the entire coast of southern Asia, from Yemen to Indonesia.

To be continued . . .

16. A multiregional tree

To get a better handle on our old problem, let's take another look at the evolutionary tree I came up with back in 1966. You will note that, unlike the new one, which has a single root, this one has three: A. "Complex Group"; B. "Solo"; and C. "Simple Group." Believe me, I would have loved to trace all these different styles to a single source, but could find no way to make that work. The evidence available at the time suggested at least three more or less independent lines of development (the "X" I placed at the top was more wishful thinking than anything else). And, as I stated in the previous post, this interpretation seems more consistent with the "multiregional" rather than "replacement" (Out of Africa) theory of human evolution. To briefly explain, the replacement theory sees all "modern" humans -- homo sapiens sapiens -- as originating in Africa and then spreading to the rest of the world, where they replace all the various "archaic" peoples, such as homo erectus and neanderthals, who'd been living in Asia and Europe for millions of years prior to the African expansion. The multiregional theory sees archaic humans developing independently in Africa, Europe and Asia for millions of years and then gradually converging more or less in-place into modern humans through a long, slow process of adaptation.

Interpreting the 1966 tree in terms ofmultiregionalism, style area A, headed by "Pygmy Style," could be interpreted as representing Africa; B, headed by the solo, "Breathless Style" characteristic of the hunter-nomads of Siberia, could be seen as prototypical for Asia (and regions of Europe influenced by Asian culture); and C, characterized by a relatively simple, "hymn-style" type of group vocalizing in rhythmic unison, might be interpreted as accounting for Western and Northern Europe. According to such a scheme, one could argue that music might well have been invented independently at least three times, in Africa, Asia and Europe, by three different groups who had evolved separately into AMH -- Anatomically Modern Humans.

There are problems, however, with this interpretation. For one thing, style area A extends beyond Africa, to Europe, Southeast Asia and even South America and Oceania. And style area C extends well beyond Europe, to Oceania, the Americas and Australia. Moreover, at that time I wasn't aware of the very strong connection between P/B style and the music of certain groups in New Guinea and Island Melanesia, truly dramatic similarities of both vocal and instrumental practice, that, as I argue in my essay, make independent invention highly improbable. While one could argue that each of these very different types of music making might have had an independent origin, it would be very difficult indeed to trace the migration of each of these styles into so many widely separated regions of the world.

Shortly after my 1966 presentation, I left the Cantometric project for SUNY Buffalo, to pursue a Ph. D. in music composition -- and from then on concentrated much more on creative work than musicological research. From time to time, however, especially when teaching classes in world music, I would puzzle over that same venerable and vexing problem of the various style families of the world, how they relate to one another, and whether or not they share a common origin.

Which returns me to the new phylogenetic tree, inspired by the new genetic research -- and the new and very exciting possibilities offered by the Out of Africa (i.e., "replacement") model. We are now, I hope, in a better position to understand the "bottleneck" that appears therein, which I will finally address, in the next installment.

Sunday, May 27, 2007

15. An Old Problem

Before dealing with the "bottleneck" that breaks the continuity of my phylogenetic tree, I need to back up a bit, to consider an old problem, a longstanding dilemma the bottleneck is intended to address. During the heyday of Vergleichende Musikwissenschaft (aka Comparative Musicology), many gifted scholars struggled vainly to place the various types of "primitive" music of which they were aware into a coherent evolutionary picture. Part of the problem was their methodology, focused almost exclusively on Western theoretical concepts: scale formations, melody types, phrase structure and tuning systems, none of which proved particularly useful for broad-based comparative research. Even more of a problem, however, was the puzzling variety of different musical types and styles with which they were confronted. It seemed logical to assume that the earliest music would be the simplest, with scales of only 2 or 3 tones and simple, repetitive structures. And indeed there were some tribal peoples whose music seemed to fill the bill. But there were other "primitive" groups whose scales contained 7 or even more notes, with musical structures that weren't at all easy to characterize. Especially puzzling was the fact that there was no one kind of music making that typified all extant hunter-gatherer groups. Some did seem limited to only a small number of notes, sung solo or in unison, but others favored far more elaborate styles, with wide ranging melodies, polyphony and even counterpoint. Still others appeared to be somewhere between these two extremes. Given such a wide array of different "primitive" types, it wasn't difficult to conclude that music might not have had a single origin after all, that it must have been independently invented in different places at different times, a notion still widely held today.

Alan Lomax conceived Cantometrics as a style-oriented methodology better suited to the complexities of world music than the analysis of scales and tuning systems, an approach he rejected as both ethnocentric and ineffective. Back in the early Sixties, as Lomax's collaborator-assistant, I had the extraordinary opportunity to participate in putting his hypothesis to the test, a process that involved the encoding and analysis of a wide array of musical practices from a great many different corners of the world. Thanks to a dramatic expansion in field recording activity, due to improved lightweight equipment, an accelerating interest in non-western music, and Lomax's tireless efforts in building his archive, our sample far exceeded anything that had been addressed before. And, as far as I was concerned, the results did, rather dramatically, bear Lomax out. The cantometric approach provided a far more comprehensive, coherent and convincing overview of world music than anything preceding it.

In December 1966, at the New Orleans meetings of the Society for Ethnomusicology, in a paper titled "Patterns of World Song Style," I presented the results of a study of my own, representing a first attempt to produce what would now be called a "phylogenetic tree" of musical evolution. I still have some copies of the hand-drawn, xeroxed graphic I distributed. (The reader is urged to right click on the link and select the "Open in a new window" option.)

Thanks to the extensive sampling provided by the cantometric database, which I queried systematically, plus a certain degree of subjective inference based on careful listening, I was able to put together an admittedly hypothetical but nevertheless coherent picture. Studying it today, I see that it strongly resembles the new phylogenetic tree I've been discussing in several very interesting respects. The differences interest me even more, however, because my old effort could easily be construed as a model based, not on the Out of Africa theory, but the more traditional multiregional theory that has been its chief rival for many years, in a raging controversy that has still to be convincingly resolved.

Friday, May 25, 2007

14. More on the Tree

If we concentrate only on what could be called "superhaplogroup A," i.e., all the categories labeled A1 through A6, on the left side of the display, we see a clear continuity from one branch to the other on the phylogenetic tree. In most cases each branch differs from the one immediatly adjoining it by only one or two traits. Note that, as far as Africa is concerned, both A1 and A2 are found only among Pygmy and Bushmen groups. It's only when we reach A3 that we see some Bantu (i.e., non-hunter) groups included. Note also that the symbol Y, for yodel, is in parenthesis for this category, meaning that the trait is not always found there. This reflects the fact the yodeling can be commonly found among almost all Pygmy groups as well as many (though not all) Bushmen groups, but is only rarely associated with P/B style among other African groups, even among those peoples whose vocalizing is interlocked in a manner strikingly similar to the P/B groups. (By the way, it's important to understand that there will always be exceptions to any generalization I make on this blog. I'm presenting a very broad and also provisional overview that may well be revised in future.)

Note also that A5, Call and Response, associated with the Bantu "mainstream," is located relatively high on the tree, reflecting estimates that the so-called Bantu groups genetically diverged from their P/B ancestors at some point roughly around 18000 years ago (see post 11. "Standard Candles," for details). Another very important feature of the Bantu "mainstream" is the extraordinary development of instrumental music, but as the tree is limited to vocal styles, that isn't shown. What's implied by the tree is that P/B style can be regarded as prototypical for Bantu mainstream style, with polyphonic call and response seen as an outgrowth from hocketed interlock.

Looking now to the far left column of the diagram, we see the phrase "Out of Africa," with a horizontal line marking, very crudely, the point in history where, hypothetically, the first band of "modern" humans left Africa for Asia. Everything below this point can be understood as representing the state of music in Africa prior to that event. It should also be noted that style families A1 through A4 were apparently spread to other parts of the world as a direct, and early, result of the "Out of Africa" migration. (See the listings above the tree for references to populations still vocalizing in P/B style.) Again, instruments are not represented here, so it's important to mention that, in all likelihood, the first "Out of Africa" migrants were probably carrying: sets of endblown pipes (unbound, but also possibly bound into panpipes), whistles, and sets of horns and trumpets, slit drums, stamping tubes, all performed in "P/B style," i.e. with interlocked hocket; endblown flutes, musical bows, jews harps and, probably, simple membranophones (skin drums). How do I know all this? From paying attention to the distribution of all these instruments among certain indigenous peoples today. (See my essay for details.)

Just above "Out of Africa," on the leftmost column, we see the word "Bottleneck." This may be at once both the most controversial and most interesting aspect of the tree. In order to understand what's at stake we'll need to delve more deeply into certain very basic issues faced by both the comparative musicologists of "olden times," and the hip, young genetic anthropologists of today.

Thursday, May 24, 2007

13. Understanding the Tree

The only way for me to fully explain the Phylogenetic Tree and Evolutionary Map would be to present a full semester (or better, full year) course on world music. To really get it, one would have to do a lot of listening and also a considerable amount of analysis, both traditional and Cantometric. For now, the best I can do is explain in very general terms how the tree works and some of its basic elements. Here, again, is the link to the Tree, which I recommend you download on a separate page (right click on the link and select "Open in a New Window") and keep handy for easy reference.

If you scroll to the lower left, you'll find the "root" of the tree and some of the earlier branches. Since it is customary when designing genetic trees to root the tree in the lineage of the closest primate (usually the Chimpanzee), I've considered that possibility for music and discovered that in fact there might be some basis for rooting my tree in the "musical" culture of certain primates. This aspect is of course an extremely speculative, but also, I think, quite interesting, feature of my tree -- but I'll leave further discussion of primate vocalizing for a future post.

Moving upward and to the right, we find the human root of the tree, labeled, appropriately, "A1." You'll note that in addition to such purely indexical symbols I have provided descriptive titles, in this case "Shouted Hocket." A more complete description of each style family (or musical "haplogroup") is provided in the abbrevitions located just below each label. The key to these abbreviations can be found in the rightmost column. Under "Shouted Hocket," we find the following "haplotypes": Hk CV It Y N RT. Hk stands for "hocket," a musical procedure in which an idea, usually a brief motive, is broken up between two or more parts, often but not always dovetailed with one another. In my view, this practice is one of the keys to an understanding of humankind's earliest music, as will be evident as we proceed. CV stands for "continuous vocalizing," i.e., the production of a continuous stream of sound, with no pauses and no coordination between breathing and any aspect of musical structure, e.g, phrasing. CV tends to sound like the musical equivalent of a run-on sentence. Where more than one performer is involved, the breathing is usually staggered and we hear no pauses. (If there is only one performer, the musical flow is often interrupted by audible gasps for breath that appear to come at arbitrary points in the melody.)

"It" stands for "iteration," where the same note tends to be repeated, while sung vocables either repeat or change. Y stands for yodel, a well known type of vocalizing, centered in the glottis, characterized by rapid and very fluid alternations between head and chest tone. Yodel is very common among Pygmies, Bushmen and certain other indigenous peoples. N stands for the frequent use of "nonsense" vocables. Since in many cases we can't be sure if any given vocable has a meaning or not, Cantometrics defines "nonsense" largely in terms of repetition. (If the same vocable or set of vocables is regulary repeated it is coded as "nonsense" regardless of whether or not it carries some meaning.) RT stands for "relaxed throat." While vocal tension (line 33 on the Cantometric coding sheet) is not always easy to assess, it would seem that relaxed, "open throated" singing is characteristic of just about all Pygmy and Bushmen vocalizing -- and tends to be associated with yodeling wherever it is found.

These six traits are by no means the only ones that characterize shouted hocket, but they do seem among the most prominent and easily identified. However, I still need to explain why I call this "shouted hocket" in the first place and why it's located at the root of our musical tree. Since there's no substitute for actually listening, here's an example, from the CD set "Musical Anthology of the Aka Pygmies," recorded by Simha Arom : Aka esime.

This is a brief example from the end of a much longer track, of a type of Aka vocalizing called (according to Michelle Kisliuk) "esime," which, as she describes it in her excellent book Seize the Dance, functions as a kind of interlude between songs. As such, it is often overlooked and has, in fact, only rarely been recorded. What particularly interests me are the places where the shouting is tossed rapidly, and rhythmically, back and forth between the "leader" and the group. Compare with an example of a similar practice from the Dani people of the Eastern Highlands of New Guinea: Dani Hocket. This is, in fact, a type of vocalizing found quite commonly among certain indigenous groups, in Africa, New Guinea, Island Melanesia and elsewhere. I think it could be important, first because of its worldwide distribution among so many peoples whose ancestry might well go back to the original "Out of Africa" migration; second, because it's easy to see how it could be the prototype for the more intricate varieties of contrapuntal hocket we find so often among the same groups; and third, because of its striking similarity to certain types of primate vocalization (more on that in a future post).

At this point, please follow the vertical line upward to the list of tribal groups at the top. These are the groups for whom I have examples, either in the Cantometric database, or from my own listening experience (usually both), which vocalize in some version of the "shouted hocket" style. That style may not be the most important or most commonly found in those groups, but it is known to be at least present among all of them.

Let's now move up to the next branch of the tree, "A2. Interlocked Hocket." You'll notice that one of the "Shouted Hocket" traits is missing: It, or iteration, which is characteristic only of the simplest type of hocket, which lacks any melodic structure. Instead, we have, in addition to all the other traits, some new ones: Int, for "interlock," WI, for "wide intervals," and P, for "polyphonic." I won't go into any more details for now, but this should give you a good idea of how the tree works in terms of musical families (the "haplogroups," such as A1 or A2) and the individual style traits ("haplotypes," such as Hk or Int) that characterize them.

Looking more generally at the tree as a whole, it's important to recognize that it is broken down into two different segments, the first labeled "Unaffected by Bottleneck," the second "Affected by Bottleneck." This fundamental break makes my tree very different from any other I'm aware of. We have a strong tendency to think of evolution in terms of gradual change, but there are very good reasons to see the evolution of early music in a different light. For more on this I must again refer you to my essay (which as I stated in one of my earlier posts I will be happy to share with anyone who emails me with a request. The address is:, though I hope to cover this extremely interesting question in a future post as well.

Note, by the way, that all the African groups are located in the leftmost section, "Unaffected by Bottleneck." There are no African groups in the rightmost section. In my view, this is an extremely important aspect of this tree. Note also that there are many non-African indigenous (or "folk") groups along with the African ones listed at the top of the leftmost section. As I see it, the close connection of these groups with variants of Pygmy/Bushmen style (haplogroups A1 through A4) is possibly of great significance, as they may well represent survivals from the musical practice of the original "Out of Africa" migrants.

Wednesday, May 23, 2007

12. A Phylogenetic Tree

I'm going way out on a limb (no pun intended) in this post, offering a phylogenetic tree, representing my own best thinking so far, based to some extent on the Cantometric data, but also on other aspects of musical style that I'm aware of -- and some personal, though IMO well informed, inferences. It represents, I suppose, a rough summary of the hypothesis presented in my "Echoes" essay. There are many years of research behind this, so I do think it's meaningful -- but it is also somewhat subjective, i.e., not automatically produced by the raw data. I've been working on this for several months, tweaking it from time to time as new wrinkles occur to me. I've been reluctant to submit it anywhere for publication, 1. because I'm not completely sure how accurate it is and 2. because I may be one of the very few capable of fully understanding (and appreciating) it. Few ethnomusicologists alive today have paid much attention to such issues.

(Click on the tree to enlarge it -- or better yet, right-click and select "Open in New Window")

Note the column on the right, a listing of various musical characteristics, mostly drawn from Cantometrics but not all. In the map I treat these as analogous to what geneticists call "haplotypes," grouped together by analogy with "haplogroups." [Correction (as of July 29, 2007): I now believe my terminology to be in error. The musical characteristics in this column are analogous to genetic markers, i.e., mutations, not haplotypes.] All "haplogroups" beginning with the letter A represent styles or style variants originating in Africa and surviving more or less intact in other parts of the world to this day. All those beginning with "B" represent styles that appear to have originated as the result of a major bottleneck, both genetic and cultural, that occurred during an early phase of the Out of Africa migration. (Bottlenecks are severe reductions in population, due to the splintering off of certain groups or some form of mass destruction due to a catastrophe of some sort, a famine, flood, war, etc.) For more on this, see my essay.

A second display, a set of maps, is an attempt to apply the "haplogroups" presented in my tree to a hypothetical picture of the evolution of musical style according to the "Out of Africa" migration picture featured in my monograph, essentially the one presented by Steven Oppenheimer in his book "The Real Eve."

In the first little map, titled "Out of Africa," all the arrows are red, representing the five variants of the original "Pygmy/Bushmen" style (A1 - 5 on the Phylogenetic Tree) that, as I see it, must have spread along with the original "out of Africa" migrants, following Oppenheimer's coastal route, all the way to Sundaland and beyond, turning the corner around the Moluccas, I guess and then continuing north along the SE Asiatic coast.

Mini-map 2, "Bottleneck Event," is an attempt to picture a catastrophic event that could account for the musical gap we find between Yemen and Myanmar, where little or no evidence of Pygmy/Bushmen style can be found today. According to Oppenheimer, there is a very similar gap in the genetic evidence, though, as I understand, not everyone agrees about this. As I see it, only some sort of catastrophe at some point from, say, 75,000 to 50,000 years ago, can explain all the very different musical styles we find in the world today, especially the styles represented in the phylogenetic tree as B2 and B3 and their derivatives. So this map, unlike any other I've ever seen, is not based solely on continuity, but contains an abrupt break, representing the effects of the bottleneck on the various surviving groups.

Oppenheimer seems convinced that the Toba explosion can account for the bottleneck, but since that theory is controversial, I decided to present an alternative possibility that could have had a very similar effect, a Tsunami centered somewhere south of the "point" of India, occurring sometime between, say, 80,000 and 50,000 years ago. If in fact "Out of Africa" was a coast-oriented culture, then such an event could have wiped out all or almost all the various colonies strewn along the coast of the Indian Ocean -- but spared those who had already made it around the corner, to the coast of E. Asia and some of the Islands to the East of Sundaland. As I see it, this would explain a great deal about a lot of things, not just music.

The other two little maps should be more or less self explanatory, but of course there would be a great deal to say about each and every arrow represented.

All of the above makes a great deal of sense, at least to me. But I'm afraid that to most others it will look extremely speculative, if not meaningless. I would appreciate comments and questions, nevertheless, to help me get a sense of whether or not others are able to find anything interesting or useful in this. I'd also very much appreciate suggestions for improving it, especially from anyone with experience in concocting phylogenetic trees.

Tuesday, May 22, 2007

11. Standard Candles

If, as I argued in the previous post, musical style can be regarded as not only a "neutral marker" but an actively conservative force, inherently resistant to change, then it has the potential to become an especially powerful tool for the reconstruction of cultural history and, in cooperation with genetics, the various "Out of Africa" migrations of early humans.

Some geneticists have claimed that the mutation rate for certain types of DNA is statistically predictable, providing them with a kind of clock with which to estimate when certain mutations occurred. A linguistic method called Glottochronology made a similar, though more dubious, claim for language. As far as I have been able to determine, no such "internal clock" would seem to exist for musical "mutations." But there is a method from the realm of Astronomy that could, I think, be applied to music: the "standard candle."

An example of a "standard candle" would be the Cepheid Variables, a type of star with a known "signature" of variability that makes it easy to recognize. One can also calculate its absolute brightness from the frequency of the variation, a discovery that made it possible to use Cepheids as "standard candles" to determine the distance to any stars or galaxies in the same vicinity.

We can apparently do something similar to help us estimate certain dates important for cultural history. For example, the genetic estimate I've already alluded to, of 72,000 to 102,000 years ago for the initial separation of the Biaka Pygmies from the hyphothetical founder group can be used as a "standard candle" to help us determine the age of P/B style, which we could then consider to be at least 72,000 years old. We could then look for any other practices common to both Bushmen and Pygmies that might also date to the same time. Since the dates for the Bantu expansion (between 2000 and 4000 years ago) seem also to be known, that would by the same token provide us with a conservative estimate of at least 2000 years ago for the origins of mainstream Bantu style. However, the same genetic studies that gave us the Biaka dates also give us dates for the divergence of a Bantu population from Senegal ranging between 17,900 and 23,200 years ago. On that basis, we might want to revise our estimate for the origins of this musical style to some time between 2000 and 17,900 ya.

As I mentioned in my previous post, a conservative estimate for the earliest arrival of humans in the Americas is around 10,000 years ago, which could put the age of Amerindian unison/one-beat/iterative style at roughly the same date. In my essay, however, I noted a strong resemblance between the music of many native North American groups and the Australian Aborigines, who also tend to sing iteratively, in unison, with one-beat accompaniment. Many Aboriginal songs have a markedely descending or even "terraced" melodic contour, making them especially close to the Indians of the Plains and Prairie, many of whose melodies are also terraced. If both styles do indeed stem from the same cultural root, we could use the estimated genetic-archaeological date for the first entry of humans into Australia as our "standard candle." That would be somewhere between 60,000 and 40,000 years ago. Since the Australians appear to have been essentially isolated on that continent from that time until the colonial period, with little or no outside contacts, it seems safe to conclude that the common root for both traditions, assuming they have a common root, would have to be dated to some time prior to the first landing of the aboriginals on the Australian shore.

We now have a set of dates that we can use in the construction of some sort of very simple phylogenetic map for tracing the "geneology" of the above musical traditions. Using the most conservative dates, we would then place the origins of P/B style at some point prior to 72,000 years ago. The next segment would branch off the root at 40,000 years ago, as the possible origin of unison/one-beat/iterative style, with a sub-branch, signalling the origin of the Amerindian version, at some time between 40,000 and 10,000 years ago. Another, separate branch, from the original root, marking the origin of mainstream Bantu style, would begin sometime between 2000 and 17,000 years ago.

These are only a few of the possibilities for estimating the origin dates and constructing phylogenetic trees of certain musical, and associated cultural, traditions, using a "standard candle" provided by genetic and/or archaeological research.

Monday, May 21, 2007

10. Is Music a Neutral Marker, continued

A genetic marker is considered neutral if it is not affected by natural selection, i.e., not affected by changes in the environment or any other outside influence, but continues unaltered until, suddenly, a mutation occurs. Then, after the mutation, the altered marker continues, generation after generation, unchanged, until the next sudden mutation. In my view there is good evidence that music, or more accurately, musical style, could also be considered a neutral marker in roughly the same sense. (The same conclusion might apply to dance as well, but that very important question will have to be tabled for now, as I lack sufficient knowledge to adequately consider it.)

Why do I feel so sure this must be so? Because there are certain musical style families whose distribution suggests that they have, indeed, remained essentially unchanged over extremely long periods of time, large geographical areas, and very different types of environment. The most dramatic case, as I have already explained, is Pygmy/Bushmen style. But there are several others, such as the "mainstream" style of most African Bantu peoples, characterized by overlapping "call and response," harmonized vocalizing, the importance of drumming, use of polyrhythms, etc., a style that not only pervades subsaharan Africa but, due to the effects of slavery, has gone on to have a profound influence on the music of the Americas. As the "Bantu expansion" is thought to have occurred anywhere from two to four thousand years ago, this style could therefore be at least two thousand years old. Another good example would be the very different, remarkably homogenous "mainstream" style of the great majority of native North American tribal groups, characterized by unison singing, "one-beat" percussion accompaniment, nonsense vocables, wide intervals, moderately tense, raspy voices, and a highly idiosyncratic manner of forming melodies, where most notes tend to be on the beat and the iteration of the same note over different vocables is common, especially at phrase endings. Since these peoples are thought to have entered the Americas at least 10,000 years ago (very likely much longer), we can conservatively estimate that this style must be at least 10,000 years old.

The existence of such broad ranging, clearly identifiable style families is, in my opinion, strong evidence that musical style can indeed be considered a neutral marker. (This is a conclusion, by the way, that goes very much against the grain of most current thinking in Ethnomusicology, where the very notion that one could take music "out of context," to consider it independently from the environment and its functions in the immediate society in which it is performed would be considered not only meaningless, but ethnocentric.)

If music (and possibly dance as well) is indeed a neutral marker, does this make it unique among all other aspects of culture? Certain genetic anthropologists, such as Luca Cavalli-Sforza, have looked to language as, roughly, the cultural equivalent of a genetic marker and have consequently paid a great deal of attention to the distribution of language families worldwide. Comparing language with music, however, we find some rather important and instructive differences. For one thing, language is much more complex than music, with a more rigidly defined syntax, and an extremely important dimension either lacking or undeveloped in music: explicit reference, the semantic dimension, the realm of words, which music completely lacks. It is also far more important than music as the basis for all sorts of everyday interactions, musings, self-expressions, introductions, announcements, flattery, gossip, swearing, taunting, courting, education, instruction, interaction with other groups, etc., and is consequently, unlike music, a totally indispensible, ubiquitous and "visible" aspect of ordinary life. All these factors make language 1. much more difficult to study, as many more elements and aspects must be taken into consideration; 2. much more susceptible to change, as there are so many more elements subject to change and so many more possibilties for changes to occur.

Music, on the other hand, seems to exist in a realm of its own, a highly ritualized realm, filled far more with redundancies than explicit messages. Unlike language, in which original utterances are continually being produced, music tends to repeat the same utterances over and over, in the form of set pieces that have names and in many cases composers. The primary function of language would seem to be communication, in the form of a long series of continually fresh and original utterances. The primary function of music, on the other hand, would seem to be the affirmation of group identity, based in tradition, operating in and through repetition, which pervades the world of music on a great many different levels, from notes to phrases to the continual reiteration of individual songs and repertoires. Language may be seen, in fact, as a force for change, while music seems to operate as a conservative force, continually linking a society to its ancestors and its origins.

When we look at the relationships between musical styles and languages in various parts of the world, we see many instances where a language has changed, but a musical style persists, suggesting that music may well be more highly valued, and consequently more conservative, than language. The African Pygmies seem to have lost their original language, usually speaking the language of their Bantu neighbors. They will often use Western articles of clothing, Western tools, utensils, ornaments, etc. But all the evidence points to their retaining their original musical style more or less unchanged, as it may have been sung tens of thousands of years ago. A similar pattern is evident in a great many cases where social forces have caused certain societies to change a great many aspects of their culture, from language to lifestyle to religion, yet their basic musical style, or at least significant aspects of it, will persist. An obvious case in point is the persistence of African elements in the music (and dance) of so many African Americans today.

It looks very much to me as though musical style (and possibly dance style as well) is in fact by far the most conservative cultural force in human society, persisting both as a "neutral marker," largely unaffected by environmental and social forces surrounding it (with important exceptions, of course), and, what is more, an actively conservative force, working to preserve the most precious and timeworn traditions of the society that treasures it.

We must remember, however, that I am speaking of music in a very general sense, in the sense that Lomax intended when he conceived Cantometrics as a tool for analysing music on the fundamental level of style. Even when, as is so often the case in the modern world, more patently "musical," surface elements, such as notes, scales, tunings, instruments, melody types, etc., change due to certain social forces and economic pressures, the very basic stylistic aspects I've been emphasizing here will, as often as not, persist.

9. Is Music a Neutral Marker?

Or, more accurately: can we regard certain musical styles or characteristics as neutral markers? Before trying to answer that question, let me explain what a neutral marker is and why it's important. For anthropological geneticists, certain types of DNA, such as mitochondrial DNA or Y chromosome DNA, are especially useful because 1. they don't recombine in each new organism, as does most DNA, so the information contained in them remains essentially intact from one generation to the next; 2. the only way they can change is through random mutation, a rare event that can be used as a marker to identify different lines of inheritance, or lineages; 3. they are not, apparently, affected by natural selection, i.e., they are not thought to be associated with any diseases or other inherited traits that could provide an advantage or disadvantage of the sort that would cause certain types to survive while others die out, due to environmental, biological or social conditions. It is due to point 3, above, that the term "neutral marker" has been applied to them. In other words, because these types of DNA seem neutral with respect to any evolutionaly process other than random mutation, they can be used to trace ancestry with considerable accuracy. There are those, by the way, who dispute this, claiming they can indeed, under certain circumstances, be affected by selection, but that controversy need not concern us at the moment.

The question of the changeability of music over time has an interesting history in the field of Ethnomusicology. Initially, many of the pioneers seem to have uncritically assumed that all "primitive" peoples could be regarded as stone-age "relics" or, indeed, "living fossils," whose cultural practices, including music, could automatically be regarded as survivals, if not from the beginning, then certainly the very distant past. Various theories of musical evolution arose that focused on certain kinds of music from such groups, especially melodies that appeared especially "primitive," e.g., one or two note chants, sung solo or in unison, or else "pathogenic" outcries based on "tumbling strains," expressing primal emotional states. It was assumed that music must have begun with something more or less that "primitive," among the ancestors of "primitive" peoples and then developed over time, to add more notes to the scale, more sophisticated types of melodic structure and, ultimately, with the advent of Western Civilization, harmony and counterpoint.

As it became increasingly clear that little or no evidence existed to support such claims, that, indeed, both harmony and counterpoint could be found among so-called "primitive" peoples in many parts of the world, and especially when ethnomusicologists began thinking more critically about the whole process of tradition and change in music generally, there was a strong reaction against such thinking, to the point that the entire field of musical evolution, and much of comparative musicology, became highly suspect. When Alan Lomax presented Cantometrics during the mid-Sixties, he already had two strikes against him, because of the many years of pent-up resentment against an earlier generation that had become so focused on what looked like a hopeless task. I won't go into the history of Cantometrics here, but its rejection was at least in part due to a general movement away from comparative studies of any kind, in favor of so-called "contextual studies," where the primary focus was on the function of music and musicians, both traditional and non-traditional, within an individual, currently active, social group. From this viewpoint, music would not have been regarded as a neutral marker, because it was now assumed that musical practices would "naturally" change as living conditions changed, due to environmental factors, contact with other groups, the innovations of especially creative individuals, etc.

As I see it, this "new" (actually, no longer new) movement in Ethnomusicology, while valid in many respects, represented a serious over-reaction to what had come before, to the extent that the field seems to have lost much of its relevance for anthropology and social science generally. The contributions of ethnomusicologists have been all but completely ignored, not only by anthropologists and geneticists but cognitive scientists interested in the nature of music and its origins. As now seems evident, both the "old school" of comparative musicology and the "new school" of contextual studies suffered from more or less the same problem: unwarranted assumptions, based largely on intuition with little recourse to either evidence or logical inference.

To be continued . . .

Sunday, May 20, 2007

8. On the Archaeology of Music

Whenever I speculate on certain possibilities regarding the "deep history" of music (i.e., its history prior to any of the written records), I am continually being reminded that it's a mistake to extrapolate backward, from the present situation to the past, because we cannot "go back to the past" to either observe musical performances or listen to how they sounded -- and since we are unable to do this, then we have no way of testing any theory regarding the nature of music in the past, especially the remote past. But isn't the same thing true of every aspect of history? Don't all historians and archaeologists base their theories on what is before them in the present, whether they are working in a library or a museum collection, or out in the field, examining sites, digging for fossils, etc.? It seems to me that I am doing the same thing they do, i.e., drawing inferences based on the methodological study of what I've been presented with in the present, and my understanding of its context, based on years of additional study and research, and extrapolating backward in a meaningful manner, based on those inferences.

How are inferences based on music heard on a recording different from inferences based on a fossil studied in a museum collection? It will be argued that the fossil is "old," a long dead relic of the past, while the music is "new," freshly recorded by a group of living, breathing humans. But how do we know the fossil is old? As far as the ordinary museum goer is concerned, it could be the skull of someone who died last week. An expert who examines the skull knows what to look for. The expert can make inferences about that object based on a lifetime of knowledge and experience. Similarly, how do we know the music is "new"? The composer is still living, you say, so it couldn't be that old, could it? Well, maybe not the specific composition, but most original works are rooted in a tradition, and what interests me is not the indvidual work but the tradition it represents. And there is no reason to assume that tradition is brand new simply because the composition and the performers are our contemporaries. Again, an expert can analyze that composition and that performance and make inferences regarding the age of the tradition behind it, which might, in fact, go back thousands or even tens of thousand of years. As I see it, therefore, there is really no fundamental difference between an archaeological site, a fossil, an ancient papyrus, and a brand new recording of a piece of music. None of the above presents itself to us directly and without mediation as a relic of the past. It can only be treated as such after the evidence is carefully examined and meaningful inferences are drawn.

I have also been told it is a mistake to make assumptions about the deep history of indigenous peoples, that we have no way of knowing whether or not they are still living in the same manner as their ancestors and that there is no reason to assume they haven't changed over time any less than we have. It was, indeed, once common, even among so-called "experts," to make all sorts of assumptions about "stone age," "primitive" and "backward" people, whose lifestyles were uncritically accepted as windows into the world of "early man." But there is a hyge difference between making assumptions and drawing inferences. While I completely agree that it is naive and even sometimes harmful to make assumptions about any group, it is, on the other hand, part of our responsibility as scientists and/or scholars, to collect evidence, and draw reasonable inferences based on that evidence. Arguments regarding the nature of any musical style and its meaning, historical or otherwise, should never be based on assumptions of any kind, on that I completely agree. If I treat certain musical styles as survivals from the past, it is NOT because I am making any sort of assumption about any of the peoples currently practicing such styles, who may or may not be currently living in a manner that reflects any other aspects of the way their distant ancestors lived. But I DO reserve the right to claim that the musical evidence might provide us with some clues as to the nature and history of their culture generally. Not based on assumptions, but evidence, and inferences based on that evidence.

Another objection I sometimes hear is that by suggesting that their musical practices could be survivals from the distant past, I am reducing certain indigenous groups to the status of "living fossils" and thereby "essentializing" them. And, indeed, in the past, it was not uncommon to think of such peoples in exactly that way. Our thinking has now changed, however, as has our vocabulary. We no longer use terms like "primitive" or "stone age peoples," nor do we anymore invoke "living fossils" (unless we are arguing with someone like me!). All people living today are fully alive. They are our contemporaries. No one is a "living fossil" for sure and it would be an insult to claim that anyone was. We do know, however, that certain peoples, and certain individuals, are carriers of certain traditions, traditions that could well go back thousands or tens of thousands of years, traditions valued quite highly by their own group, if not all educated, open minded people everywhere. The term sometimes used is: "Living Treasures."

Saturday, May 19, 2007

7. Some Examples

Here's a brief example of Ju'hoansi (!Kung) Bushmen music, in the form of an mp3 file (from the CD "Chants de Bushmen Ju'hoansi," recorded by Emannuelle Olivier): The Eland -- Girl's Initiation

For comparison, here is an example of the music of the Aka (Biaka) Pygmies of the Central African Republic (from the CD set "Musical Anthology of the Aka Pygmies," recorded by Simha Arom): Divining Music

There are several things to be said about these recordings, which are, by the way, only brief, digitally compressed excerpts (both CDs are available online, at and elsewhere). First of all, in both Bushmen and Pygmy traditions vocal music predominates, often with the accompaniment of handclaps only, as in the first example. While certain Pygmies have become excellent drummers, as is evident from the second example, drums are not considered a part of their native traditions, but were borrowed from neighboring Bantu peoples.

In both examples we can hear many of the striking points of similarity between the two traditions: the use of yodel; the interlocking of voices, to produce an intricate counterpoint; a frequent tendency for one part to be completed by another part, with the effect of a melody tossed back and forth between two or more voices, a practice similar to what, in Medieval Europe, was called "hocket" (or "hiccup"); the extraordinarily well matched and fluent blending of the voices; intricate, precisely executed, polyrhythms; the predominance of meaningless vocables, usually open vowels, such as "oh" or "ah"; highly repetitive, but also varied, melodic structures, based on short motives (but with an underlying melodic phrase as an implied, but often unheard basis); frequent wide melodic leaps; almost complete lack of embellishment; a continuous flow of interwoven sound with no pauses. Most of the above characteristics are part of the Cantometric coding system, by the way, a methodology I'll have more to say about in future posts.

Here are two more clips, from two other sources, first a !Kung Bushmen Giraffe Medicine Song, from Gilbert Rouget's LP, "Bushmen Music and Pygmy Music." Next, an Elephant Hunting Song from the Mbuti Pygmies, located in a completely different part of Africa from either the Bushmen or the Aka Pygmies -- the Ituri Forest, in the northeastern region of the Republic of Congo (formerly Zaire), as recorded by the noted anthropologist Colin Turnbull (now available as "Music of the Ituri Pygmies").

The following example, from a completely different part of the world, but along the theoretical "Out of Africa" migration path, will give you a sense of what is at stake. Does this style of vocalization, from the Bosavi region of Papua New Guinea, a recording of a Men's Work Group, represent a survival of the same Pygmy/Bushmen tradition, an echo from 40,000 to 60,000 years ago, or are the striking resemblances simply a coincidence? (From the CD "Bosavi Rainforest Music of Papua New Guinea," recorded by Steven Feld.)

Here's another one to think about, a beautifully yodeled Women's Song from Guadalcanal, in the Solomon Islands (from the CD set "Voices of the World," recorded and edited by Gilles Leothaud, Bernard Lortat-Jacob and Hugo Zemp).

Thursday, May 17, 2007

6. Year One Part Two

OK, where was I? There were for many years all sorts of speculations regarding the possible significance of Pygmy/Bushmen (P/B) style, which due to its very distinctive nature and unusual distribution, among so many indigenous peoples in so many very different parts of the world, looked very much like a survival of truly ancient or even archaic provenance. However, it was only recently, with the advent of the "Out of Africa" theory, based on new and very exciting types of genetic research, that the real significance of this tradition became evident. Because over and over again it was the Pygmies and Bushmen of Africa whose DNA was being referenced as representative of some of the oldest populations in the world. In the words of a recent (2000) study by Yu-Sheng Chen et al., "these data showed that the Biaka Pygmies have one of the most ancient RFLP sublineages observed in African mtDNA and, thus, that they could represent one of the oldest human populations. In addition, the Kung [Bushmen] exhibited a set of related haplotypes that were positioned closest to the root of the human mtDNA phylogeny, suggesting that they, too, represent one of the most ancient African populations." This is only one of several such studies, pointing over and over again to certain Pygmy and Bushmen groups whose DNA is suggestive of great age.

Interestingly, the Bushmen group closest to the human mitochondrial root was the group Rouget initially studied, the !Kung, who clearly exhibit all the characteristics of P/B style, from interlock to yodel. Another Bushmen group, the Khwe, also studied by Chen, had mtDNA characteristics that were different, closer, in fact, to the African mainstream: "Comparison
of Kung and Khwe CR sequences with those from other African populations confirmed the genetic association of the Kung with other Khoisan-speaking peoples, whereas the Khwe were more closely linked to non–Khoisan speaking
(Bantu) populations." The Khwe would seem to be closer to the African mainstream musically as well. I have not, thus far, been able to identify any of the more distinctive characteristics of P/B style in any of the recordings of their music I have heard to date.

For Chen et al., therefore, just as for Rouget and Lomax, the genetic research suggested that the Biaka Pygmies and !Kung Bushmen stemmed from "a common root." The geneticists went on to estimate that the ancestors of the Biaka Pygmies diverged from the hypothetical founder population between 76,200 and 102,000 years ago, with a divergence time for the Kung Bushmen between 41,000 and 54,100 years ago.

To understand the full, and indeed rather staggering significance of the above, let us once again focus on the musical evidence. We have two populations consisting of nomadic hunter-gatherers with the simplest of material cultures, no permanent residence, no iron or steel tools (until very recently), without domesticated animals, moving about on foot, and located in what amounts to three very distant parts of the African continent, the Pygmies in the forests of both West and Central Africa, the !Kung Bushmen in the desert of southern Africa. Yet they have musical traditions that, for most of the musicologists who have studied them, are so close as to be almost indistinguishable. Since the genetic evidence so strongly suggests that both groups stem from the same root, it would make a great deal of sense to conclude that the similarity of musical practice must stem from the time when their ancestors were part of the same population, a history that may go back at least 76,000 but possibly 102,000 years, according to Chen et al.

So what could the ancestors of the Pygmies and Bushmen have been singing 76,000 to 102,000 years ago? Would it have been something roughly similar, but perhaps more "primitive," maybe less intricately contrapuntal or not contrapuntal at all -- or perhaps not as well organized, less precise rhythmically, more loosely structured? Astonishingly enough, if the Pygmies and Bushmen of today indeed sing in a manner that even the experts find it difficult to distinguish, then it's hard to imagine how their common ancestors of 76,000 years ago could have been vocalizing much differently -- in any respect. Why do I feel so sure of this? Well, let's assume that the ancestral population had a different, or at least somewhat different, musical tradition at the time when the first group of Biaka Pygmy ancestors broke away from the "founder population," 76,000 to 102,000 years ago. Once the two groups had separated, then, in order for them to be so musically similar today, their music would have had to evolve in more or less exactly the same way despite the fact that they were no longer in contact with one another. And for the life of me I can think of no "sufficient reason" (to quote Liebnitz, whose "principle of sufficient reason" is one of the backbones of science) to explain such parallel evolution, or what evolutionists call "convergence." The only explanation that makes sense to ME is that the ancestral group must have been making music in a manner that would also be indistinguishable from "modern" musical practice, among their descendents, to the same experts, if they had had an opportunity to hear it.

We can only conclude that this particular musical tradition must have been passed down from generation to generation over a period of at least 76,000 years (assuming the genetic estimates are correct) essentially unchanged -- a conclusion that, if corroborated, would totally transform our notion of cultural evolution and the role of tradition in its history.

By the way, there are other, closely related, aspects of culture that do NOT seem to have been passed down in the same way. For example, today's Bushmen have a very strong tradition of what has been called "shamanism," where many if not most of the men participate in certain very long and arduous rituals where they can go into trance and then operate as healers. The Pygmies, on the other hand, do not have shamans at all -- and while they do have initiation rituals, these are thought to have been due to the influence of the Bantu groups with whom they have developed symbiotic relationships. Therefore, unlike the situation with music, we are unable to conclude that the original "founding" group already had either shamanism or initiation rituals -- in fact such practices may well have developed after the two groups became separated.

Wednesday, May 16, 2007

5. Music in Year One

So, what WAS music like in the year 000001, you ask? Astonishingly, there is more than enough evidence for us to speculate meaningfully on the nature of humankind's first musical style. Linguists would love to reconstruct the very first -- or "Ur" -- language, based on what is known about the nature of all the various language families in existence today. There's not much hope that any such effort could be successful, as there are very few (or perhaps too many) clues to work with and the whole process of reconstruction would have to be based on a long series of untestable assumptions and speculations. However, music would seem to operate in a very different manner than verbal language and as a result there may be no need to reconstruct humankind's earliest music, because it is still being performed today -- we can simply listen to it. But how can that be? If languages have changed so much over time, wouldn't musical styles also have changed? One would think so, certainly. But the evidence would seem to tell a different story.

What is that evidence? For the details, you will need to read my essay, "Echoes of Our Forgotten Ancestors" (I'll send you a copy if you request it), in which mountains of evidence are considered. For now, I'll simply say that there is one particular type of music, a very beautiful, distinctive and in every way remarkable style of vocalizing found among most Pygmy groups, and also certain groups of Bushmen, in Africa, that, for a variety of reasons, would seem the best candidate by far. For one thing, the two musical traditions are so close that even experienced musicologists with an intimate knowledge of African music have difficulty telling them apart. Yet the Pygmies are scattered over widely separated regions of tropical forest in various parts of central Africa while the home of the Bushmen is in a completely different area, the Kalahari Desert of southern Africa.
In 1956, Gilbert Rouget, then director of the Ethnomusicology program at the Musee de l'Homme, in Paris, wrote: "If, as it is traditional to do, one should consider the Pygmies and the Bushmen as belonging to two races entirely distinct, how can one explain the troubling relationship between their music and their dances? It cannot be a phenomenon of convergence, the resemblances constituting a system too complex and too coherent to allow for an explanation of this order. A reciprocal influence is also to be rejected, being given the distance as much geographic as climatic which separates the ones from the others. Is it necessary to believe, then, that the Pygmies and Bushmen are of common stock, and that their dance and music represent the remainder of a common cultural heritage?"

Alan Lomax, in the same year, also wrote about this remarkable musical tradition, calling it "Pygmoid Style," suggesting that it had a special significance that might well go beyond Africa. And when, some years later, I was working with Lomax, we often played some of the remarkable recordings of Rouget, Colin Turnbull and others, and discussed this music and its possible significance. One of the things we kept noticing was the similarities we found between Pygmy and Bushmen vocalizing and other vocal traditions among certain indigenous peoples in widely separated parts of the world. Lomax suspected that this was a truly ancient, or even archaic type of music making and I agreed. At the time, however, it was not yet possible to speculate as to how old it might actually be, or whether there were other styles that could be older. (to be continued . . .)

Tuesday, May 15, 2007

4. Alan Lomax and Cantometrics

I don't want to get too far into this blog without making it clear that the work I'm doing now has its source in the insights and achievements of certain predecessors, notably some of the early pioneers of Comparative Musicology, such as, for example, E. T. von Hornbostel, Curt Sachs, Jaap Kunst -- also Gilbert Rouget and Walter Wiora -- but primarily my mentor and teacher Alan Lomax. So that there should be no misunderstanding regarding Lomax's importance, or the nature of the Cantometric system he pioneered, I have asked the editor of The World of Music to include the following statement in the next issue:

Since certain aspects of the treatment of Cantometrics in my article, “Echoes of Our Forgotten Ancestors,” could be misleading, especially for younger readers not familiar with the history of that project, I would like to take this opportunity to clarify. As I indicated in the article, Cantometrics was from the start Alan Lomax’s project, based on ideas stemming from years of insight and research on his part. I was hired in the summer of 1961 to work as his musicological assistant, and continued to work under his direction as Research Associate from early in 1963 to the end of 1966. While we collaborated on the creation of the Cantometric system and a portion of the research based on it, Lomax was the senior and I the junior, member of that team. While I would describe myself as an active participant who contributed significantly to the project, the overall tenor of Cantometrics, the fundamental nature of the system, its purpose, and goals, were determined by Lomax, who continued work on the project until 1995.

Several others who also participated in a significant manner to this effort should be recognized. The noted anthropologist, Conrad Arensberg, served as Co-Principal Investigator from 1963 on. Among others who contributed importantly to various aspects of the research during the period of my participation were: statistician Norman Berkowitz; musicologist and composer Roswell Rudd; anthropologists Edwin Erickson, Barbara Ayres and Monica Vizedom; linguists Edith Trager Johnson and Norman Markel, and dance notation specialists Irmgard Bartenieff and Forrestine Paulay, who collaborated with Lomax on the creation of the Choreometrics system.

While I credited Alan Lomax as a major influence and included several references to him and his writings throughout my essay, the exact nature of his influence was not always made clear, and as a result the reader may have been left with the impression that certain key ideas originated with me. Alan’s interest in comparative studies probably began very early, with his involvement in African American music, both in itself, and in relation to its African roots, a subject that interested him deeply throughout his career. Already by 1956, he presented a preliminary classification of world folk song style, proposing some of the same style areas that were to be so convincingly supported by Cantometric testing years later. Here we find a description of a musical family labeled “Pygmoid,” in which he already puts his finger on the most salient characteristics of the same “Pygmy/Bushmen style” that plays so important a role in “Echoes” (Lomax 1956:49). In the same paragraph, moreover, Lomax refers to groups “possibly also in India, Formosa, etc.” that share some of the same, very distinctive, characteristics. Clearly he already had an awareness of the potential significance of P/B style far beyond the confines of the African continent, an awareness that deepened and broadened in subsequent years as more and more instances came to light. My understanding of, and interest in, P/B style, the very special nature of its distribution worldwide, and the possible historical implications of that distribution, was based on what I learned from Lomax during the years I worked with him. Many other observations appearing in “Echoes” regarding the patterned distribution of style traits and their possible meaning similarly stem from this period of “tutelage” with Lomax, along with concurrent research carried out under his supervision.

In 1959, Lomax produced a much more extensive essay, containing a wealth of important and indeed prophetic, insights, paving the way for the Cantometrics project to come (Lomax1959). While I collaborated with him on the creation of the Cantometric system per se, and much (though certainly not all) of the research that followed over the next few years, the foundations of Cantometrics had thus already been established by Lomax at least two years prior to my involvement -- and considerably more was accomplished by him and his team after I left.

Another possible source of misunderstanding centers on two statements appearing toward the end of the “Prologue”: first, the assertion that “I treat Cantometrics as an essentially heuristic methodology”; and second, the subsequent declaration that, in my view, it is “… not capable in itself of producing totally objective results of the sort required by strict scientific method.” I hope this did not give the impression that I do not regard Cantometrics as a legitimate scientific methodology, because that was certainly not my intention. On the contrary, I consider it a particularly powerful scientific tool. The purpose of these statements was to address, as succinctly as possible, a prevailing tendency on the part of a great many to demand too much of this methodology, rejecting it out of hand for not meeting what are in fact unreasonable expectations. This attitude may have been encouraged by Lomax himself, who sometimes gave the impression that certain of his results had been proven beyond the need for additional investigation.

Heuristics is an important aspect of scientific research, so to describe a methodology as “essentially heuristic” is by no means to declare it “unscientific.” More fundamentally, research in the social sciences generally can never approach the objectivity and precision characteristic of “hard” sciences, such as physics, chemistry, biology, etc. For what it sets out to accomplish, Cantometrics would seem to work remarkably well. It is, in fact, not easy to imagine how encoding over so diverse a range of parameters and styles could have been achieved in a significantly more objective manner.[i] Nevertheless, because of certain inherent limitations necessarily built in to any methodology of this sort -- sampling issues, subjectivity, the possibility of unconscious rater bias, certain ambiguities that will inevitably arise due to the difficulty of anticipating all possibilities, etc. -- cantometric findings should, wherever possible, be supplemented with additional research and testing. There is no hard and fast rule here. In some cases, as with, for example, the factor analysis reported in Lomax 1980, the cantometrically derived categories are so clearly correlated with well known and widely accepted cultural and geographic boundaries, that the result is, for me at least, especially convincing. In other cases, however, where the meaning of the results may not be so clear, additional research may be necessary before any hard and fast conclusions can be reached. What makes Cantometrics stand out from so many other methodologies in the social sciences, is the fact that it lends itself so well to re-examination of this sort, being, at least in principle, fully open to independent testing and review.

With such considerations in mind, I am pleased to add that, thanks to the efforts of Alan’s daughter, Dr. Anna Lomax Wood, the staff of the Association of Cultural Equity (ACE), which she now heads, and especially computer gurus John Tam and Michael Del Rio, the Cantometrics dataset has been updated, to the point that it is now compatible with currently available database software. Dr. Wood’s intention to ultimately make available the database, along with the original teaching materials, will hopefully encourage renewed interest in this extremely useful, but too often misunderstood and undervalued, methodology. Thanks to the generous support of ACE, I have been able, over the last few months, to access the revived database as part of a new research project, in collaboration with Dr. Sarah Tishkoff of the University of Maryland, one of the leading figures in Genetic Anthropology, and her associate, Dr. Floyd Reed. It is my hope that this and similar efforts in future will prove effective in vindicating and carrying forward Lomax’s extraordinarily valuable legacy.


Grauer, Victor and Fred McCormick
2005 “Cantometrics: Song and Social Structure – A Response.” Music Traditions 159., last accessed 27 January, 2005.

Lomax, Alan
1956 “Folk Song Style: Notes on a Systematic Approach to the Study of Folk Song.” International Folk Music Journal 48-50.
1959 “Folk Song Style.” The American Anthropologist 61-6:927-953.
1980 “Factors of Musical Style,” in Stanley Diamond, ed., Theory and Practice: Essays Presented to Gene Weltfish, 29-58. The Hague: Mouton.

[i] The nature of the scientific problems entailed in any such project are treated in some detail in Grauer and McCormick 2005, wherein I analyze some of the “hidden pitfalls inherent in almost all attempts to be objective when evaluating style.”

Monday, May 14, 2007

3. The neglect of music by anthropologists

I find it both puzzling and frustrating, given what is known about the various families of traditional music worldwide, to read, in the field of anthropology, book after book, report after report, where theories about the history of languages, and their grouping into various families, are continually being referenced, especially in the literature on the origins and early migrations of "modern" humans, and yet what is known about the musical picture is all but ignored. In a sense this is a "good thing" for me, since I now find myself apparently without any competition at all, from ethnomusicologists or anyone else. But it's a huge problem for me as well, because the notion that the study of current or recent musical practices among traditional societies cannot be applied to the distant past is so ingrained into the thinking of so many.

Even in a book explicitly devoted to speculations about the nature and origins of humankind's earliest music, The Singing Neanderthals, archaeologist Steven Mithin makes a point of ignoring such practices (thus conveniently relieving himself of the need to do much of any reading in the ethnomusicological literature) on precisely those grounds. He bolsters this position by arguing that music changes at a much more rapid rate than language: "The speed at which music can evolve is readily apparent from the development since the early 1970s of rap, part of a more general subculture of hip hop, which includes break-dancing, 'cutting and scratching' records and graffiti." (p,. 21). This monumentally naive and amateurish observation, offered as evidence, would probably, and very sadly, be seconded by many professional ethnomusicologists, who now seem so obsessed with the notion of change that they have forgotten what the word "tradition" means.

First of all, it makes no sense to lump the practices of indigenous tribal peoples with current developments in Western society, where information can be transmitted instantaneously around the world to millions if not billions of people at a time. Second, it completely ignores all the many aspects of rap and hip-hop culture that can be seen as perpetuating very fundamental aspects of subsaharan African culture. It's even been argued that such practices constitute a revival of African traditions, as a reaction to the more Europeanized types of music that constitute the pop mainstream. As I argue in my essay, we have very good reason to believe that music, or at least certain musical traditions, are far more conservative than just about any other aspect of culture, certainly more than language. One of many reasons why it's important for anthropologists to start paying attention.

2. The Importance of Music

Is music truly important? Or is it just another aspect of culture, like, say, pottery making, bead stringing, cave painting, hunting, fishing, farming, etc.? To answer my own question: YES, music really is all that important. But you'd never know it from the work of the great majority of ethnomusicologists over the last, say, 25 years. Which may have something to do with their neglect by anthropology -- or every other field, for that matter -- over roughly the same period. I could write a book on how the field for which I had so much enthusiasm in my youth has become so ingrown, so parochial and, very frankly (and with some important exceptions) so irrelevant -- but for now just want to emphasize my main point here, which I suppose could be phrased thusly: music is too important to be entrusted to (most) ethnomusicologists.

To understand why, let's compare music with language. And for the sake of argument, let's assume that linguistics hasn't yet been invented, so anthropologists have no methodology for dealing with the languages of the peoples they study. Thus, every language encountered would have to be treated on an ad hoc basis, with little or no possibility of comparing it with any other language, no way of determining what language family it might belong to, nor any generally agreed upon method for assessing what elements are essential, or what its formal structures, including its "deep structures" might be. Our hypothetical Anthropologists would take note of certain more or less obvious facts: language is found everywhere, among all people; language is an important aspect of day to day life, reflecting many significant aspects of the value system of any given society; language is an important factor in the "construction of identity" and plays a significant role in the "negotiation of gender." The Anthropologists might therefore do things like interview informants to elicit their thoughts on the importance of language in their life, get them to reminisce about their first encounters with language, and through this and similar methods attempt to place language and the use of language in context, arguing that we cannot really understand any language unless we understand the context in which that language takes place. They might also agree that one cannot separate language from the rest of culture, that to do this would be to "reify" it, thus ruling out ahead of time any possible attempt to develop a science of linguistics in the future.What they might also agree on is that it would be a good idea to actually learn the language of whatever group one is studying, so maybe they'd find a "master speaker" to apprentice themselves to and, after a year or so, learn (more or less) to speak that language. They'd then return to their college or university, get a teaching job (hopefully) and teach that language to their students.

While clearly no Anthropologist or Linguist works this way, I think the above is a more or less fair description of the current state of Ethnomusicology. In other words, ethno without the musicology. Or, more accurately, ethno, with some sort of ad hoc musicology, that works differently in each case. This would be roughly equivalent to a situation where, for each language studied, each individual researcher would need to define for him/her-self a unique version of phonology, morphology, syntagmatics, etc. or simply ignore such problems altogether. A situation, also, where little or no attention was paid to whether or not a particular musical practice in a particular place was part of some "musical family," by analogy with the language families that are now (however controversial) so basic a tool of both historical linguistics and anthropology.

There is a history here, that I won't go into much, except to say that once upon a time there was a field known as "Comparative Musicology" that did concern itself with such issues, but is now regarded as hopelessly outdated, ethnocentric, reductive and whatever other fashionably "post-modern" dismissals you can call to mind. Just think of the consequences for anthropology if comparative linguistics were deemed ethnocentric and reductive, and it was no longer considered morally acceptable to place languages into large-scale categories, distinguishing between, say, Bantu and Nilotic or Austronesian and Papuan, Indo-European and Altaic, etc.

What I'm getting at here is that in my view music is NOT, as so many of today's ethnomusicologists seem to think, just one of many cultural practices worth studying, such as fishing, hunting, tool making, basket weaving, pottery making, etc., but, like language, a far more fundamental, if not elemental, cultural FORCE. That shouldn't be so difficult to see, given the fact that, in our own society, the music industry has for many years been among the most profitable on the planet. You won't make anywhere near as much money investing in the pottery, basket, cloth, meat, or even automobile, industry. But even in the most remote tribal societies, music usually would seem to play a necessary role in a host of absolutely central areas, such as religion, ritual, courtship, work, war, the establishment of group identity, sense of community, etc.

Also, music and language would seem to have a great deal in common and may well have developed in tandem. Music operates in many ways like a language, especially if we put the issue of semantics aside and concentrate on structural issues, such as musical "phonology" and syntax. As linguist Roman Jakobson once pointed out, musical notes are very close in function to phonemes -- Charles Seeger called them tonemes and rhythmemes.

Anyone who's ever taken one of those ubiquitous "World Music" classes comes away with some very rough, but real, sense that musical families exist. You may be surprised to learn, however, that musical families are not recognized among todays ethnomusicologists and any attempt to think in such terms is actively discouraged as overly speculative and, again, reductive. In my view, musical families most certainly do exist, and might well be of great use to those ethnographers, archaeologists, geneticists, cognitive scientists, etc., who pay so much attention to linguistics but show so little interest in musicology, ethno- or otherwise. Actually cognitive scientists have, of late, shown a great deal of interest in music and have speculated rather widely (and wildly) regarding the origins of music and its relation to language. In so doing, they have rarely, if ever, made use of ethnomusicological research -- especially since the ethnomusicologists themselves are assuring them that such research would be of no use to them, since, according to the current dogma, we cannot, must not, read any aspect of the distant past into the current musical activities of any peoples alive today. For historical linguists, and the anthropologists who draw on their work, such a practice is routinely accepted. Archaeologists draw on so-called "ethnographic analogy" probably far more often than they'd like to admit. It's a controversial, but nonetheless generally accepted aspect of this field. But for (most) ethnomusicologists it is forbidden territory.