43. Battle of the Maps

On page 13 of the essay I've been discussing, Reconstructing African music history, Roger Blench presents, under the heading "Historical Interpretation," his theory that "the polyphonic wind ensemble developed in the early period of the genesis of Nilo-Saharan in the Ethio-Sudan borderlands and then spread westwards to become part of the cultural repertoire of Niger-Congo in the early period of its expansion." Blench is referring to two well-known African language families, Nilo-Saharan, the older, "appearing in the terminal Pleistocene, perhaps as much as 18,000 BP," and Niger-Congo, the more recent, "assigned to the early Holocene," roughly 11,000 years ago. Niger-Congo is the larger unit within which all the various Bantu languages, which now dominate most of subsaharan Africa, have been subsumed.

I find it interesting that Blench, noted primarily as a linguist, would find a linguistic explanation for the distribution of a musical style family, under the apparent assumption that linguistic and musical families travel together as a unit. (If you notice a trace of sarcasm in the previous sentence, I must immediately clarify, and also apologize, because as a matter of fact Blench's essay pleases me enormously, and I have the greatest respect for what he has both attempted and accomplished therein. He is doing the sort of thing I'm trying to do myself, i.e., seeking out large-scale patterns and attempting to account for them as meaningfully as possible -- and in so doing, challenging some of the most deeply entrenched dogmas of both ethnomusicology and anthropology -- and probably linguistics as well. While I disagree with him over certain points, I admire him enormously for what he has attempted -- and am also very much in his debt, for what he has revealed about this important musical practice, the details of its performance, its social significance, and its distribution.)

To better understand what's at stake, let's take a look at the map Blench presents under the heading, "African Distribution of Polyphonic Wind Ensembles."

He compares this map with the "Outline Map of African Language Families" presented on p. 14 of his essay:

The Nilotic-Saharan language family he associates with the origin of the wind ensembles is in the upper part of the map, represented in green. I'm sorry, but I don't see much of a resemblance between the green portions of this map and the distribution of the wind ensembles as presented on his previous map. As should now be clear, I (respectfully) disagree with Blench. I believe the wind ensembles to be intimately associated with the hocketed vocal styles I've been discussing as possible candidates for humankind's earliest music, practices that would most likely have emerged, according to all the evidence I've been presenting here, among the same ancestral population the geneticists have associated with the origins of both the Pygmies and the Bushmen. I'm grateful to Blench for his distribution map, which looks reasonably accurate, but puzzled as to why he doesn't see the obvious connection with the distribution of both the Pygmies and Bushmen in central and southern Africa, respectively. To make my point rather dramatically, I'd like to present his wind ensemble map side by side with a map I've put together showing the distribution of both these groups. My map is something of a Frankenstein monster, complete with very visible seams, but nevertheless clear, I hope. The ivory colored rectangle is taken from the following Wikipedia article, on Pygmies. If you scroll down about halfway, you'll see a partial map of Africa, entitled "Distribution of Pygmies According to Cavalli-Sforza." This is Luca Cavalli-Sforza, one of the pioneers of genetic anthropology, who has devoted a great deal of attention to the various Pygmy groups over many years. I've superimposed his map over a more complete map of the entire continent, to which I've added two words: "Tellem," to indicate the location of a Pygmy group thought to have once been based in southern Mali, though no longer present in that area today; and "Bushmen," to indicate the general location of the Bushmen, now based in the Kalahari desert area of Botswana and Namibia, but thought to have formerly inhabited the entirety of southern Africa:






To get a clearer view of each map, just click on it and an expanded version will appear. As I see it, this speaks for itself, but you know me -- I've be having a lot more to say, as usual.

42. Cave of the Winds

All the evidence I've been considering in the last several posts strongly suggests a direct line of "development," from pre-human duetting/ chorusing (style X on my phylogenetic tree), to Shouted Hocket (style family A1), to Interlocked Hocket (A2), to the hocketed wind ensemble tradition(s) discussed by Blench. In other words, the wind ensembles can be seen as either direct outgrowths from the hocketed vocalizing of A2 or, very possibly, as a practice that arose in parallel with the vocal style, something roughly similar to what we heard in the audio clips from the Babenzele and Huli, where a single performer hockets between his voice and his pipe; or the examples from the Ouldeme, in Africa, and the Russian panpipers studied by Velitchkina, where the same group of performers interweave voices and piping to produce an intricately blended texture.

If we assume, therefore, along with Rouget, that the Bushmen and Pygmy vocal practices stem from "a common root," then it's not difficult to see how the hocketed instrumental tradition could have originated from that same ancestral culture, some time between 72,000 and over 100,000 years ago (according to the genetic estimates) and spread along with the spread of that culture to certain places in Africa where, in various forms, it survives today. This is somewhat different from the view expressed by Blench, who sees no structural relationship between P/B style and the one-or-two-note-per-instrument hocketing of the wind ensembles -- an opinion reinforced by Simha Arom, who, as I mentioned earlier, has stated that he's never heard any hocketing among any of the Pygmy groups he's studied.

As I see it, Blench's denial of that crucial connection may stem from a failure on his part to distinguish between three different types of P/B vocalizing, as presented in my Phylogenetic Map: A1, "Shouted Hocket," A2, "Interlocked Hocket," and A3, "Contrapuntal Interlock." The latter, characterized by relatively long, fluid lines, interweaving "contrapuntally" does, indeed, sound rather different from the more choppy one or two note interchanges of the wind groups. But the other styles, A1 and A2, often simpler, possibly less interesting musically, and thus perhaps overlooked by Blench (and also Arom), do most definitely employ a very straightforward and clear form of hocket, in a manner quite similar to what is found in the wind ensembles.

The details of such performances are often obscured in recordings, since Pygmy and Bushmen polyphony is notoriously difficult to sort out by ear. Fortunately, I've been able to find a very clearly notated transcription in an excellent book by Michelle Kisliuk, an ethnomusicologist who spent enough time among the Aka to learn their songs (and dances) well enough to participate. (See Kisliuk, Seize the Dance, Oxford Univ. Press, 1998, p. 42.)

Kisliuk's notation, based not simply on listening and/or analysis, but her intimate knowledge of what happens in every detail of this music, represents a kind of "summary" of certain basic elements as they might be juxtaposed in a given performance. Her intent is to give the reader a general sense of what can happen at any point in the three "measure" song, as it is continually repeated and varied, and is thus more effective in reflecting the basic structure than a more literal transcription would be.

It's hard to understand, given such a transcription, how someone so well acquainted with Aka music as Arom could deny any knowledge of hocket in their music. The very striking similarities with the wind ensemble traditions he's documented are immediately apparent from the layout of the lowest three parts, each of which is limited to two notes which do indeed interlock with one another in classic hocket style. There are other examples of very similar sorts of vocal hocket in Kisliuk's book as well, e.g. on pp. 93, 137 and 138. She explicitly refers to this practice by name on p. 136. And, of course, many other very experienced musicologists have noted the importance of the hocket principle in Pygmy and Bushmen music. While, indeed, certain details are often obscured in recordings, and many performances do feature sustained melodic passages with fewer gaps, as in what I've called style A3, there is no question in my mind that a very deep structural affinity between P/B style and instrumental hocket exists.

41. Voices of the Winds

In the section of her essay on Russian panpipes (see previous post) entitled "Breathing Techniques" Olga Velitchkina explicitly refers to a very specific danger inherent in pipe playing generally -- hyperventilation:

Uninterrupted blowing from one pipe to the next is very air consuming, because in this way part of the air is wasted during the movement. On the other hand, if each push of air is treated as a separate breathing cycle, it exhausts the player and quickly leads to hyperventilation. For this reason, Russian panpipe players employ two levels of breathing simultaneously, one "deep" level maintaining a normal breathing pace and another level supplying air for each note played on the instrument. . .
The articulatory breathing in panpipe playing is a visible and audible phenomenon. It can be observed in the movements of the upper abdominal muscles, pushing the puffs of air through the rest of the air column. The way this is done holds the 'secret' of not hyperventilating when playing kugikly.

For the Russian performers breathing is especially tricky because they must coordinate their pipe articulations with "vocal sounds," as described in the following section of her essay. One cannot help, while reading this, and recalling the sound of both the Russian and Ouldeme recordings of combined piping and vocalizing (see previous post), but make a mental comparison with the situation we found among the Siberian and Inuit throat-singers, the Bushmen Tcoqma performers and the Maasai warriors, all of whom are skating at the edge of hyperventilation, unconsciousness and trance. Could there be something about this type of hocketed interplay generally, whether vocal or instrumental, that tends to induce hyperventilation and trance? And if so, could this have something to do with the apparently close association between the vocal and instrumental versions of both "Shouted Hocket" and "Hocketed Interlock" we've been discussing?

Did the vocal tradition come first? If it developed directly out of the duetting, chorusing and pant-hooting practices of our pre-human forbears (whose air sacs would have forestalled hyperventilation), then it does make sense to assume the first musical notes could have been produced vocally, very possibly originating in yodeled (thus pitched) shouts, or hoots. On the other hand, it could have been the invention of simple one-note whistles, pipes or trumpets that produced discrete pitches for the first time -- and inspired the yodels. Another possibility, perhaps more likely, is that both practices emerged at roughly the same time, each feeding the development of the other.

I'd like at this point to return to Roger Blench, whose doubts regarding the possible relation of his polyphonic wind ensembles to Bushmen vocalizing got me going on this particular track. Blench correctly points out the differences, but at the same time implies a deeper affinity by pointing to a possible "link" to be found somewhere in southwest Ethiopia. There are, indeed, some very interesting traditions of polyphonic vocalization in that region, made all the more intriguing by the archaeological and genetic evidence pointing so strongly to this part of Africa as the possible homeland of the original "Out of Africa" migrants.

Here is an example of a rather intricate form of style family A1, "shouted hocket," from the Adjuran, cattle herding nomads of southwest Ethiopia, sounding very much like a cross between the Bushmen Tcoqma ritual and the panted vocalizing of Maasai warriors (from Ritual Music of Ethiopia -- Smithsonian/Folkways). Indeed, according to the liner notes, the Adjuran "dance themselves into trance" during such performances. Here's another example, but this time in style A2, "hocketed interlock," from another group also from southwest Ethiopia, the Ari, sedentary farmers living in the highlands (from Ari Polyphonies, Ocora). I find this performance especially fascinating as 1. it's so close in style to the preceding "shouted hocket" selection; 2. it so strongly resembles the one or two note per instrument hocketing of the wind ensembles, yet 3. is at the same time so very similar to certain types of Pygmy and Bushmen vocalizing. Here indeed we find a type of vocal interlock that's structurally all but indistinguishable from its instrumental counterparts. And as a matter of fact it's not unusual to find hocketed and/or interlocked compositions that can be played by wind ensembles or sung, interchangeably.

40. Sweep Away the Morning Dew and Sweet the Winds Shall Blow

For Roger Blench, the polyphonic music of San speakers (i.e., Bushmen) "structurally . . . appears quite unlike the wind polyphony recorded from elsewhere in Africa, although it has been argued that it resembles the vocal polyphony of the Pygmies. There are important structural differences between this type of two- and three-part polyphony and the multi-octave ensembles characterised here." He then goes on to suggest that a "link" might be found, nevertheless, in Ethiopia, where both styles are still being practiced.

Blench's remarks regarding structural differences make a great deal of sense -- indeed, the differences seem obvious, since in the vocal tradition each part is in itself melodic, while the instrumental parts often consist of only one note each. However, if a link is possible then doesn't that imply a structural equivalence, if not on the surface, then at some more fundamental level? We run into this sort of problem all the time when comparing one practice to another -- some things seem similar, others different; so are the two "really" different after all, or basically the "same"? In my view, the vocal and instrumental traditions are, if not the same, then nevertheless very closely related. For one thing, each individual pipe, horn or trumpet part in these ensembles is both hocketed and iterative, suggesting a possible connection with the hocketed vocal iterations so characteristic of style A1. Additionally, I think the more complexly interlocked hocketing of A2 may also have served as a model on which the more complex instrumental styles were based, as the end results can sound quite similar, despite the differences in performance technique. But there is also some evidence that the two could have developed in parallel, reinforcing and enriching one another.

It's important to understand that certain Pygmy groups have instrumental traditions similar to the ones mentioned by Blench, although the instruments rarely play on their own, without the participation of voices. Here is an example of hocketing between a Babenzele "hindewhu" pipe and a vocalist -- the same man performs both parts (from Anthology of World Music Series). And here, incidentally, is something very similar, from the Huli, in a very different part of the world -- New Guinea (from Music of Papua New Guinea, Musique du Monde.). Returning to Africa, the Aka Pygmies call such pipes mo-beke (from Musical Anthology of the Aka Pygmies, Ocora). A more complex interchange between pipes and voices is heard in another Babenzele example, "Song After Returning From a Hunt."

Hocketing pipes and voices are combined in other African traditions also, as in this example from the Ouldeme people, in the Mandara Mountains of Cameroon (from Cameroon:Flutes of the Mandara Mountains). I find this recording strikingly similar to the music being played in the following video, made by Olga Velitchkina as part of her study of the movement style of Russian panpipers (you'll need QuickTime installed to view this little movie). Listen carefully for the "hooted" voices in both clips, strongly reminiscent of what we hear in the Pygmy recordings above, as well as some of the shouted hocket examples we've already heard in an earlier post. If my juxtaposition of African, Melanesian and European traditions puzzles you, I wouldn't be surprised. But there is in fact considerable evidence, as presented in my essay, that all the musical practices quoted above share a common root

That's all I have time for today, folks. See you back here soon.

39. Blow Ye Winds Ay-oooooo

When I first came across the essay on African music by Roger Blench (see previous post), I was delighted to learn how much his views corresponded with my own, as for example, in the following passage, from p. 3:

Obviously, no amount of archaeology or other types of
reconstruction are going to recover the music in Africa’s past. However, this is not really the point; the structures of African music in the present can teach us a great deal about the way cultural patterns become distributed over large areas. Indeed one intriguing aspect of broader categories of musical forms is their relative conservatism over extensive regions of the world.

Or this passage, from the following page, which deserves to be quoted at length:

Reconstructing African cultural history through the distribution of material culture or related practices such as musical forms is of course deeply unfashionable. Its real exponents were the German ethnologists of the Kulturkreislehre school . . . and their Swedish successors, notably Lagercrantz. Detailed and painstaking as much of this work was, it made little real impression on scholars from other disciplines, largely because of their lack of an interpretative framework. . . This paper proposes that this tradition is worth reinventing in the light of a much better grounding in the prehistory of Africa and in particular a more comprehensive framework of language distribution to which it can be affixed. It will assume that sometimes diffusion does occur and that this is not necessarily an affront to the dignity of a people who borrow an instrument from their neighbours. . . Polyphonic wind ensembles resemble an underlying pattern [that] surfaces in different forms across the continent but which is sufficiently specific musically to suppose that it is very ancient and is perhaps associated with the language phyla of Africa.

I was excited to learn that Blench, influenced by the work of the distinguished Africanist and musicologist, Klaus Wachsmann, had independently taken a comparative approach so similar in spirit to mine. (I was also pleased to see the remarkable photo of a horn and flute ensemble from Chad on p. 9 -- the one I appropriated from Blench's essay to place at the top of this blog. There's another great photo of a similar ensemble on the next page.) The list he provides of the "main musicological characteristics of such ensembles" is especially valuable, particularly what he has to say in sections d and f:

d) The tuning of such ensembles is almost invariably pentatonic or heptatonic. In most cases, instruments produce a single note. One instrument is assigned to each degree of the scale, even where
the instruments are capable of producing a wide variety of notes, for example, notch-flutes.

f) Each musical part is of approximately equal importance; canon or hocket-like structures are usual.

The observations that follow regarding social context seem also to be both relevant and useful, as is the map he provides on the following page, to which I drew attention in my previous post. I had no serious problems with any of his observations and interpretations until I came to the following passage:

San speakers also have polyphonic music but it is
essentially vocal. Structurally, it appears quite unlike
the wind polyphony recorded from elsewhere in
Africa, although it has been argued that it resembles
the vocal polyphony of the Pygmies. There are important structural differences between this type of two- and three-part polyphony and the multi-octave ensembles characterised here. Nonetheless, the link may be found in southwest Ethiopia. Omotic and Nilo-Saharan speakers in this area retain polyphonic styles reminiscent of the Pygmies as well as more complex styles and it seems likely enough that from this centre of diversity emerged the characteristic one-note-to-a-part wind polyphony.

Blench continues, drawing some very interesting historical inferences from the special qualities he finds in the polyphony of southwest Ethopia, ultimately proposing a diffusionist expansion of wind ensemble polphony from this area, based at least in part on his interpretation of certain linguistic evidence.

I agree that there is a difference between the polyphonic vocalizing of the San (i.e., Bushmen) and Pygmies on the one hand and the "one-note-to-a-part" hocketed polyphony of the wind ensembles. A leading authority on both Pygmy vocalizing and African instrumental polyphony, Simha Arom, is of essentially the same opinion. In fact, Arom has stated that he has never observed any form of hocketing among the Pygmy groups he's studied (personal correspondence). In addition, I find Blench's remarks regarding the possible importance of certain polyphonic traditions in southwest Ethiopia to be remarkably apt, as I too find these traditions significant.

Where I have problems with Blench, and also Arom, is in the way certain (admittedly very real) differences are interpreted, how we are to understand them, and how much weight they should be given when considering certain basic issues pertaining to historical and cultural relevance. There is, as we all know, a long history of "academic disputes" over all sorts of evidence, but I would prefer not to think in such terms here, especially since there is so much in the thinking of both Blench and Arom that appeals to me -- and with which I agree. There is even more along these lines to be dealt with, since two of Arom's former students, Susanne Furniss and Emmanuelle Olivier, have, after very careful studies of both the Aka Pygmies and Ju'hoansi Bushmen, concluded that the fundamental concepts underlying the music of these two groups are "radically opposite." Here again, despite my great respect for both of these very diligent and insightful researchers, I believe there to be an important difference of interpretation that needs to be resolved.

So! What I'd like to do in the following posts is delve more deeply into the relationship between Pygmy/Bushmen "hocketed interlock" and the structure of the wind ensembles discussed by Blench, with an eye to 1. explaining why I believe the latter to be a development from the former; 2. examining the possibilty that both practices may have developed in tandem; 3. offering a historical interpretation somewhat different from that of Blench, though it's by no means my intention to dismiss the very impressive insights behind his thinking, especially his views on the key role played by the southwest Ethiopian groups.

I'll need to wait a bit before fully airing my differences with Furniss and Olivier on this blog (though a refutation has already been published, in the current issue of The World of Music, as part of my reply to Jonathan Stock), since I am currently preparing a detailed paper on their extremely interesting and important, though ultimately misleading, research.

In any case, look for me back here soon, hopefully tomorrow, with some specific examples to listen to -- and ponder.

38. Blow Ye Winds of Morning

Picking up the thread from post 36, I'd like to consider the role of certain very significant instruments, notably certain wind instruments: whistles, ocarinas, horns, trumpets and especially pipes -- including panpipes. There are many ensembles in Africa consisting of such instruments, either all of one kind or, in some cases a mixture -- and typically each instrument plays only one or two notes, interlocking with one another according to principles very similar to those of the vocal style we've been discussing, i.e., "Interlocked Hocket." It looks very much as though the instrumental practices might have developed from the vocal style. It's also possible the two could have developed together.

Linguist/ethnomusicologist Roger Blench discusses such ensembles in a recent essay, available online, entitled Reconstructing African music history. See especially the section entitled "3. Polyphonic Wind Ensembles," beginning on p. 8. On page 12, he presents a very interesting map, indicating the distribution of such ensembles in Africa. What struck me immediately on a first viewing was the correspondence of this distribution with the distribution of many Pygmy groups in the central area, and the Bushmen in the south. While Bushmen do not currently have such ensembles, there is good evidence they did in the past, as Blench notes, referring to their presence among the closely related Khoi (Hottentots) "at the time of Portuguese contact." Also, the smaller area he outlines in the northwest is centered in southern Mali, where a group of Pygmies called the Tellem apparently lived at one time, before relocating to central Africa.

Blench does not hear the resemblances I hear between Pygmy/Bushmen hocket/interlock and the types of instrumental hocket he describes on these pages. In the next few posts I'll be looking into various aspects of this relationship in an attempt to sort out some of the similarities, differences, possibilities and disagreements, drawing on a variety of sources and concepts.

Sorry this post must be so short, but I've been spending most of my time lately going through some of the literature on this very interesting and to my mind extremely important issue. I'll be taking a break on Saturday (I DO have a life, in case you're interested) but hope to be back with much more on Sunday.

37. Is This For Real?

Many people have an odd idea about science, that something isn't really truly scientific unless you can prove it. And if you can't prove it then it's better if you never even tried, because that would be speculation and speculation is dangerous. This is the attitude I've encountered time and again in music departments, and especially among music historians, who, under the impressive sounding but questionable title of "musicologists," were usually the ones who ruled the roost. These people, who spent most of their time in libraries, dredging up musty, dusty tomes or the occasional score, and keeping track mostly of really exciting things like dates, opus numbers, watermarks and handwriting, thought they understood something about science, though my guess is that very few had ever taken a college level science class in their lives.

My original hope was that Ethnomusicolgy was going to free us from the grip of the historians by opening up the study of music to a whole set of new ideas, new ways of listening, new ways of thinking about music and making it, and all sorts of new possibilities for interdisciplinary research into completely fresh and inspiring realms of possibility. Instead what seems to have happened is that almost all Ethnomusicology programs are still embedded in music departments, run all too often by people with the exact same sort of mindset that drove me nuts as a student: spend your time gathering information, catalog it, then look for more, and by no means speculate because speculation is dangerous, it isn't objective and therefore isn't scientific, and what you really need to do is be narrowly objective at all times, the narrower and more constricting the better, so you can prove that you're "right" even though what you've proven is so trivial that no one other than yourself and your thesis adviser gives a damn.

Most professional scientists see things very differently: that science is at heart an inquiry, a disciplined inquiry for sure, an inquiry based on evidence, of course, but first and foremost an inquiry into certain possibilities, meaningful possibilities that are worth exploring and could lead to further, even more fruitful inquiry. And at the heart of this sort of science is: speculation. What the "musicologists" never seem to get is that, as far as science is concerned, speculation, i.e., meaningful, disciplined speculation, based on the objective assessment of significant evidence, is what it's all about.

So, to answer my own question, there is no way of knowing for sure whether or not the various speculations I've been offering here about the meaning of certain patterns of musical style and the origins of music are really, truly "for real." I do think there are aspects of my argument that could be tested and indeed I'm currently working with geneticists from the University of Maryland to put some of my ideas to the test. Certain of these ideas have already, as I see it, been tested and passed the tests, which is why I have been sufficiently emboldened to share my thoughts here.

But, as with some of the most widely discussed scientific theories of our time, such as the Big Bang, cosmic inflation, quantum theory, supersymmetry, string theory, etc., along with various theories currently being debated about the nature of human origins and early migrations, there will probably never be any way of definitively proving that any are either correct or incorrect -- though some may very well fall by the wayside. What I've been doing here is attempting to demonstrate, not that my theories are "right," but that I am on to something worth looking into, that there are certain possibilities, overlooked in the past, that could lead to some real insights into the nature and origins of music -- and culture as well -- opening the door onto some very fruitful paths of exploration.

36. Arriba!

As I argued in my "Echoes" essay, there are many reasons for tracing the musical traditions of the original "Out of Africa" migrants back to Pygmy/Bushmen style ("haplogroups" A1 through A4 on my Phylogenetic Tree) -- and there are in fact a great many instances of strikingly similar practices, both vocal and instrumental, to be found among indigenous peoples living in enclaves along more or less the same route they are thought to have taken, anywhere from 50,000 to 100,000 years ago. Moreover, as I have been arguing in the last several posts, it would seem that the P/B variant I've labeled A1, or "Shouted Hocket," is widely enough distributed among such groups, and has enough distinctive features in common with certain types of primate vocalizing, to be taken seriously as a possible survival of some pre- or proto- musical practice of our pre-homosapiens or possibly even pre-human ancestors.

The question then would be "what came next"? And, as I suggested in the previous post it looks very much like the very interesting and distinctive practice of yodel might provide us with a valuable clue. Yodeled hocket can be found along with or in place of shouted hocket in at least some of our A1 examples, the most significant difference being its use of discrete pitches. While yodeling can of course be considered a musical practice, it can also be regarded simply as a type of vocalization -- and is indeed used not only for musical expression but also signalling, in which case it becomes a special case of shouting. We might therefore consider yodel to be just one simple step away from the "shouted" or hooted vocalizing of primates. Which would be a fairly simple and straightforward way to think about how the use of yodeled pitches could have turned some sort of hooted duetting or chorusing among our ancestors into something we can today recognize as music "proper."

Once pitches are introduced, by the way, then all sorts of other possibilities emerge, because pitch relationships are goverened by harmonic relationships and harmonic relationships have a very important triple identity: as musical, mathematical, and also semiotic entities. So once early AMHs (Anatomically Modern Humans) began to play around with discrete pitches, initially just for fun I'd imagine, they are not only on the way to music, but also math and language. I'll probably have more to say about such possibilities in future, though I've already touched on some of it in my essay.

Consulting once again my Phylogenetic Tree, we see that I've placed style family A2, "Interlocked Hocket" as the next higher branch, implying that A2 can be understood as derived from A1. Note that A1 is characterized by the following musical traits, or "haplotypes": Hk (Hocket), CV (Continuous Vocalizing), It (Iterative), Y (Yodel), N ("Nonsense" vocables), RT (Relaxed Throat). A2 shares most of these traits, Hk, CV, Y, N, and RT, but not It, meaning that unlike the simple back and forth pitched or unpitched iterations so characteristic of shouted hocket, each voice uses more than one pitch, i.e., is melodic rather than iterative. In addition, three new traits are introduced: Int (Interlock), WI (Wide Intervals), and P (Phrased vocalizing). While shouted hocket can to some extent be regarded as a form of interlock (and has often been coded as such, cantometrically) the interlocking aspect is very simple, more like a rapid interchange, whereas the interlocking we find in A2 is more complex and can in fact be quite elaborate. While A1 is so iterative that we rarely hear any intervals at all, but just repetions of a single "hoot," shout or pitch, A2 does have intervals, which are usually coded as "Wide," i.e., anything from intervals of a third to a fifth or more, up to and beyond the octave. While this might seem like a considerable "advance" from an evolutionary perspective, yodeling actually makes such intervals easier and more natural to produce than the smaller "diatonic" intervals we are used to in the popular and concert music of the West. The symbol P, for Phrased, reflects the fact that most of the musical traditions represented by A2 are apparently organized according to the presence of an underlying musical "phrase," though this is often far from obvious. This can get us into fairly deep waters technically, so I won't go into too much detail at this point, but it's important to understand that A1 appears to have no structure aside from simple re-iteration of more or less the same sounds endlessly. A2 represents a considerable "advance" (if you will) as it does often involve a kind of musical "syntax" built around the repetition and/or variation of a basic musical unit -- or "phrase."

In the sense that A2 is in fact very similar to A1 in many respects, but also quite different in others, richer and more complex, suggests that there could be an evolutionary relationship between the two styles. The fact that both tend to be found among the same populations in Africa appears to strengthen this hypothesis.

Thus, if the duetting, chorusing and pant-hooting of certain primates might be understood as some sort of pre- or proto- music, and the shouted hocket of A1 as a kind of "missing link" taking us to the threshold of music "proper," then with A2, "Interlocked Hocket," we might be able to claim, however tentatively and provisionally, that we may finally, in some sense, have arrived at our long sought goal, humankind's earliest musical utterance.

35. Musical Beginnings

Having covered our first two questions, we move to the third: can style family A1, "Shouted Hocket," be regarded as prototypical for more elaborate types of interlocked vocalising (and instrumental performance) as found among the Pygmies and Bushmen and other indigenous groups in various parts of the world? Other questions present themselves at this point as well. Because if A1 is indeed a survival of some type of protomusical vocalization tradition of our pre-homosapiens ancestors, then what might have developed out of it that could be considered music "proper"? And what would be the difference?

To answer the last question first, we must return to the working definition of music I proposed some time ago: music can be understood as the use of discrete pitches or regular rhythmic patterns or both. Many A1 performances, like primate pant-hooting, duetting and chorusing, come close, but not quite. Shouting, chanting and primate vocalizing do not employ discrete pitches. The repetition of a shouted or hooted vocable does come close to what we could recognize as rhythm, especially when, as in all the cases we've considered above, the time intervals between such repetitions range from fairly regular to precise. But there is no pattern, only simple repetition.

At this point, the value of my working definition should be apparent. Because, regardless of all the many different types of audible expression, ranging from bird song to the roaring of lions and the howling of wolves, that could, in principle, be accepted as "music" (and under certain circumstances probably should be so accepted), we need to draw the line somewhere if we want to recognize that music is, like speech, a uniquely human achievement. While the working definition may be flawed, overly simple and incomplete, at least it enables us to make this very basic and in my view essential distinction.

So, assuming A1 was some sort of prototype, what could have been the next step that actually took us there? Or, more realistically, since I've posed an almost metaphysical problem: what clues do we have that might help us orient ourselves to such a question? Here's one, a recording we've already heard as an example of "shouted hocket," from the Dani of New Guinea. Here's another, from a group living in the same general area, the Huli. Note the way the Dani example combines shouting with yodeling. The Huli example sounds quite similar but the "shouts" are, like the yodels, pitched. The example I presented from the Esime section of the Aka Mokondi ceremony is, on the other hand, unpitched throughout. It is, however, only an interlude between episodes of fairly elaborate interlocked and pitched counterpoint, also featuring yodel.

What I'm getting at is first of all the strong similarity among all three examples, but also the ease with which precisely pitched yodeled cries fit with the shouted hocketing. Yodeling is an important part of the mix we haven't considered yet, but I have a feeling it's especially important at this juncture. For one thing yodeling is only one step removed from "hooting," as good a word as any to describe many types of primate vocalization. For another, it's centered in the "glottal" area of the vocal canal, thus also one step removed from the "gutteral" vocalizations so characteristic of the Tcoqma ceremony and the chanting of the Maasai, not to mention, once again, certain types of primate vocalization. Yodeling is in fact a very interesting practice, developing almost naturally, it would seem, from certain physical properties of the voice -- yet found only rarely among all the many musical traditions of the world.

One thing I find particularly interesting about yodel is the way it operates as a kind of vocal "overblowing," not very different, acoustically, from the overblowing so characteristic of wind instruments. As with a pipe, whistle, flute or trumpet, once the performer gets into the right "place" both mentally and physically, the lower notes of the natural overtone series can be produced almost effortlessly. And, as is well known, these pitches, transposed to within a single octave, form more or less exactly the sort of scales we so often find in so many different musical traditions worldwide. Musicologist Robert Fink has been particularly active in researching the implications of this relationship, and has formulated some theories of his own worth looking into.

I don't know of any primates or any other animals that seem interested in "overblowing" their voices in this manner to produce clearly articulated, discrete pitches. But this might have been one of the crucial differences between our ancestors and ourselves: the interest in both producing and "playing" with such tones in the context of duetted or chorused hocketed interchange.

34. Music Degree Zero?

Could music have originated as a result of the same speciation event that created what anthropologists call "Anatomically Modern Humans" (AMH)? And was this a sudden or gradual process? I've been doing some reading on an evolutionary theory called "Punctuated Equlibrium," as formulated by Ernst Mayr, Niles Eldredge and Stephen Jay Gould, that makes a great deal of sense to me. According to this theory, speciation is not a gradual process, as most evolutionists once believed, but relatively sudden, the result of the marginalization and drastic reduction in size of certain populations -- i.e., the sort of bottlenecks that can lead to founder effects, the same sort of mechanism the population geneticists are exploring with respect to human history and what I've been considering as the most likely explanation for the origin of certain musical style families. It occurred to me that perhaps the best way to deal with the question of musical origins was to consider human speciation in such terms and work the musical evidence into that context -- but on second thought I'm afraid such a discussion would be too speculative and take us too far afield, so have decided to put it on the back burner, at least for now.

Let's return instead to our most problematic question and try to deal with it as simply and straightforwardly as possible: is there a basis for thinking that any or all aspects of A1 might have originated in some form of pre-homosapiens vocalizing, as exemplified in the vocalizing of certain of our primate cousins today? I've already discussed, in section 21, some of the clues I've been following, primarily the "duetting" and "chorusing" of various species of apes, notably bonobos, as described by Merker and de Waal, but also Siamang gibbons, as their duetting has been carefully studied and some excellent videos are available. Here again are the links to the Siamang videos, to refresh your memory. And here's one more, an excellent example of Siamang chorusing. Pay special attention to the tendency of the animals to shout, or if you prefer "bark," back and forth in a reasonably steady rhythm. Scientists in fields such as primatology, paleontology, cognitive science and linguistics are paying close attention to this sort of phenomenon, apparently quite common among many different ape species, which is being carefully studied for its possible relevance to the origins of both music and speech. While all sorts of speculations are being proferred, pertaining to the essential nature of both of these very distinctive human accomplishments, considered in the abstract, in terms of genetics, cognition, infant psychology, brain anatomy, linguistic theory, etc., no one aside from myself appears to believe that ethnomusicological research could have some relevance and might lead to some insights in this area.

My own thinking in this regard has been profoundly influenced by the research presented in my "Echoes" essay, where I was able to demonstrate, to my satisfaction at least, that certain musical style families can persist for tens of thousands of years essentially unchanged. While some of the ideas presented in that essay are admittedly speculative, that conclusion is, for me, rock solid, as the evidence is truly overwhelming. And if that's the case, then it does make sense to consider whether any of these families could possibly be survivals from the earliest history of the human species.

I've zeroed in on what I call "shouted hocket" for several reasons: first, because, as I argued in the essay, it has a "pedigree" taking it all the way back to Africa well prior to the "Out of Africa" migration; second, because, for many reasons also discussed in the essay, "Pygmy/ Bushmen" interlock appears to be the earliest of all musical style families, with shouted hocket representing P/B in its simplest, most elemental form; third, because shouted hocket so dramatically resembles exactly the sort of primate duetting and/or chorusing that's been getting so much attention from so many scientists interested in musical and linguistic origins.

What makes the connection even more compelling, however, is the evidence I've been considering in the last few posts with respect to A1a, the subfamily of A1 characterized by the sort of gutteral, "panted" vocalizing involving rapid alternation of in- and out- breaths verging on hyperventilation -- as exemplified by circumpolar "throat-singing," the Bushmen Tcoqma ceremony and the ritualized chanting of Maasai warriors. Many primates vocalize in a remarkably similar manner, as exemplified by the so-called "pant-hoots" commonly heard among both bonobos and chimps, but possibly also the very rapid duetting and chorusing sequences I've been focusing on. According to a recent article in the International Journal of Primatology, "Apes and larger gibbons may be able to produce fast extended call sequences without the risk of hyperventilating because they can re-breathe exhaled air from their air sacs. Humans may have lost air sacs during their evolutionary history because they are able to modify their speech breathing patterns and so reduce any tendency to hyperventilate." (The whole issue of the role of primate air sacs in continuous vocalizing raises all sorts of interesting questions with respect to possibly related issues in musical evolution, such as the the role of certain instruments requiring recycling of the breath, such as the didjeridoo, or instruments like the bagpipe, with air sacs built in.)

On this basis, we could postulate a situation in the distant past where a newly "speciated" band of Anatomically Modern Humans (AMH) were attempting to continue an older tradition involving some combination of pant-hooting, duetting and chorusing that might have sounded very much like A1a, only without air sacs, which would have placed their vocalizing dangerously near the threshold of hyperventilation, unconsciousness and trance. The only significant difference between the vocalizing of their predecessors and the newly minted AMH version would have been the need to work out some sort of strategy for avoiding or at least delaying hyperventilation and its effects, which does indeed seem to be an important feature of A1a. However, the continual danger of falling into trance while vocalizing in this manner might provide a clue to the common association of music with trance and the origins of shamanism as well.

According to the above theory (admittedly highly speculative, but nevertheless rather interesting I should think), style family A1, with its two branches A1a and A1b, would represent not so much the origin of music as the survival of a hominid tradition that preceded it.

33. Almost There

In post 28, I asked three questions, all pertaining to my hypothetical style family A1, "Shouted Hocket": 1. is it truly a family, or simply an agglomeration of different practices that happen to sound alike? 2. do any or all the practices included in A1 contain elements that could be regarded as survivals from our pre-homosapiens, or even pre-human past? 3. can A1 be regarded as prototypical for the more complex types of hocket and/or interlock I've identified as A2, A3 and A4? If the answer to all three is "yes," then we will have arrived at the origin of music, and our passage from 000000 to 000001 will be complete. (Could there have been more than one origin? That's a rather tricky question I'll defer for now.)

One thing we've learned from our investigation so far is that there may be at least three subgroups within what I've been calling the A1 "style family": 1. the "throat-singing" practice identified by Nattiez, with certain very specific accoutrements not apparently found outside the circumpolar region; 2. the shouted, gutteral style of hocketed "panting" found among the Bushmen, Maasai and other African groups; 3. all the instances of non-gutteral shouted hocket in various parts of the world, including those for which I've specifically provided examples and at least some or possibly all of those retrieved by my Cantometric search.

Subgroups 1 and 2 are strikingly similar in so many respects that I think it reasonable to group them together as a single family. Nattiez himself tends to minimize the importance of certain features of the circumpolar tradition that might set it apart, such as animal imitations, the presence of dancing and drumming, the relative position of the singers, whether they are seated or standing, dancing or stationary, etc., as these are all associated with specific functions and contexts, thus belonging to the realm of the ephemeral signified, as opposed to the more basic features of the signifier: the intricate interlocking of parts, the "panted" vocalizing on both in- and out- breaths, the continuous flow of sound, etc.

The recordings I've heard from subgroup 3 don't sound particularly gutteral or breathy, and tend to lie higher in the voice. The Balinese "Monkey Chant" does involve trance and probably hyperventilation and some form of shamanism as well. It's difficult to say what role, if any, these factors might play among most of these groups, an assessment that will have to wait on additional research. For now, it's probably reasonable to divide A1 into two subfamilies, A1a, for the gutteral, "panting" styles, and A1b for the styles involving more "normal" use of the voice. It should be noted, by the way, that gutteral vocalizing and yodel are both centered in the glottal area of the vocal cavity.

So, to answer the first question, regarding the status of A1 as a true style family, I'd have to say that, at least as far as subgroup A1a is concerned, there are so many striking points of similarity that yes, even despite the great geographic distances involved, the evidence does appear to support such a conclusion. As for A1b, a comparison of the Aka "esime" with the examples we've heard from Madagascar and New Guinea, and even South America, for which I could play several other very similar examples, reveals striking similarities that would be almost impossible to distinguish in a blind testing -- with the complex, interlocked hocketing of the Balinese example lying somewhere between these and the Bushmen Tcoqma. Moreover, all performances from all groups we've heard in both A1a and A1b, regardless of the differences in voice placement and breathing technique, share a very distinctive type of interaction and overall sound, organized as a continuous, uninflected flow, that strongly suggests, for me at least, a common source. This assessment is reinforced by other factors involving other sorts of evidence that I'll be discussing presently.

In the past such a conclusion might have seemed totally fantastic, and without any support whatsoever. In the light of the genetic evidence, and the completely new picture it presents of humankind's earliest migrations, it is possible for the first time to get a fairly clear handle on how all these different populations in so many very different parts of the world could be related.

Now for the second and most problematic question: is there a basis for thinking that any or all aspects of A1 might have originated in some form of pre-homosapiens vocalizing, as exemplified in the vocalizing of certain of our primate cousins today? With such a question we are definitely in the realm of speculation -- but not without a certain amount of support from genetic science, coupled with some basic principles of evolution.

To answer such a question we need to consider not only the origin of music, but the origin of modern humans, which may well have resulted from a unique speciation event, resulting from a severe population bottleneck.

To be continued . . .

32. Are We There Yet?

I believe I've established 1. that throat-singing in essentially the same style described by Nattiez must have been practiced many thousands of years ago, well prior to the 4,000 to 5,000 year figure he suggests; 2. that shamanism must also have been practiced by the same ancestral population during the same period. We can go on to conclude that the association we now find between the spiritual practice and the vocal style must be equally archaic, since that association is also so widely distributed today, among the indigenous peoples of the north. While the Inuit might appear to be an exception, there is strong evidence, as Nattiez has shown, that their throat-singing "games" once had a shamanic function as well.

What is the nature of that relationship? In section 29, I referred to Nattiez' descrition of throat-singing as "the production of a continual stream of sound through rapid alternations of audible exhaling and inhaling, 'which create what can be called a 'panting style' . . . .'" [p. 401] I then referred to an essay on hyperventilation and trance in Nganasan shamanism, which concludes that this type of vocalizing, described in terms of a "panting motive," can be and often is used to induce trance. According to sources referenced by the authors, this type of ceremony can last either all night or until the dancers collapse. Here is a recording we've already heard (in section 22) from Kamchatka with many of the same characteristics.

Now I'd like to do essentially what Nattiez did, only on a much larger scale, by comparing similar vocal "signifiers," but over a far more extensive geographic span, taking us far beyond the circumpolar region. Because there are other instances in the world where "shouted hocket," combined with hyperventilation, singing in the throat and "panting style" work together to invoke trance in a context of shamanic ritual. The most dramatic example is the Ju'hoansi Tcoqma we've already heard. While the context is different in certain respects, as this is a boy's initiation ceremony and not a bear dance, they actually have a lot in common: both ceremonies go on for hours to the point of exhaustion or collapse; both involve hyperventilation and "panting"; most of the vocalizing in both is "gutteral," i.e., centered in the lower throat, or "glottal" area; both are organized according to the characteristically shouted form of hocketed interlock that's been the focus of our discussion for some time now; and both invoke shamanic trance.

There are in fact many other tribal groups in Africa whose vocalizing combines shouted hocket with gutteral "panting," hyperventilation and trance, though not always associated with shamanism per se. A remarkable variant can be found, for example, among Massai warriors, who chant, according to Malcom Floyd, in "a semi-vocalised, semi-pitched, rhythmic hyperventilation accompaniment technique" called nkuluut, characterized by low pitched, gutteral sounds. According to Floyd, this type of performance serves both as a source of arousal and containment of that arousal, in delicate balance. "It will also be noted, however, it is not uncommon for the arousal to reach depths which make containment impossible, for nkuluut to overpower melody, resulting in extreme cases in seizures leading to catatonic states." While such chanting can be organized according to the call and response litany format so commonly found in Africa, it can also take the form of a type of shouted interlock very close indeed to both Bushmen and Paleosiberian practice.

What we need to consider at this point is how all these practices relate to one another. It's tempting indeed to ascribe the origins of shamanism itself to the ancestors of the Bushmen, rather than the Paleosiberians, as has generally been assumed. And the musical evidence certainly seems to support such a theory, since we apparantly find so many of the essential ingredients of Paleosiberian shamanic vocalization in the Tcoqma ceremony. If our earliest fully human ancestors did indeed emerge from Africa to populate the rest of the world, then one could make a strong case for some sort of proto-Bushmen origin for shamanism -- or at the very least an origin somewhere in Africa. The argument would look very much like Nattiez' argument for the unity of circumpolar throat singing, for which he posited a "protoform" dating to at least the time when the three language families, Eskimo-Aleut, Chuckchi-Kamchatkan and Altaic, diverged. If we feel comfortable including Bushmen shamanism (and/or Massai catatonism) in the same mix, then we would need to think in terms of much larger and far older linguistic/genetic super-families which would most likely have diverged in Africa prior to the "Out of Africa" migration. By the same token, if we are looking for the same "protoform" posited by Nattiez, his dates would have to be adjusted to a period roughly 100,000 years ago, very close indeed to our year 000001.

I must hasten to add that the whole question of the origins (and even the nature of) shamanism is not something I am qualified to pass any sort of judgement on. What I've presented above can be described as informed speculation at best. Maybe also uninformed. Or half baked. I do like it. And it does make sense, to me at least. But I'd want to know a lot more about this topic before proceeding any further in that direction. It was a detour anyhow, because our main concern is with a topic much dearer to my heart: the origins of music.

31. On a Clear Day You Can Hear 000000

As I argued in the previous post, there is no reason to assume the throat-singing tradition studied by Nattiez could not be far older than the time span he considers, ca 4000-5000 years. (By the way, this type of "throat-singing" should not be confused with the Tuvan type of throat singing found in Mongolia, where more than one pitch is produced by a single voice -- that's a different tradition.) For now, however, I want to proceed as he did: cautiously. In this spirit, let's do a little thought experiment by projecting ourselves back 4000-5000 years and asking if we have any means of determining what his protoform might have been like at that time. Most ethnomusicologists would throw up their hands at such a question, claiming we have no way of telling what was being sung by any group at a period so remote and undocumented. They would be wrong.

If linguists were to find several different tribal populations, scattered over such a vast territory, all speaking essentially the same indigenous language, they'd be stunned. Because this would imply that either the language was only recently disseminated -- but how could that be, by what means could it have traveled? -- or that it stems from a much earlier prototype that must have been essentially the same as its derivatives, which could not have changed in any significant way in all the time it took for all these different groups to diverge from one another and migrate to all these remote places. This would cause a sensation because, as is well known, language doesn't work that way. All languages that have ever been studied show unmistakable signs of having changed over time, and linguists have worked out sets of rules according to which they believe such changes take place. When considering all the many languages of all the different circumpolar groups encompassed in Nattiez' research, it is certainly possible to assume a common prototype, but another matter entirely to reconstruct it, since all the languages it is thought to have given rise to are now so different from one another.

But the throat-singing traditions are not very different from one another at all! While some are associated with shamanism and others not; some danced and others not; some sung in groups, others in pairs; some sung standing, others sitting, such differences reflect the different meanings this practice has among different populations, i.e., the various signifieds produced from what is essentially the same signifier. This is, in fact, a situation encountered quite often in the realm of music, where we find populations distributed throughout an entire continent, speaking different languages, but very often singing and playing in remarkably similar ways. This is especially true of Australia, but also much of the Americas. And, as I have already argued, there are populations scattered among other populations, such as the Pygmies and Bushmen, who also share a distinctive musical style despite major linguistic differences. Clearly musical traditions operate historically in a manner totally different from those of language.

So, returning to our question about the nature of Nattiez' prototype, the answer is simple: the prototype of 4,000 years ago could not have been other than essentially the same musical "signifier" he has identified among all the various circumpolar groups today. If we want to know what it sounded like we need only play some recordings. How do we know this? If the prototype were different from its derivatives in any important way, then the striking stylistic similarities apparent in all the different groups today could be neither explained nor understood.

Once we fully digest the reasoning behind this fairly straightforward, though by no means obvious, conclusion, we are in better position to press even farther backward in time. Because there is no reason to assume this style or any other musical style would just happen to develop at the same time some languages branched off. Logically, we can assume that throat singing must have already been entrenched as a cultural tradition for a very long time prior to that divergence. And if a musical tradition can persist for well over 4,000 years, then why not 8,000? Or 16,000? Or 32,000? Or longer? How far back can we hear?

Let's consider more evidence. The linguistic divergence identified by Nattiez as taking place 4,000 to 5,000 years ago tells us something not only about throat singing but also shamanism. Since all these groups have shamanic roots (even the Inuit apparently, though they no longer practice it), we can only assume that this practice must also date back to at least the same time as the linguistic divergence -- since again, as with their music, there would be no way to otherwise explain its widespread distribution. And, again, as with the music, we have no reason to assume that shamanism originated at the same time. Like throat-singing, it must have originated many years earlier.

We can say more. Because throat-singing is also, as Nattiez has demonstrated, quite strongly associated with shamanism, either explicitly or implicitly, among all these circumpolar cultures. How far back might that association go?

30. 000000 to 000001 part 3

As Nattiez demonstrates, the circumpolar "throat-singing" tradition he's identified can be understood as a unified style family despite being interpreted differently in different social contexts, as either a shamanic practice or a game. It's important to understand that Nattiez achieves his results phylogenetically, by working his way backward, with the aid of archaeological, linguistic and genetic evidence, to a "protoform" for all manifestations of this style, a common root dating to thousands of years in the past. Is it possible for us to connect all or at least some of the other types of "shouted hocket" we've been considering, from Melanesia, Indonesia, South America, Africa, etc., via the same protoform? And if so, would the association Nattiez has found with shamanism still be relevant?

As should be clear by now, a significant amount of additional research into both the musical practices and cultural background of all the groups we're considering would be necessary before any solid conclusions could be drawn. Nevertheless, there is a great deal we are in a position to consider at this point, if only provisionally. If all these instances could indeed, in one way or another, turn out to be associated with shamanism, that would certainly strengthen the connection. But as Nattiez has shown, it's not necessary for the social function, the signified, to remain the same, since the same signifier may, over time, come to take on different significations. What's most important, it seems to me, is the ability to trace the various manifestations of the "signifier" we are examining back to a single source phylogenetically, i.e., historically. If we could do that, then I think we might at least be able to develop a convincing circumstantial case. The presence of shamanism in certain contexts might indeed provide an important clue, but its absence may not matter at all.

Let us therefore consider the protoform deduced by Nattiez. This must be understood not simply as a theoretical construct but a real musical style, practiced by real people at a certain time and place, i.e., both historically and geographically. What could this musical practice have been? How would it have differed from the throat singing of today? What sort of people could have been singing in this manner? And where could they have been living? Taking historical linguistics as his guide, Nattiez associates the protoform with the time when three language families, Eskimo-Aleut, Chuckchi-Kamchatkan and Altaic were, according to certain linguistic theories, one and the same -- a connection reinforced by certain genetic evidence produced by the noted population geneticist Luca Cavalli-Sforza. (I'm delighted to note, by the way, that Nattiez has preceded me not only by considering the genetic evidence but also by suggesting that the musical evidence might actually have a bearing on the way the supposedly "more scientific" linguistic and genetic theories are evaluated.)

Where Nattiez was guided by the linguistic and genetic evidence pointing to a circumpolar root culture 4,000 or 5,000 years old, I am allowing myself to be guided, at least for now, by the far more ambitious "Out of Africa" theory, with characteristic time spans ranging into the tens of thousands of years. I am also being guided by the "standard candle" (see section 11) afforded by Chen's estimate of at least 72,000 years for the branching of the Aka Pygmies from the hypothetical founding group of homo sapiens sapiens, several thousands of years prior to the estimated date of Bushmen divergence. As I've already argued, the extraordinary similarities between the vocalization styles of the Pygmies and Bushmen, combined with the genetic results, warrant a complete rethinking of our sense of how long a particular musical tradition may persist unchanged, at least as far as its most salient stylistic elements are concerned.

There is no reason to assume, therefore, that Nattiez' throat-singing protoform is only 4 or 5 thousand years old, simply because the divergence of some language families might date to that time. As it is now possible to argue that musical style can be a far more conservative force than language, the stylistic nexus studied by Nattiez might well be rooted far more deeply in the mysteries of our human past.

29. 000000 to 000001 part 2

Can what I've been calling "Shouted Hocket" be legitimately considered a musical "haplogroup" (style family)? As a first step in the systematic investigation of this question, I've done a Cantometric query based on four characteristic traits (see previous post). The query produced a list of cultures, almost all indigenous tribal groups, many from Africa. Ideally the next step would be to listen systematically to each recording to be sure it really fits. Unfortunately I don't currently have access to most of the recordings represented in the sample, though all are presently being archived by the Association for Cultural Equity (the organization Lomax founded, now directed by his daughter, Dr. Anna Lomax Wood). The presence of so many African groups is to be expected if the style did in fact originate in Africa as I've postulated. Not every group may actually be performing in true "shouted hocket" style, however, so it's best to defer judgment until the recordings can be checked.

We can add to the list the groups I included in my examples (see post 22) that weren't found by the query. The Aka Pygmy esime is almost always performed as an interlude between songs. This would have complicated the Cantometric coding, which might not have picked up on the salient characteristics of that one segment. The Balinese "Monkey Chant" also involves a mix of performance types that would have made the coding more tricky than usual. The Mehinacu recording is new and not yet coded. The Mikea example came up positive for three of the four traits in our search, but should clearly be included on the basis of its overall sound. The Ainu example also posted three out of four and does, in fact, sound a bit different from the others -- nevertheless in my opinion it should still be included. Clearly the Cantometric results work best as a provisional overview and should be supplemented whenever possible with a close analysis of whatever recordings are available.

We must now ask ourselves if there are other aspects of any or all of these performances that could give us some clue to their meaning, cultural, historical or both. We can begin with Nattiez' conclusion that Inuit throat singing was in all likelihood, along with the Siberian and Ainu variants he cites, part of a circumpolar shamanic tradition originating, by his estimate, 4,000 to 5,000 years ago. The Balinese "Monkey Chant" is also thought to have originally been part of a shamanic ritual. The Bisorio example is described as a "house building," which might imply that it was some sort of work song. Since house building in Melanesia is often accompanied by shamanic rituals, however, we can't be sure. Since music among groups like the Dani and Huli is so often associated with ritual, it seems quite likely the hocketing we hear in those examples also has shamanic associations.

The Aka Pygmy "Mokondi" ceremony, a ritual involving a masked dancer who, in all likelihood, goes into trance, may or may not be associated with shamanism. The Ju'hoansi Tcoqma unquestionably is shamanic, associated with an all night initiation ceremony for young boys, in which many men go into trance and perform healing rituals. Among the Ju'hoansi, the great majority of males are considered shamans, with healing powers.

While shamanism was for many years thought to have originated among the Paleosiberians, the "Out of Africa" model may force anthropologists to rethink that hypothesis. The fact that shamanism plays so important a role in Bushmen culture suggests that it may have originated in Africa, making its way to the rest of the world via the initial "Out of Africa" migrations and their post-bottleneck aftermath.

According to Nattiez, a pervasive characteristic of throat-singing is the production of a continual stream of sound through rapid alternations of audible exhaling and inhaling, "which create what can be called a 'panting style' . . . the main feature common to the three cultures under consideration." [p. 401] In a remarkable essay currently available on the Internet, Triinu Ojamaa and Jaak Aru demonstrate the relation between hyperventilation and trance in the Nganasan Bear Dance, yet another example of the shamanic circumpolar tradition explored by Nattiez. According to Ojamaa and Aru, "Inspiration and expiration alternate in a certain rhythm. We can characterize the accompaniment as rhythmically organized panting." Their essay presents convincing evidence, both musical and biological, of the relationship between hyperventilating and trance, a pervasive feature of shamanism.
Signs of a somewhat similar type of "panting" suggestive of hyperventilation, can be found in our examples from Kamchatka, the Inuit, Ainu and Hupa, as might be expected, but also the Ju'hoansi -- while there are strong indications of something similar going on in the uncannily rapid, trance inducing, interlocking of the "Monkey Chant."

28. From 000000 to 000001

We are now in a better position to appreciate the value of performance style per se, which may indeed be taken out of its immediate setting to be recontextualized as part of larger, more broadly applicable, contexts. As Nattiez' paper illustrates, a number of different research methods can be applied to the evaluation of performance style, including direct observation, filming, photography, audio recording, musical (as well as verbal) transcription, consultation with local informants, and the study of first-hand reports and analyses, both musicological and anthropological. Cantometrics was designed to take advantage of certain properties of recorded performances that make them especially amenable to comparative research: repeatability, consistency, availability, ease of storage, replicability, and, above all, comprehensiveness, as there now exists a wealth of recordings of traditional music from almost every corner of the world. While Cantometrics, with its database of roughly 5,500 encoded performances, can indeed provide a broadly representative, consistent and systematic set of research tools, it is essentially a heuristic approach which should be supplemented wherever possible by some or all of the other methods listed above.

Now that certain basic methodological issues have, hopefully, been clarified, we may return to the issue at hand, which can be summarized thus: 1. can style area A1 on my phylogenetic map, i.e., "Shouted Hocket," be regarded as a genuinely meaningful musical "haplogroup," or is it simply a collection of different practices with unrelated histories that happen to sound alike? 2. can certain aspects of any or all of the practices I've called "Shouted Hocket," be understood as survivals of some form of "duetting" or "chorusing" as practiced by our pre-human ancestors? 3. can "Shouted Hocket" be regarded as prototypical for more elaborate types of interlocked vocalising (and instrumental performance) as found among the Pygmies and Bushmen and other indigenous groups in various parts of the world? In other words, could style area A1 really be the "missing link" that can take us from 000000 to 000001 on my musical calendar?

Let's begin by consulting the Cantometric database. While I didn't define "Shouted Hocket" (or any of the the other style families) exclusively in Cantometric terms, we can get a rough approximation by looking for four specific traits: interlocked vocal organization; short or very short phrase lengths; one or two repeated phrases; and forceful to very forceful accent. A query for all performances exhibiting all four traits produces the following set of hits: the Ju'hoansi Bushmen tcoqma ritual we've already heard; 3 Mbuti Pygmy performances; 5 Inuit performances; 1 from Kamchatka (Siberia); 2 Yukaghir (Siberia); 1 Hupa (North American); 1 Motilon (South American); 1 Amahuaca (S. American); 1 Shuar (Jivaro, S. Amer.) 2 Ata Krowe (Flores, Indonesia); 1 Ajie (Melanesia); 2 Biami (New Guinea); the Bisorio performance we've heard (N. Guinea); 4 Dani, including the example we've heard (N. Guinea); 2 Huli, including he example we've heard (N. Guinea); 1 Yali (N. Guinea); 1 Hanunoo (Phillipines); 1 Georgian (Caucasus, Asia); 1 Russian; 1 Hungarian Gypsy; 1 Village India; and several African groups: Bundo, Mbala, Ndongo, Lese, Anaguta, Forest Bira, Toma, Meru, Tandroy, Masai, Samburu, Pondo, Wodabe Fulani, Hamar, Tuareg, Ajuran, Dorze, Gamo, Konso and Gio.

Aside from the Russian, Gypsy and Village Indian examples, which apparently don't fit, all the others could in fact be good candidates for the style area in question, which, according to the phylogenetic map, would be very broadly distributed among tribal groups in Africa, Melanesia, New Guinea, Siberia, and the Americas, notably Inuit and California. The South American examples may also be meaningful, as I explain in my essay, though the connection is by no means obvious. The Balinese Monkey Chant is not on the list, probably because the elements of shouted hocket are embedded in a more complex structure.

The above query is comprehensive, in other words it is as significant for the very large number of groups not represented as those which are, telling us, in fact, that the great majority of the worlds peoples very likely do not have musical traditions characterized by shouted hocket. Which makes those who do all the more interesting. Examination of the list gives us, I would think, a pretty good sense of both the strengths and weaknesses of the Cantometric approach. A well designed query, based on a good sense of what to look for, can provide a very useful overview of the stylistic terrain, an excellent starting point for further investigation. It will almost always, however, contain certain things that probably don't fit and omit others that do. This is to be expected from a heuristic system, necessarily provisional and preparatory.

27. How to Listen to a Recording

The question before us, as presented in post 24, is: "what can we make of all these [recorded] examples, what do these resemblances mean, how seriously should we take them, are they significant or superficial?" I have been arguing, with the help of Agawu and Nattiez, that a great deal of importance can in fact be gleaned from a recording, which can indeed be understood, in Agawu's terms, as a "text," or, as for Nattiez, a "sonorous symbolic form" functioning as a "signifier." But a certain element of caution is necessary at this point, because, before one can read recordings like texts, one must first learn how to read such texts.

Whenever I present musical examples for comparison, either on the Internet, or when teaching or lecturing, I tend to cringe a bit inside, because I never know how my listeners will react. I've had people tell me that a recording I've just played, from Polynesia, or Central Asia, or Japan, sounds "just like" a song their grandmother from Poland used to sing. Or, on the other hand, when I present two performances in styles that sound quite similar to me, I'll hear something like "Oh well, what can one say about such comparisons, it's all subjective, you can always find something in one song that sounds like something in another." Which may of course be perfectly true. If one doesn't know what to listen for, then anything could sound like anything else.

If one were handed the text of a song from one of Shakespeare's plays, for example, and one had never read much poetry or any Shakespeare, then it might remind one of a TV jingle or popular lyric. After many years of study, however, you might feel confident you could distinguish a line of Shakespeare from that of any of his contemporaries.

Even in the case considered by Agawu, in which an African musical "text" is "read" primarily for contemplative or aesthetic purposes, such music might well, at first hearing, sound chaotic or simply monotonous. While, as Agawu has argued, we need not understand how this music functions in its social context in order to listen meaningfully, we may need to spend some time familiarizing ourselves with a new style before we are able to hear anything meaningful in it at all. And if we want to understand such "texts" in the spirit of Nattiez, not simply as objects of contemplation, or simply free floating signifiers freed from their signifieds, but potentially meaningful indices of traditions rooted in the deepest recesses of history, then we must certainly spend some time learning to "read" those traces.

By the time Alan Lomax conceived Cantometrics, he had spent many years in the field, recording a great many traditional performances not only in the United States, but the West Indies, Britain and parts of Europe. As editor of the Columbia World Library of Folk and Primitive Music, he'd spent hundreds of hours listening to and evaluating field recordings from a wide variety of different peoples all over the globe. When I came to work with him, in the summer of 1961, I was far less experienced, but did at least have two years of intensive analytical listening behind me, as an ethnomusicology graduate student. As we worked together on developing the Cantometric method, we did a great deal of additional, very careful and critical listening to all sorts of performances from all corners of the world, to be sure we were making the sorts of distinctions that would prove both meaningful and useful.

Some years later, Lomax, with the assistance of jazz musician/composer Roswell Rudd, put together an extensive series of examples, from all over the world, on seven audio cassettes, as part of the training system in Cantometrics they'd developed [Cantometrics: an Approach to the Anthropology of Music, by Alan Lomax, The University of California Extension Media Center, Berkeley, 1976]. Familiarization with a wide variety of performance styles from a variety of different contexts worldwide is thus a very basic and necessary aspect of Cantometric training -- as it should, indeed, be expected of anyone seriously involved in the broad-based comparative study of traditional music.

Returning to the initial question, therefore, of whether or not the resemblances I've been pointing to are significant or superficial, the answer would be that this is by no means a simple or straightforward matter, certainly not one that could be decided by incidental listening rather than in-depth study. A lot would depend on ones familiarity with all the other major traditions of world music, as a basis for comparison. But in the long run there would be, as I see it, no substitute for the sort of systematic research made possible by a methodology such as Cantometrics.

When, therefore, I present some recorded examples of traditional performances on this blog, in order to illustrate some point I am making, the reader/listener must realize that the connections I'm hearing may not be immediately apparent to a novice -- though, with the aid of my comments I do hope it's possible for most here to listen with enough sympathy and understanding to follow the gist of my argument. And I hope I can count on at least some of you to keep things interesting by disagreeing from time to time.

26. All Contexts Great and Small

For the great majority of ethnomusicologists today, context is fundamental. One is expected to study the manner in which various types of music function within a social context, narrowly defined: a particular tribe, or set of closely related tribal groups, a village, town, religious community, age group, etc. Such an approach conveniently squares with the prevailing research model, i.e., field work within a small community. What is not encouraged is any effort to broaden the context beyond the fieldwork model. The prevailing assumption would seem to be that one cannot, and in fact must not, compare a certain practice in one place with what might look or sound like something similar in another place because that would entail the removal of both practices from their "context." And since it is only in terms of the local context that one can assess the real meaning of any such practice, its true function in the community, then any but the most cautious and limited attempts at comparative study would, literally, be meaningless.

There is a long history behind this attitude, now hardened into a dogma, which I've touched on in earlier posts and won't return to here. Suffice it to say that in my opinion the decision to favor, and indeed enforce, such a narrow approach to the study of world music has been a mistake. I've already offered the testimony of a leading ethnomusicologist who is also a distinguished semiotician, Kofe Agawu. In the quoted segment Agawu focuses on "contemplative" listening as opposed to contextual "understanding," arguing that African music can be fully enjoyed via the "sonic trace" without any need to justify such involvement by poring over ethnomusicological field studies to become more fully aware of how it functions in its "social context." What Agawu is not saying is that we can enjoy African music simply as a disembodied formal structure, removed from any human context whatever. For him, music is a text, to be "read," as one reads a book, the implication being that the music itself can already "speak" to us with a meaning of its own, i.e., that it can indeed be comprehended "out of context," as in fact it routinely is, anytime anyone plays a recording.

Why is this the case? In my view, a recording of a piece of music already contains encoded within it a context of its own, a broader context, with a deeper meaning, than anything to be gleaned from an ethnographic study of how it functions in some specific community. It is this very deep well of meaning and emotion that, in fact, sets musical performance apart from any other mode of human behavior and ought to have given the comparative study of the musical practices of the world a place at the center, rather than the periphery of anthropology.

As an illustration of what's at stake, I would like to discuss a remarkable study by Jean-Jacques Nattiez, known primarily, like Agafu, as a semiologist: "Inuit Throat-Games and Siberian Throat Singing: A Comparative, Historical, and Semiological Approach" [in Ethnomusicology 43, 3, 1999]. Nattiez begins by presenting what looks like the sort of standard observation one might expect from a typical ethnomusicologist: "I wish to demonstrate how throat-games and throat-songs look alike but have different meanings in various cultures around the pole . . . " How often have I read, or heard stated some variant of that notion -- how "we can't" or "it is dangerous to" or "one should never" make connections between things that may look or sound alike but "have different meanings" in different cultural settings. Significantly, however, Nattiez is not an ethnomusicologist and his understanding of the problem is quite different from what might be expected.

He begins by presenting examples of "throat singing" from various "circumpolar" groups: certain Siberian tribes, the Ainu of the Sakhalin Peninsula and Hokkiado, and various Inuit groups of Alaska and Northern Canada. In his original oral presentation, he played recordings of various examples, but in the printed version he offers only photographs, descriptions and notated transcriptions of a style that from my perspective would be regarded as a variant of A1, or "Shouted Hocket." (For examples of throat singing roughly corresponding to the traditions studied by Nattiez, see section 22 below, where I've provided recorded clips from Kamchatka, the Inuit and Ainu.)

Nattiez notes that there is a significant difference in meaning between the various Inuit traditions, understood simply as games, and the traditions of the Siberians and Ainu, where throat singing has strong associations with shamanism. Understood "in context," one might thereby dismiss the very strong stylistic resemblances among all these different practices as of no importance since one functions merely as a game while the other has a very different function, as part of a shamanic ritual. Not easily discouraged, Nattiez brackets the issue of function to consider more generally applicable explanations for all the many similarities, along three categories: "universalist," "diffusionist" and "phylogenetic." Rejecting the first as unlikely and the second as improbable (because the vast geographic distances all but rule out direct influence), he embraces a phylogenetic interpretation quite close in fact to the hypothesis I have been outlining here: "Among the Inuit and the people of Asia, analogies of distribution between linguistic features . . ., archaeological artefacts . . . and genetic data . . . have been established. This strongly suggests that these connections are the result of a migration which occurred 4,000 to 5,000 years ago . . . " [411-412] He concludes that what is true for the linguistic, archaeological and genetic connections must probably be true for the musical practices, especially in the light of the long series of stylistic similarities he then enumerates. They must all stem from "common protoforms, as is the case for genes and languages..." [413]

Nattiez then goes on to consider "why these symbolic forms do not necessarily have these religious connotations today, particularly among the Canadian Inuit. The semiological distinction between the signifier and the signified in an historical perspective will help us to understand how a similar form (a similar signifier) gets a new meaning (a new signified) in a different culture... From this situation,we may draw broader conclusions of interest for general musicology and semiology. In sonorous symbolic forms, the form, the signifier, best resists transformations through time. However, the signified, the religious significations of the animal and nature imitations associated with these forms, are evanescent." [414]

The conclusion stated above is extraordinary, literally turning on its head the long cherished assumption that the only meanings to be considered are those signified in the context of a particular society, meanings which, for Nattiez, must be considered "evanescent." As his research clearly demonstrates, it is the musical signifier, what Agawu has called the "sonic trace," that has the power to persist through the ages, from one social context to the next, thus offering the more reliable index of human history and, potentially, the more convincing and satisfying insight into the meaning of music in the broadest and deepest sense.

It is necessary to make one more point with respect to all the above argumentation. While Nattiez explodes the hegemony of the functionalist and contextualist assumptions so dear to so many ethnomusicologists, he clearly could not have arrived at the conclusions he did without a very deep prior investigation into the functions and immediate contexts of throat-singing as manifested in all the different cultures studied. In fact, it was the testimony of an Inuit woman, who recalled some things her grandmother had said about the association of Inuit throat "games" with hunting magic, that provided him with an important clue to the possible origin of such games in shamanistic practice. [405]

Nattiez's paper by no means tells us that it is safe to ignore social context and function, especially since so much in it is based on exactly the kind of close ethnographic investigation that so many ethnomusicologists do so well. What it tells us is that this is not enough, that there is a larger context that must be considered as well. As I have often stated, both the generalist and specialist need each other. While the Cantometrics method is indeed limited to some of the many things one can learn by "reading" the "text" offered by the signifier of recorded musical performance, the Cantometrics project, like the project under discussion here (also based in Cantometric methodology), took a great many other things into consideration, including the unquestionably valuable contributions and insights of so many workers in the field. I sometimes like to joke (but it really isn't a joke) that, though I am often critical of a great many of my ethnomusicologist colleagues, I am probably the most avid reader, and appreciator, of their work -- from which I have learned and continue to learn a great deal.

25. Music, Text, and Context

An amost obsessive catechism for ethnomusicologists over the last 30 years or so has been the phrase "Music cannot be taken out of context." And by "context" is invariably meant not only the specific social context in which music is performed, but above all the way in which music functions in that context. So the typical mode of research for ethnomusicologists has been to go out into the field to do what amounts to an ethnographic study. Recordings are made to be sure, but the primary goal is to work as a participant observer in order to get under the skin of the phenomenon, by talking with and interviewing performers and above all, learning to become a performer oneself. Once an insider's view of a culture is achieved, then the true function of its music in context and the true role of its performers in context can be understood -- and then communicated via the writing of field reports, academic papers, books, etc. but also via teaching, not only in the classroom, but also by training students in the musical traditions one has learned, directing an ensemble made up of traditional instruments from the culture in question, inviting traditional master performers to come to campus to lecture, give lessons, perform, etc. In other words by, you guessed it: taking the music 0ut of context. As this is only the first of many such contradictions, I'll let it pass for now without comment.

So where were we? The music of some specific society has been studied closely, in situ, its various functions have been determined and the ethnomusicologist returns, as an expert in how certain specific types of music function in one particular "native" context, to pass her or his knowledge on to others. All well and good. In fact highly commendable. But to what end? Are such activities so important that they can be justified as ends in themselves? Ethnomusicology for the sake of Ethnomusicology, as in "Art for Art's sake"? If ethnomusicology is to be regarded as a form of scholarship or perhaps even a science, then clearly the study of one particular body of musical practices in one particular place can be justified only if that study can be put to use as part of some larger project, with broader goals. In other words, any ethnomusicological study, to justify itself, must at some point be taken, you guessed it, out of context so it can be compared with other studies, with the goal of comprehending different types of musical practice in a variety of different contexts.

Or, in the words of philosopher Paul Ricoeur, "An essential characteristic of a literary work, and of a work of art in general, is that it transcends its own psychosociological conditions of production and thereby opens itself to an unlimited series of readings themselves situated in different sociocultural conditions. In short the text must be able to . . . "decontextualize" itself in such a way that it can be "recontextualized" in a new situation -- as accomplished precisely in the act of reading."

Though Ricoeur refers to "the text" and "the act of reading" only in passing here, the notions of text, textuality and reading are of central importance in his thinking, and by no means to be understood as strictly literary. That music can also be regarded as "text" in this sense has been argued by ethnomusicologist/ semiotician Kofi Agawu in an essay entitled, significantly enough, "African Music As Text." I want to quote extensively from this important document:

Traditional African music is not normally described as contemplative art (see Euba, "The Potential of African Traditional Music"). It is thought rather to be functional. Functional music drawn from ritual, work, or play is externally motivated. Thus funeral dirges sung by mourners, boat-rowing songs sung by fishermen, lullabies performed by mothers, and songs of insult traded by feuding clans: these utilitarian musics are said to be incompletely understood whenever analysis ignores the social or "extra-musical" context. This music is then contrasted with elite or art music, whose affinities with European classical music are for the most part unmediated. Such contemplative music is not tied to an external function. Although it is in principle consumed in a social setting, it demands nothing of its hearers save contemplation, meditation, an active self-forgetting. According to this distinction, then, analysis of traditional music--which is sometimes generalized to encompass all African music--must always take into account the particular activity to which the music is attached, whereas analysis of European music, unburdened of attachment to external function, can concentrate on the music itself, its inner workings, the life of its tones.
.......
The point that African music can be legitimately listened to still needs to be made in view of long-standing views linking music and dance. Gerhard Kubik, for example, has argued that African music constitutes a "motional system" (9-46). The implication is that physical negotiation of various musical patterns is as important as the sonic trace itself. The resulting emphasis on context easily leads to a denial that listening to African music without extramusical props can be a rewarding or even legitimate activity. In explaining why "it is a mistake `to listen' to African music," John Chernoff draws attention to a number of supplementary texts that prop up the music sound (75). Putting aside the one sidedness of the argument (isn't European music as much a "motional system" as any other? isn't it similarly rooted in the extramusical?), we might point out that at no point is traditional music turned over to the deaf; at no point do its critics abandon aural engagement. The functional attribute, then, is restrictive, and it is part of my purpose here to urge its elimination...[in Research in African Literatures, 6/22/2001.]

Agawu’s basic argument is that African musics can legitimately be treated as “texts, texts that demand (and deserve) to be contemplated.” That is, just as we profitably read a book, so may we profitably listen to a recording of African (or, by implication, any) music, because the “sonic trace” itself already represents so much of importance. I emphasize the word “represents” because Agawu, as a semiotician, is sensitive to the processes in and through which signifieds can contain so much more than what is immediately apparent in their signifiers.

I'm editing a bit here, since I just now realized that the last part of the above sentence might be better stated thus: the processes in and through which signifiers can contain so much more than what is immediately apparent in their signifieds.

(to be continued)