Chapter Nine: The Migration

I want at this point to move from the migrants to the great migration they are thought to have initiated, sometime between 80,000 and 60,000 years ago, with a short but probably quite hazardous trip from the Horn of Africa to what is now Yemen. According to the preponderance of the archaeological and genetic evidence, they and their descendants subsequently made their way slowly (or possibly rapidly) across the continent of Asia, following a “southern route,” along the Indian Ocean coast.

In his pathbreaking book, The Real Eve, Stephen Oppenheimer, one of the most thorough explicators and synthesizers of the “Out of Africa” model, summarizes the mitochondrial (mtDNA) evidence as follows:

[O]nly one small twig (Out-of-Africa Eve) of one branch, out of the dozen major African maternal clans available, survived after leaving the continent to colonize the rest of the world. From this small group evolved all modern human populations outside Africa.... I cannot overemphasize the importance of the simple and singular fact that only one African line accounts for all non Africans. (The Real Eve:64)

To clarify (or confuse, as the case may be), here’s a family portrait of homo sapiens, from the female (mtDNA) point of view, as seen by population geneticists (phylogenetic tree designed by “Maju,” for his blog Leherensuge) (click on image to expand):

Figure 9.1 mtDNA Tree (from Leherensuge)

That’s mitochondrial Eve, at the upper left, followed by her two daughters, LO and L1. The “twig” Oppenheimer calls “Out-of-Africa Eve” branched off from one of the major haplogroups, L3, and subsequently divided into two more twigs, which eventually grew into very thick branches indeed: M and N. As Maju’s tree indicates, literally all non-African populations belong to either the M or the N lineage.

On his website, The Journey of Mankind, Oppenheimer traces the Out of Africa route in red, starting roughly 85,000 years ago, from the horn of Africa across the Red Sea, northeast along the coast of the Arabian desert, then southeast along the coast of India, and from there following the coastline of the Indian Ocean eastward all the way through Indonesia, at that time part of the Asian mainland, and then northward, up the Pacific coast to what is now southern China.

Figure 9.2 Early Migration Paths (The Journey of Mankind)

Following the Trail

Oppenheimer estimates the entire trip, starting with the initial crossing of the original band from the Horn of Africa to Yemen and thence from the Arabian coast all the way to Indonesia, might have taken no more than 10,000 years (Oppenheimer 2004b:70-71). How does he know all this? Through evidence gleaned from indigenous peoples still living along the same pathway:

If all non-Africans share one ancestral origin [i.e. Africa]...all their trails should lead back to one point in space and time; and all the colonies, left behind en route, should hold genetic and even physical keys to who went that way. This is the case.... Along the coastline of the Indian Ocean we still find small colonies of aboriginal peoples who may be descended locally from those first beachcombers.... (156)

Let’s attempt as best we can to recreate the original “beachcomber” journey. Apparently the migrants, having made their way across the Bab el Mendab and the coast of the Arabian Peninusula, continued eastward to the Indus Delta and from there some may have proceeded up that river while the rest continued along the coast of what are now Pakistan, India and Bangla Desh. Sri Lanka would, at that time, have been part of the mainland, so they would have passed through that region as well. The question is whether or not they left colonies in place as they traveled or simply bypassed South Asia altogether, gradually progressing eastward as a mobile unit. Everything depends on how we interpret the genetic, archaeological and ethnographic evidence.

One possibility is a so-called “wave of advance,” with each generation taking root in a particular spot and subsequent generations moving on only when feeling increased pressure as the region becomes saturated. This would produce not so much isolated colonies as an ever growing population continually expanding in all possible directions. If this had been the case, then we would expect to find evidence that the Out of Africa migrants expanded into all inhabitable regions of South Asia before continuing east along the coast -- the migration would not in fact have been a coastal migration at all, but a progressive saturation of the South Asiatic peninsula, leaving numerous archaeological sites north of the coastal regions, dating from the same, very early period; and the genetic haplotypes found throughout central and northern Asia today would be far older than any found further along the coast.

However corroborating evidence for such a “wave of advance,” from either archaeology or genetics, is nowhere to be found. In fact some of the oldest evidence of the migration is located well beyond India, in Southeast Asia, New Guinea and Australia. Such a pattern strongly suggests that the migrants must, as Oppenheimer argues, have had a coastal culture, constraining them, largely, to the ocean front as they progressed. So if there was a “wave of advance” it would most likely have been confined to the coast. 

What seems most logical is a relatively rapid migration, with the migrants moving along the seacoast by boat or raft, leaving various colonies behind as they went, eventually progressing beyond South Asia to Southeast Asia, and what is now Indonesia, as well as many other places in that same general area, including Sahul. This would have left the Out of Africa immigrants strewn out in various colonies along the Indian Ocean coast, all the way from the western border of Pakistan to Indonesia, the Philippines, southern China, and also New Guinea and Australia.

A Straightforward Scenario

If the migrating populations simply marched or sailed across the full length of the Asiatic beach, leaving colonies as they went, it stands to reason that all these colonies would have been pretty similar in a great many ways. Since I went to so much trouble to characterize HMC, I could save myself further effort by characterizing the culture of their migratory descendants in essentially the same terms. Why not? If they were direct descendents of HBP and HMP, why wouldn't they be maintaining the same traditions?And there is some evidence that would tend to support such an idea.

For example, I have before me a paper titled The Muduga and Kurumba of Kerala, South India and the Social Organization of Hunting and Gathering, by George Tharakan. Both groups are characterized as hunter-gatherers, though they also practice “non-intensive” agriculture. They “hunted and gathered in rain forests” but also, like so many African Pygmy groups, “frequently interacting with outsiders.”

Gathering in the forest is of great importance both as a means of obtaining food and also as a source of raw materials. Major plant foods include tubers, edible roots, mushrooms, leaves, berries, nuts, seeds and seasonal fruits (Table 2). A considerable portion of the diet comes from roots, tubers, yams and green leaves (p. 9). . . Gathering and collecting is done both by men and women, although it is mainly a women’s activity. . .

Collection of honey is done only by men who are highly skilled in activities such as climbing big trees, driving away the bees, and also tracing the bees and locating the honey comb in the thick forest (p. 10). . . Hunting is mostly a male activity where groups of men, both agnates and affines, gather together and proceed into the forest in search of game for one or two days (p. 11).

So far this sounds very much like the African hunter-gatherer norm. There are differences as well. The Muduga and Kurumba now hunt mainly with rifles and traditionally hunted with traps rather than bows and arrows, a tradition which was, apparently, lost at some point. They do hunt with dogs, however, as do African Pygmies.

As far as core values are concerned,

game is shared equally among all those who participate in the hunt and if the game is sizable a share is given to all other households in the hamlet. Apart from the normal share, the inner meat (i.e., heart and liver) and a thigh go to the person who shot the animal. . .

Sharing and food exchange among the Muduga/Kurumba is a highly institutionalized daily activity. It is necessary that those who obtain game share with those who did not. The Muduga/ Kurumba believe that even small game should be shared among all members of the hamlet so as to avoid the craving (daham) they feel for meat. However, small game is often shared only among the members of the hunting party and their close kin. Large game animals are always widely shared (pp. 12-13 ).

All of the above is strikingly similar to what we find among African Pygmies and Bushmen, though in the case of the Mbuti, the best cuts go to the owner of the arrow rather than the one who did the shooting.

As far as kinship is concerned, we find a flexibility quite similar to what we've seen for both Pygmies and Bushmen: “Though they are patrilineal by descent, the system shows bilateral tendencies of a flexible and loosely structured system (p. 15).” Also, as with the Pygmies and Bushmen, we see no signs of warlike or competitive tendencies, but on the contrary, a willingness to peacefully share with others, even from other groups:

Among the Muduga/Kurumba it is the corporate group, the clan, which owns the land and has primary rights over its plant and animal resources. However, people from other groups and hamlets are never restricted from hunting and gathering in the clan’s territory.

Though there are significant differences as well, the most important being their clan structure and their long-term involvement with swidden agriculture, these Indian tribals do in fact seem quite close, in a great many ways, to African Pygmies and Bushmen, and thus to their HBP and HMP ancestors. And as I hope everyone reading here understands, such strong affinities should no longer simply be attributed to “hunter-gatherer-ivity.” Here we have very strong cultural evidence, consistent with the genetic evidence, of a historical continuity between the Out of Africa migrants and at least some of the many tribal peoples of India.

The picture of a continuous migration path from Africa through South Asia is reinforced by additional cultural information gleaned from the Cambridge Encyclopedia of Hunters and Gatherers, on seven Indian tribal groups, the Andamanese Islanders (Jarawa and Onge), the Birhor, Chenchu, Nayaka, Paliyan, Hill Pandaram and the Veddahs of Sri Lanka. The picture for most of these groups is roughly similar, with gathering and hunting (often with bows and arrows) supplemented with some form of swidden agriculture in all but the Andaman groups, who completely lack farming traditions. The term “egalitarian” is used in almost all instances to characterize their political and economic situation and most (though not all) of these groups are described as generally non-violent and informally communal, with no permanent leaders.

The Gap

The picture I have drawn thus far, of a relatively straightforward west-east progression, seems reasonably logical and relatively unproblematic. But there are some serious discrepancies that cannot be ignored -- it does not completely fit all the evidence. Oppenheimer offers a hint when he notes that

[c]uriously, some of the best, if not the only archaeological evidence...comes not from India, South Arabia or Africa, but from the later parts of the trail—the Malay Peninsula, New Guinea, and Australia (ibid.:156, 159).

Oppenheimer’s observation is particularly apt in view of a curious and disturbing gap in the genetic evidence, due to

the paradox of the Indian genetic picture, in which the genetic trail of the beachcombers can be detected, but the bulk of Indian subgroups...are unique to the subcontinent, especially among the tribes of the south-east. This is what we would expect for a recovery from a great disaster (The Real Eve, p. 193).

What sort of disaster could have taken place, and how could we possibly determine what it was and when it occurred? Oppenheimer finds a very interesting and possibly very important clue in “the greatest natural calamity to befall any humans, ever,” the eruption, c. 70,000-74,000 years ago, of Mount Toba, in Sumatra. The explosion was so vast it left a plume of ash over the entirety of India for approximately five years, what Oppenheimer has called a “nuclear winter,” in which almost every living thing in that area would have been wiped out (Oppenheimer 2004b:82). This is one of the very few events in prehistory that can be precisely dated and measured, since “a metres-thick ash layer is found throughout the region.”

Additional genetic evidence for such a gap is provided by an apparent discrepancy in the distribution of mtDNA haplogroups M and N (which he nicknames “Manju” and “Nasreen”):

In West Eurasia there is only Nasreen; in most of East Eurasia there are even mixtures of Nasreen and Manju, but on the east coast of India there is nearly all Manju. The latter is consistent with near local extinction following the Toba explosion with recovery only of Manju on the east coast.

Oppenheimer next proceeds to a consideration of the most important of Nasreen’s “daughters,” haplogroup R, which he dubs Rohani:

What is perhaps most interesting about the unique Indian flowerings of the Manju and Rohani clans is a hint that they represent a local recovery from the Toba disaster . . . A devastated India could have been recolonized from the west by Rohani types and from the east more by Manju types (183).

Turning to evidence from the male line, as found in the Y chromosome, he finds a yawning gap in the distribution of the Y haplogroup referred to as YAP (which he calls “Abel”):

A puzzling aspect of the Abel trail is the big gap in his distribution between West Eurasia and the Far East and, notably, his complete absence from India. That he was on the beachcombing trail is evident from the presence of Asian YAP in the Andaman Islands, Cambodia and Japan (my emphasis -- p. 188).

For Oppenheimer this disconcerting gap can best be explained

by the devastating effect of the Toba blast on the Indian subcontinent, as suggested by the geology and the maternal genetic story. . . Toba could have created a genetic bottleneck in India, which was followed by a predominant local recovery of the Seth line [haplogroup S] at the expense of his two beachcombing brothers Cain [haplogroup C] and Abel. (189-190)

The notion of a major genetic bottleneck in South Asia, occurring after an initial “beachcomber” colonization of the entirety of the Indian Ocean coast, is potentially very significant. For geneticists, a “bottleneck” represents a sudden population loss, due either to migration or some sort of disastrous event, which severely reduces the genetic diversity of the original population. This usually leads to a “founder effect” through which a new lineage, with a very different genetic profile, is born.

While Oppenheimer’s evidence could be considered somewhat out-of-date (his book appeared in 2003), results consistent with the mtDNA discrepancy he highlighted can be seen in a more recently published paper, Phylogeographic distribution of mitochondrial DNA macrohaplogroup M in India, by Suvendu Maji, S. Krithika and T.S. Vasulu (2009), where we see a map of haplogroup M distribution very similar to the one displayed on p. 181 of Oppenheimer’s book:

Figure 9.3 Haplogroup M in India (Krithika and Vasulu 2009)

Of the 13 different mainland tribal groups represented in the leftmost map, 10 are located in the Eastern and Southern portions of India and only 3 elsewhere, consistent with the distribution reported by Oppenheimer, and reflecting, for him, the effects of a devastating ancient event, centered to the East and South of the Indian subcontinent.

Still more genetic research reinforces Oppenheimer’s surprising finding. A recently published article, Correcting for Purifying Selection: An Improved Human Mitochondrial Molecular Clock, June 2009, by Pedro Soares et al., notes a genetic inconsistency between South and East Asia:

In the context of the southern-coastal-route model, it should be noted that although the distribution of haplogroup M has . . . been used to support the southern route model, the age of haplogroup M in India, at 49.4 (39.0; 60.2) kya, is significantly lower than in East Asia, at 60.6 (47.3; 74.3) kya . . . At face value, this could suggest an origin of haplogroup M in East Asia and a later migration back into South Asia, suggesting that it may have been a ‘‘pre-M’’ lineage that initially crossed South Asia. . . Southeast Asia may [therefore] be the point of origin of haplogroup M . . . Alternatively, if M dispersed with N and R through South Asia, M may have been caught up in a subsequent bottleneck and founder effect so that its age signals the time of re-expansion rather than first arrival (p. 752 -- my emphasis).

Additional evidence for the same gap comes from an unusually thorough and critical study by Richard Cordaux et al, Mitochondrial DNA analysis reveals diverse histories of tribal populations from India, 2003. While it is sometimes assumed that the tribal peoples of India are directly descended from the initial wave of Out of Africa migrants, Cordaux et al found little sign of that:

Our analyses of mtDNA variation in tribal populations of India indicate that groups in different geographic regions have different demographic histories. In general, southern tribes have reduced mtDNA diversity and mismatch distributions strongly indicative of recent bottlenecks. The distinctiveness of southern groups is also emphasized by the MDS analyses and AMOVA. However, it is difficult to distinguish from these data between old and severe bottlenecks or more recent and less severe bottlenecks (my emphasis).

The picture they found is, in fact, more consistent with a gap:

three typical east-Asian mtDNA haplogroups (A, B and F) are absent or virtually absent from non-northeast India . . . Furthermore, the fourth typical east Asian mtDNA haplogroup M has a different structure in India as compared to other Asian areas. This suggests that, although they show close affinities, the east Asian and Indian mtDNA gene pools are fairly distinct. This result is consistent with the suggestion that the east Asian and Indian mtDNA pools have been separated from each other for about 30 000 years (262 -- my emphasis).

The genetic discrepancy is paralleled by morphological and linguistic distinctions, as noted by Oppenheimer:

In Nepal, Burma and Eastern India we come across the first Mongoloid East Asian faces. These populations generally speak East Asian languages, contrasting strongly with their neighbors who mostly speak Indo-Aryan or Dravidian languages (pp. 181-182).

 While such a gap might at first seem to be a stumbling block in our understanding of the earliest migrations of modern humans out of Africa, it is also, as I see it, a potentially important clue, which could explain certain other discrepancies, some of which are so obvious as to have been taken for granted by the great majority of investigators.

A Musical “Signature”

What struck me with tremendous force while reading this portion of Oppenheimer’s book for the first time, was the truly remarkable correspondence with the musical evidence. There is in fact a surprising gap in the musical picture, which conforms very closely indeed to the gap identified in so much of the genetic, ethnographic and archaeological evidence. To understand the meaning of this gap, we must first make ourselves aware of certain striking affinities among musical traditions in many different parts of the world.

As we learned in Chapter One, one of the most distinctive aspects of Pygmy/Bushmen style (P/B) is the interlocking or interweaving of parts. In many cases the interwoven effect is produced by a practice known as “hocket,” where one part is completed by another part, with the effect of a phrase or melody tossed back and forth between two, or sometimes several, interlocking and often overlapping voices or instruments, to produce an integrated “resultant” texture. Similarly interlocking, contrapuntal and/or hocketed traditions, often featuring yodeling as well, have been found among indigenous peoples located along portions of the “Out of Africa” trail, along a path stretching from Southeast Asia, Taiwan, and the Philippines, down to Indonesia and beyond, to New Guinea and many of the smaller islands of Melanesia. 

Here, for example is a recording from Bosavi, in the New Guinea highlands, with many of the same features we’d expect to find in a Pygmy or Bushmen performance: Audio Example 17:Bosavi Yodeling (From Bosavi: Rainforest Music of Papua New Guinea, cd2, track 1, recorded by Steven Feld.) Compare with the Bushmen “Giraffe Medicine Song” we heard in Chapter One: Audio Example Eight:Giraffe Medicine Song. Here’s another example of yodeled interlock, this time from Guadalcanal, in the Solomon Islands: Audio Example 18: Women's Song (from the CD set Voices of the World, recorded and edited by Gilles Leothaud, Bernard Lortat-Jacob and Hugo Zemp).

Closely related types of instrumental hocket, similar to that of the pipe, flute, whistle, horn and percussion ensembles of Africa, have also been found in roughly the same regions, among indigenous peoples of New Guinea, the Solomon Islands, Indonesia (Flores, Bali, Java, among others), the Philippines, Vietnam, and certain other enclaves in Southeast Asia, and China. We find somewhat similar practices, both instrumental and vocal, in certain villages in Russia and other parts of Europe, both East and West, and also the Andes and other regions of Central and South America. In some cases we hear vocal-instrumental interactions of a very similar type.

Particularly telling is the distribution of the panpipe, an instrument known today through the performance of virtuoso soloists, but, among indigenous peoples, played almost exclusively in hocket, by ensembles of two or more performers. In many cases, each plays a single note on a single pipe; in other cases, a group of two or more pipes are bundled, either bound or unbound, into sets of multiple pipes, played by a single individual. While, strictly speaking, only the bundled pipes are referred to as “panpipes,” both practices are closely related.

Hocketing ensembles of pipes or panpipes are currently found among so called “aboriginals,” “tribals,” or “peasants,” of Southeast Asia, Indonesia, Melanesia, Central and South America, Africa, and Europe; archaeological records place panpipes in Polynesia, though the performing tradition seems to have died out there.

Since found in so many different parts of the world, so apparently unrelated to one another geographically, historically, or culturally, the distribution of the panpipe, and the distinctive method of ensemble hocketing associated with it, has been especially difficult to understand. In the light of the new genetic findings, however, all such ensembles seem likely to be of African origin, archaic survivals widely spread among indigenous peoples in many corners of the world, as part of a process of dissemination stemming from the original “Out of Africa” migrants (HMC).

Essentially the same explanation would account for the distribution of many other closely related types of wind ensemble, consisting of either trumpets, horns, and flutes, or certain types of hocketing percussion, organized along very similar lines, as commonly found in Southeast Asia, Indonesia, the Philippines and Melanesia, e.g., stamping tubes, slit drums, and gong ensembles, including aspects of Balinese and Javanese gamelan performance.

The independent invention of certain instruments in so many different places, unlikely as it seems, might be regarded as a remote possibility. But the strong association with hocketing just about everywhere we find them makes such an explanation far more difficult to accept. And it works both ways. If we might want to regard the presence of hocketing as an independent invention, we have to ask why it is so often associated with particular instruments, such as pipes, panpipes, horns, stamping tubes, gong ensembles, etc. — not to mention very distinctive forms of vocalizing, such as yodeling and so many of the other features associated with P/B. One might argue that a particular type of instrument or a particular type of musical interaction could be independently invented. But for both to be independently invented in tandem, in so many different places, seems quite a stretch.

There is one more turn of the screw. Throughout Southeast Asia, Indonesia, China, New Guinea, island Melanesia, and South America as well, the various instruments associated with hocket, especially panpipe ensembles, but also slit drums, and, in Indonesia, gongs and gamelans, are regularly divided into male-female pairs. The strong association with this type of symbolism in so many disparate places would seem to make independent invention especially unlikely.

Some Examples

Here’s a sampling of instrumental hocket/interlock performance both in and out of Africa, as described above, beginning with some very distinctive hocketed interactions between pipes and voices:  

Audio Example 19: Voice with Hindewhu, BaBenzele Pygmies (from Anthology Of World Music: Africa - Ba-Benzele Pygmies, recorded by Simha Arom);

Audio Example 20: Hocket with Voice and Pipe, Huli people, highland New Guinea (recorded by Artur Simon).

In this example, first heard in Chapter Seven, we hear multiple voices and pipes woven together: Audio Example 14:Ouldeme Pipes (from Cameroon:Flutes of the Mandara Mountains, recorded by Nathalie Fernando et Fabrice Marandola).

Pipes and voices are interwoven in a similar manner in the following video, as filmed in the Russian village of Plekhovo, by Olga Velitchkina: Video Example One: Russian Pipers.

For comparison, let’s listen, again, to some Mbuti Pygmy pipers, hocketing in a manner similar to both of the above, though without singing (also from Chapter Seven): Audio Example 13:Mbuti Pipers.

Next, a hocketing panpipe ensemble from the Ede, a “Montagnard” group of Vietnam: Audio Example 21: Ede Panpipes (from Anthology of Ede Music, Musique du Monde).

Another hocketing panpipe ensemble, in a somewhat different style, this time from the Are’are people of the Solomon Islands: Audio Example 22:Are’are pipers (from Solomon Islands:The Sound of Bamboo, recorded by Buaoka & Sekine).

Now for the most spectacular leap, all the way to Central America. Compare the panpipes we've been hearing with this hocketed duet from the Cuna Indians of Panama: Audio Example 23:Cuna Pipes   (Primitive Music of the World, Folkways).

Similarly organized hocketing trumpet ensembles are not uncommon in Africa. Here’s one from the Banda Linda people: Audio Example 24:Banda Linda Trumpets (from Centrafrique: Trompes Banda Linda, recorded by Simha Arom).

Trumpet ensembles can also be found in Melanesia, South America and Europe. Compare the preceding with these Ragai trumpets from, of all places, Lithuania: Audio Example 25: Tytytitit.

As I hear it, the remarkable distribution, among so many indigenous peoples, of all these highly distinctive musical elements -- vocal interlock, counterpoint, hocket, yodel, panpipes, hocketing wind and percussion ensembles, etc. -- all of a type apparently originating in Africa and strongly associated with Pygmy/Bushmen style (P/B), can be regarded as a kind of “African signature.” Wherever this particular constellation is found, we may well be hearing echoes of the ancestral culture, as filtered by HMC.

A Musical Gap

Surprisingly, however, the “African signature” is almost completely absent from both South Asia and the Near and Middle East, yet reappears in Southeast Asia and many other points farther east and south. Indeed, hardly any evidence of hocket, interlock, counterpoint, yodel, or anything else resembling P/B, can be found anywhere along the portion of the “southern migration route” now known as the Middle East, Pakistan and India. And, as with the vocal traditions, there is no sign of P/B related instrumental ensembles, of pipes, panpipes, flutes, trumpets, or percussion.

Both the strong evidence we find of the “African signature” in the music of so many indigenous groups to the east and southeast of India, and its almost total absence throughout the extent of West and South Asia, fit surprisingly well with essentially the same gap noted by Oppenheimer, as revealed by the genetic evidence.

[March 4: Since the audio examples I originally provided in this section fit better into the discussion of "Elaborate Style" in Chapter Eleven, I've cut them out and moved them there.]

 

A Cantometric Table

The situation is clearly encapsulated in the following table, drawn from the Cantometric database, showing the distribution of sung contrapuntal polyphony in the samples from Asia, Island Southeast Asia, Melanesia and Australia:

Table 9.1 Contrapuntal Polyphony in Asia and the Pacific

Note the dramatic differences in the representation of this very distinctive musical feature, with so many instances among indigenous groups in Southeast Asia, South China, Indonesia, Island Melanesia and New Guinea, and yet only a single instance among the entire 165 song sample for Tribal India, one other for the entirety of Village India, and no instances whatsoever for Northeast Asia (including Han China), the Near East, Middle East and aboriginal Australia.

Since contrapuntal polyphony is a highly distinctive feature of Pygmy/Bushmen style, along with hocket/ interlock, yodel, etc., and clear echoes of this style can be found among indigenous peoples in so many different parts of the world, it’s not difficult to conclude that some form of P/B must have been carried by the beachcombers to southern and southeastern Asia as part of a tradition already established among HMP, and from there to enclaves of traditional culture in Oceania, Europe and the Americas.

If that were the case, and we accept the theory of a steady progression of the migrant lineage through southern Asia from west to east, then we are once again forced, but this time on the basis of the musical evidence, to conclude that something drastic must have happened at a very early stage of their journey which caused them to lose this tradition, since there is no trace of it anywhere in West or South Asia. Such a sudden loss shouldn't, in itself, be surprising and in fact many such abrupt cultural shifts are known to have occurred in human history.

What is surprising is the sudden reappearance of strikingly similar practices in so many indigenous enclaves of Southeast Asia and beyond. The notion that an important tradition can suddenly be lost at a certain point in history and then, thousands of miles and God knows how many years later, be revived out of thin air, is, to say the least, difficult to reconcile with the Out of Africa model, in which all roads tend to lead backward to a single source. Interestingly enough, we find a very similar pattern reflected in yet another important cultural realm: language.

The Linguistic Evidence

Since we have no way of knowing what the original Pygmy language might have been, or if Pygmies ever had a fully formed (i.e., fully syntactic) language of their own, it has not seemed possible to formulate a meaningful hypothesis regarding the language spoken by either HBP or HMP. While many have speculated that Khoisan (the click language of the Bushmen) could represent the earliest language, there is no evidence that Pygmies ever used clicks and for that matter no reason to believe clicks were used by either HMP or any of its Asiatic, European, Oceanic or American descendants, since clicks are absent from just about all world languages outside of south and east Africa.

However, another significant, possibly highly diagnostic, aspect of language has rarely if ever been given its due. Almost every single language in SubSaharan Africa is, with very few exceptions, a tone language, i.e., a language in which differences of meaning are determined by differences in pitch. Thus, if homo sapiens originated in Africa, as is now generally believed, the first language is very likely to have been tonal, which tells us that if HBP had language at all, it would have been tonal, and the language of HMP would almost certainly have been tonal. Even if HBP didn't have a fully formed, fully syntactic language (as I suspect they may not have), then whatever vocabulary they had would most likely have contained tonally differentiated phonemes.

The extremely interesting and suggestive tone language evidence has never been fully appreciated, however, because a very puzzling gap can be seen in its worldwide distribution:

Figure 9.4 Tone Languages in blue (WALS)

Note the absolute saturation of tone languages (blue discs) in Africa; their near saturation in Southeast Asia and also Melanesia; and their almost total absence in exactly the same region where comparable gaps in the genetic and musical evidence have been found: South Asia and the Middle East. This map is taken from a remarkable website, The World Atlas of Language Structures Online (WALS), consisting of 141 maps of distinctive linguistic features, and as far as I've been able to determine from studying the contents of this site, it's very unusual for any one region to be so completely dominated by a single linguistic trait. Thus the chances of this highly structured distribution being a coincidence have to be pretty close to nil.

If modern humans originated in Africa, then what this map tells us is that the earliest languages, including the language spoken by the Out of Africa migrants (HMP), were almost certainly tonal, and if the Out of Africa migration had been straightforward, with no complicating factors, then we would have no reason to assume that all the languages of the surviving indigenous peoples along this route would not also be tonal.

The idea that language may originally have had pitch as an important phonemic and/or morphemic differentiator is, in fact, consistent with a commonly held theory that music and language may well have shared the same roots and, at least initially, developed together. If the cultivation of tonal awareness were an important aspect of both speech and music from the start, then the ubiquity of tone language in Africa is no longer a mystery. But what could have happened when HMP left Africa for points East?

When we turn our attention to the southern route, from the Bab el Mendab to the Arabian peninsula and beyond, tone languages all but completely disappear -- until we reach Southeast Asia. And once again, as with the musical evidence, we are faced with the conundrum of an important cultural tradition that suddenly vanishes, only to reappear many years later, in a completely different part of the world. Tone languages reappear, in fact, at the exact same point that our P/B-related “African signature” reappears.

Which leaves us with the third possibility: an early diffusion of both P/B and tonal language, as two different aspects of HMC, across a truly vast expanse, followed by some sort of event that would have resulted, ultimately, in the fragmentation and isolation of the cultures that maintained the original traditions, a process that would no doubt have had other consequences as well, both cultural and “racial,” and would almost certainly have left its mark on the genetic record. Since the gap I've been pointing to is so enormous, and since it is centered on a crucial region of the “southern route,” it is here that we must look for clues as to the nature of the event that produced the gap, an event that may well have been a turning point in human history. Which returns us to Stephen Oppenheimer and his “Toba” theory.