24 January 2014

Origins of Myth: The Other Evidence

image credit: intrepidtravel.com
A few weeks ago I reviewed Michael Witzel's book Origins of the World's Mythologies. In that review I focussed, as Witzel does, on the evidence from comparative mythology. It's fairly obvious that if we share a grand narrative into which our myths fit, that there ought to be other evidence that follows a similar pattern. And Michael Witzel devotes a chapter of Origins exploring this evidence.

It must be said that none of the evidence is unequivocal and much of it is still rather ambiguous. More information is being added all the time. For example in the main text the book claims that there is no evidence of interbreeding between Homo sapiens sapiens and Homo Neanderthalensis. However between writing and publishing just such evidence was found indicating that all humans outside of Africa, New Guinea and Australia share a small number of genes with Neanderthals and Witzel acknowledges this in the forward. In the meantime further examples hybridisation have been discovered. (See Evolution: Trees and Braids)

To briefly recap, Michael Witzel sees a shared grand narrative in the mythologies of Europe, Asia, Oceania and the Americas (what Witzel calls "Laurasia") that is distinct from the grand narratives found in Sub-Saharan Africa, New Guinea and Australia ("Gondwana"). The Laurasian narrative involves creation of the universe from nothingness (void, chaos) via an egg or giant; the emergence of the earth from an abyssal ocean; the birth and lives of gods who fight amongst generations; the pushing apart of (father) sky and (mother) earth; the genesis and age of humanity, with an heroic age followed by our more mundane times; and finally the destruction of this universe, sometimes followed by the creation of a more perfect one. Local variations exist in abundance, but the overall story-arc seems to follow this broad outline. Gondwana mythology, by contrast, places no importance on creation.

Comparative Linguistics

The science of comparative linguistics is Witzel's home turf. He has, for example, studied the regional vations in Vedic Sanskrit and mapped the geographic areas that can be associated with various Vedic texts. He has also extensively studied loan words in Vedic showing that Munda may well have been the substrate language in Northwest India where the Vedic speakers first became firmly established. The early successes of comparative linguistics in the 18th and 19th centuries were impressive. It initially became clear, for example, that Greek, Latin and Sanskrit sprang from a common ancestor language. Germanic, Celtic, Iranian, and Slavic languages were soon added to the family, now most commonly called Indo-European. Systematic changes (such as /f/ for /p/ in Germanic as compared to Latin, part of Grimm's Law) across whole languages make it certain that they share a common ancestor and that language can be reverse-engineered on the basis of its surviving transformations. The reconstructed ancestor language is called Proto-Indo-European. The theory for example predicted three laryngeal sounds (related to our /h/) for PIE, which were not found in any living language, but were subsequently discovered in a written form of Hittite.

More recently the effort has been to try to create superfamilies by trying to locate systematic relationships across families or in reconstructed proto-languages. One result of which is a super-family called Nostratic (= "our language"). Nostratic includes Indo-European, Semitic, Sino-Tibetan, Austric (South-east Asia and Polynesia) and American Indian languages. The reconstruction is yet to find universal support amongst linguists. Witzel thinks this is in part due to artificial limits placed on the possibility of language reconstruction, but also due to the inherent difficulty of comparing so many languages at once. After all, few have the linguistic skill to do so. If we accept the proposed Nostratic language superfamily then we are immediately struck by the fact that its range is almost identical to the Laurasian mythology. For Witzel this is no coincidence, though he concedes that more work is required to establish Nostratic as a reality. But the tantalising conclusion is that the Laurasian mythology might have originally been framed in something like a Nostratic language. We are not there yet, but we have fine null hypothesis to try to disprove.


Where language provides tantalising hints at an underlying unity in the form of a common ancestor tongue, the field of genetics provides further insights into the relatedness and movements of peoples around the world. Two main subjects make up this evidence: phylogenies (or family trees) of mitochondrial DNA which is passed from mother to daughter; and phylogenies of Y chromosomes which are passed from father to son. Both types of DNA change only slowly, but at a rate we can estimate. However of all the evidence, the genetic evidence is most difficult to follow. The results of experiments are somewhat confused at times and the use of acronyms is intense.

In outline genetic studies show that all modern humans are related and that our ancestors lived in Sub-Saharan Africa. Anatomically modern humans emerged ca. 150 kya (1000's of years ago) plus or minus about 50 kya. More than one migration event seems to have taken place, but the one that succeeded in populating the earth seems to have happened about 65 kya. There are competing models for exactly how this was accomplished, but most include a small group of between 1000 and 10,000 travelling along the coastline eastwards. Sea levels were between 50m-150m lower, the figures cited vary wildly even within Origins, so evidence for this migration is mostly now covered by the ocean. But modern humans arrived in Australia (having crossed the open ocean) by about 45 kya for which we have good archaeological evidence. They continued North as well settling in China between 42-39 kya. Across Eurasia, modern humans encountered other species of hominids, but in every case survived, probably at the expense of the predecessors (and probably also interbred with them to some extent). The image below shows an up-to-date outline of the migrations based on mitochondrial DNA.

Migrations: approximate routes and times from Guha et. al (2013)

Theories on how the rest of Eurasia was settled are much less clear. There are two most likely scenarios. Firstly a second wave of migrants left Africa ca. 45 kya and went north into Western Asia and spread from there. Or secondly part of the first wave, perhaps based somewhere in West Asia, were the source of the expansion (this is what the image above shows). Eurasia being backfilled from China is also a possibility. In any case modern humans entered Europe 40-50 kya where they met, and to some extent interbred with, Neanderthals. From about 20-11 kya successive waves of migration occurred from Siberia into the Americas which were very quickly settled all the way to Tierra del Feugo. From about 5 kya inhabitants of Taiwan began the epic ocean voyages that peopled the islands of Polynesia, reaching New Zealand ca. 800 CE, but not before making contact (directly or indirectly) with South America or people from there. These dates are broadly supported by archaeological and anthropological evidence.

The dates for the settling of the Americas are important in dating the Laurasian mythology. Since the mythology is shared between all of the Americans and Eurasians the main outlines must have been in place before the first American migrations across the Beringia land bridge ca. 20 kya. This is long before any evidence of civilisation in the form of agriculture or large-scale permanent settlements. However, if Witzel is right about the implications of shared narratives then we have to accept that the narrative was in place by 20 kya at the latest.

Beringia Land-bridge from Balter (2013)

Recently a complete genome was sequenced for a child who died some 24 kya in Mal'ta, southern Siberia (Balter 2013). This boy is closely related to Amerindians, but also, surprisingly to populations in west of the Altai mountains. "Before 24,000 years ago, the ancestors of Native Americans and the ancestors of today's East Asians split into distinct groups. The Mal'ta child represents a population of Native American ancestors who moved into Siberia, probably from Europe or west Asia. Then, sometime after the Mal'ta boy died, this population mixed with East Asians. The new, admixed population eventually made its way to the Americas."

And just as with language studies the broad outlines of this evidence is consistent with Witzel's hypothesis. If a people, probably (initially) sharing a language or group of related languages, spread through Eurasia then we would expect to see evidence of relatedness in their genes. The best fit to Witzel's myth and language data involves two out-of-Africa migrations. The first, beginning ca. 65 kya, along the southern coastal route to Australia took with it the Gondwana mythology and certain mitochondrial genes. They left behind a string of languages with no connection to the languages of Laurasia. The second began around 45 kya and pushed first north and then both east and west populated Laurasia. These people spoke languages unrelated to the first migration, had a new, or at least different, mythology, and shared variants of mitochondrial genes not common amongst the first wave.


It is often commented on that although anatomically modern remains are found by about 150 kya, other features we associate with ourselves - burial, complex art, music - are first seen only about 40 kya. Witzel notes that more recent research indicates a slow build up to this so-called explosion of culture. But none-the-less there does seem to be a turning point. Most of the complex cave art begins around this time. The first evidence of musical instruments in the form of bone flutes are found. And burials with valuable items or indications of a belief in an afterlife also date to around this same period.

I've already cited the migrations to America as a latest date by which the Laurasian mythology can have been known in a more or less complete form. Since we know that the Gondwana mythology was unknown in Africa, New Guinea or Australia ca. 45 kya we have a upper limit for it's existence. It would seem then that the creation of the Laurasian mythology broadly coincides with the expansion of culture into Europe and Asia ca. 40 kya., but not later than 20 kya. 

Thus, with many caveats and hedges, we can draw out from the evidence a coherent picture in which ca. 40 kya a change took place amongst the ancestral Laurasian population that they subsequently spread, along with their genes and their language, across all of Eurasia, the Americas and the Pacific. Perhaps it is no coincidence that this story itself fits the Laurasian mythology, what we might call the Laurasian worldview. It is a peculiar feature of human beings that we are constantly seeking out new frontiers, while at the same time obsessing about our origins.

It's important to note that in the case of genetics and language we see evidence of hybridisation - shared genes across species on one hand, and loan words and regional language features on the other. The image we tend to have in our minds is a tree structure branching out from a singularity. This singularity almost certainly never happened. If you view railway lines going off to the horizon they appear to converge due to parallax error. I think we need to be aware of the historical equivalent of this. Just because we can find common factors underlying present complexity, does not mean that everything converges. History is complex at whatever magnification or scale we choose.



Balter, Michael. (2013) 'Ancient DNA Links Native Americans With Europe.' Science. 25 October 2013: Vol. 342 no. 6157 pp. 409-410. DOI: 10.1126/science.342.6157.409.
Guha P, Srivastava SK, Bhattacharjee S, Chaudhuri TK. (2013) Human migration, diversity and disease association: a convergent role of established and emerging DNAmarkers. Frontiers in Geneticsdoi: 10.3389/fgene.2013.00155. eCollection 2013.  Aug 9;4:155. 
Witzel, E. J. Michael. (2012) Origins of the World's Mythologies. Oxford University Press.

Note 27 Jul 2015

A recent large scale genetic study (scopeblog.stanford.edu) strongly suggests a single movement of people from the Siberia region into the Americans ca. 23,000 year bp.
The new genetic analysis suggests that the first immigrants to America left Siberia no more than 23,000 years ago, and then lived in isolation on the grassy plains of the Beringia land bridge for no more than 8,000 years. Those plains disappeared beneath rising seas 10,000 years ago. 
Once in the Americas, ancient Native Americans split into two major lineages about 13,000 years ago. One lineage populated both North and South America and one stayed in North America.

Related Posts with Thumbnails