Homo naledi and the braided stream of humanity. It’s miscegenation all the way down.

Above are reconstructions of three famous fossil hominins by paleo-artist John Gurche. From left to right: Australopithecus afarensis (Lucy), Homo erectus (Turkana boy) and recently discovered Homo naledi (paper). The Homo naledi announcement last month of fossils discovered in a cave in South Africa was a blockbuster for human origins. Even better, two more big origins papers were published in the past two weeks. The first on the oldest modern human teeth found in China, pushing the date of modern humans in Asia from ~45,000 to 80-100,000 years ago (paper). The second on the genome of a 4,500 year old Ethiopian (paper).

In announcing the Homo naledi discovery, and reflecting on how early species from our genus Homo have distinct mixtures of ape and modern features, Lee Berger has suggested we use the metaphor of a braided stream. Here’s Berger at the 1:50 mark of this Nova/PBS TV special (my transcription):

So imagine in your mind a glacier at the top of a valley, and what happens is at it melts, it creates many many rivulets. And some of them are large and some of them are small. And they all move off down the valley. And almost inevitably, at the end of that valley is going to be a lake. Of which some, maybe the majority but not all, are contributing too. I think we need to look at these species we’re finding as almost individual channels in a  braided stream. It’s clear they have something to do with the end population, that’s us, billions of human beings alive today, but it’s hard to tell which ones are the most responsible for us being here.

In contrast, the traditional bush (or tree) metaphor of evolution has lineages diverging but never reticulating back. Yes, the braided stream has been pushed pretty hard with this announcement. But I absolutely love this metaphor. So great. Though I’ll concede part of the reason may be because it’s about sex. Between species. In old school terms: miscegenation. What makes this metaphor even better is braided streams have a fractal quality, retaining character at different scales. You can zoom in close and see small rivulets flowing apart and together underfoot (closer kin interbreeding), or rise up so high you see powerful river channels sweeping out and back together (different species interbreeding). This fractal quality gives the metaphor an applicability across vastly different time scales and geographies. So let’s attempt just that.

Several million years ago – African Origin of Humanity’s Braided Stream

Let me quote John Hawks in an interview about Homo naledi:

Maybe 15 years ago — certainly when I was a student — we thought that the transition from early hominids that we call Australopiths to our genus, homo, was an event that was accompanied by a change in the way that we walk, a change in the way that we deal with culture. We started making artifacts, we developed technology. That change was tied to a change in diet. We got smaller teeth and we ate higher quality things. We shared food, we began to relate to groups of people in tighter way with more cooperation. And that meant that the brain got bigger. And all of these things were closely linked to each other. And the fossil record — the archaeological record — seemed to support that. Looking at Homo naledi, its skull looks a lot like the skulls of other early homo types of creatures — Homo habilis and Homo erectus — except it’s too small. The brain hasn’t increased in size the way that humans have. It walked like us. It’s feet are the most human aspects of its anatomy. It manipulated objects. We see that their hands were apparently well made to make things. And its teeth looked like they were made for a high quality diet – they’re more human-like in some respects than Homo erectus, Homo habilis for certain.

But other parts of its anatomy are strikingly primitive. It’s hips are Australopith-like, they look nothing like the hips we’ve attributed to homo before. Its brain is so small it tells us that these thing that we thought went together as a package actually happened separately, and maybe in separate branches of our evolutionary tree.

It is very striking the brain size of Homo naledi (~500 cc) is not just much smaller than a modern human (~1200 cc), or smaller than Homo erectus (~900 cc). It’s only slightly larger than a chimpanzee (~400 cc).  Though to be clear the range on the sizes quoted here have enough variability there’s some partial overlap (see links for details).

The cave chamber where Homo naledi was discovered is so arduously difficult to get into it was left undisturbed by animals. So how did the bodies get in? From the same Hawks interview: “the most logical explanation is that Homo naledi did this — that they were bringing bodies to this place.”  And “from the science side, we don’t know what they were thinking. The bodies were not intentionally placed in graves. So, we try as much as we can to divorce it from human cultural cases that we immediately imagine. We know that these bodies were taken there by someone who knew this chamber and who did it repeatedly. That implies a kind of cultural knowledge that was shared with the group, and it implies some sort of recognition of mortality — maybe that’s all it implies.” Perhaps what this combination of social behavior with their dead, modern hands and feet, but small brains implies is modern human social behavior evolved prior to larger brains. Or “maybe in separate branches of our evolutionary tree,” that is, maybe Homo sapiens was birthed as a mosaic species.

100,000 years ago – Out of Africa

The new paper on human teeth found in Daoxian, China pushes the date of modern humans in Asia to 80-100,000 years ago, far earlier than known dates in Europe. This means anatomically modern humans had far longer to mix with archaic hominins than previously suspected. Biologically, most mammal species separated by a few million years of evolution can interbreed. Wolves and Dogs. Polar bears and Brown bears. For humans: “any hominine species whose ancestries diverged less than 4 ma previously may well have been able to produce hybrid offspring.” So yes, eventually the genetic distance between species becomes large enough that a non-reticulated bush or tree metaphor makes perfect sense. But all species in our genus Homo diverged recently enough they’re potentially willing and viable mates, as is apparent to anyone who has stayed out at a bar past closing time.

Homo erectus, dating from 1.9 million to 70,000 years ago, was the first widely dispersed hominin in Eurasia. Others left Africa in turn. So when modern humans left Africa, there were plenty of hominins to choose from. Melanesians have 4-6% Denisovan ancestry. Tibetans can thank genetic introgression from Denisovans for their high altitude adaptations. Eurasians have 1-4% Neanderthal ancestry. And while I want to be a bit cautious quoting twitter instead of papers, Adam Van Arsdale recently passed along a John Hawks estimate that living humans sample about 50% of Neanderthal DNA in “short rare bits”. Complicates the notion of whether to characterize this as extinction or assimilation.

Let’s jump forward to 10,000 years ago, where we have richer data.

10,000 years ago – Agriculture and the Neolithic Revolution

The last glacial period ended about 12,500 years ago. The climate got warmer and the northern parts of Eurasia  became habitable. Shortly thereafter we see the rise and spread of agriculture as shown in the map above. The Neolithic revolution greatly increased population sizes and was coupled with a massive shift in environment, moving people from a nomadic to farming existence. From a theoretic population genetics point of view, both of these (larger populations, novel environments) should increase the tempo of selection and evolution. See for example The 10,000 Year Explosion. Some canonical examples of novel genetics arising in the past 10,000 years are lactase persistence (adult ability to digest milk), changes to amylase genes (impacting digestion of starchy crops), malaria resistance, blue eye color, and Eskimo’s enhanced ability to tolerate a marine diet high in omega-3 fatty acids.

Ancient European DNA has also shed light on the successive waves of people populating Europe. Carl Zimmer summed up recent findings as:

today’s Europeans descend from three groups who moved into Europe at different stages of history. The first were hunter-gatherers who arrived some 45,000 years ago in Europe. Then came farmers who arrived from the Near East about 8,000 years ago. Finally, a group of nomadic sheepherders from western Russia called the Yamnaya arrived about 4,500 years ago.

It’s worth noting that Basque genetics are unique in being primarily descended from the first wave of farmers, avoiding the Yamnaya pulse. For our purposes what I want to emphasize is how these multiple waves overran each other, but then blended. Plus highlight something Razib Khan has written well about, how surprisingly recent the European phenotype is. This is best illustrated with eye and skin color, given their well understood genetics. Blue eyes originated in dark skinned hunter gatherers, while light skin arose in farmers. The light skinned/blue eyed phenotype is thus a relatively recent blend. And with the Yamnaya influx coming only 4,500 years ago, we can consider the European phenotype (i.e., white people) to be on the order of 4000 years old. It’s quite daunting to realize the oldest pyramids predate the existence of people we’d recognize as Europeans.

3,000 years ago – Eurasians to Africa

While it’s been known for a while that Eurasians had gene flow into Africa, the first ancient African genome, recovered from a 4,500 year old Ethiopian, is a big deal. It provides the first baseline of African genetics prior to the influx of Eurasian DNA. The most surprising finding is the Khoisan Peoples have 6-7% Eurasian ancestry. Previously that population had been considered free of Eurasian ancestry because the Khoisan genetic lineage split off from other groups 100,000 years ago.

[Update Jan 2016: Backmigration paper has bioinformatics error]

Just like Europe, large demographic changes swept across Africa with the advent of agriculture. The Bantu Expansion started 3000 years ago and reached South Africa by 300 AD. So it’s quite possible Eurasian admixture tagged along with this wave. As in the case of Europe, it’s fascinating to see how recent most of the demographics we take for granted in Africa really are.

Today – The Braided Stream Flows On

Above is a map of a three way split showing European, African and Native American genetic ancestry across Brazil. If we could skip back in time to Europe 4000 years ago, we might see a very similar three way, not yet stable mix. Though the European version would be composed of European hunter gatherers, near east first farmers, and Yamnaya (hat tip). Once again we see a pattern of streams splitting and rejoining across geography and time. In fact right now in Europe the Syrian refugee crisis has raised questions about the capacity of societies to absorb streams of refugees. Personally, I tend to be cautiously optimistic about immigrants. As it ever was, so it shall be. Yet history and pre-history show populations getting swamped by migrants is not something to be casually dismissed. Think North America after 1492. Population migration taps into some of humanity’s deepest hopes and fears. Politicians underestimate its power at their peril.

One big caveat with the braided stream metaphor is it’s just that: a metaphor. Not reality. One alternative, which I would associate with Yuval Harari’s recent book Sapiens (review), is that anatomically modern humans weren’t all that different from other hominins until about 70,000 years ago. At that time in East Africa humans made a cognitive leap forward, for the first time producing complex language, art, and hypersocial coordination. They then expanded out of Africa and overran all archaic hominins. This is more of a bush/tree model of human evolution. And clearly there’s some truth in it. But that framing has lost a bit of popularity recently as the complexity of the rise and fall and blend of ancient populations is becoming more apparent through DNA analysis. It also rubs against genetic data showing Khoisan Peoples (mentioned above), split off far earlier than the proposed date of the great leap forward. What’s exciting is the science of ancient DNA is moving at a rapid clip, and some of the questions now before us will get answered. Many no doubt in quite surprising ways.

One more caution. There’s a naive interpretation of the braided stream metaphor claiming genetic differences between individuals and populations are too small to matter. We’re all blended. Sure, it’s quite false to claim the white race (or any other modern race for that matter) is primordial and pure. Humans, like stacked turtles from cosmology, are miscegenated all the way down. Nonetheless. Even if genetic ancestry is not the only thing, it’s still a thing. And the tidal wave of modern genomics will make that more clear with every decade to come. If climate change is the inconvenient truth for conservatives in our time, perhaps genetic difference is the inconvenient truth for liberals.

After diving so deep into the past, let’s continue on, and speculate about the future. Where might humanity’s braided streams go? In Lee Berger’s telling, quoted above, they flow to a common lake. And there’s profound truth to this. Modern technology and travel have joined the world’s disparate populations back together again in an unprecendented way. But as we’ve seen, this does not mean some streams won’t stay separate, nor prevent new streams from splitting off. Way before that happens though, barring some catastrophe, we’ll have the technology to let parents select the genetic makeup of their children. In that sense we’ll all be joined to a common genetic pool, from which parents can pick and choose. And yet, wisely or unwisely, I suspect some streams will prove far more popular and fashionable than others. A GMO human monoculture might in fact get a bit tiresome, but that’s a problem for our grandchildren, not for us.