Introduction

I introduced a set of algorithms in my paper, A New Model of Computational Genomics [1], that allows you to predict ethnicity with an accuracy of about 80% using mtDNA alone. See Section 5 of [1]. It follows that mtDNA must contain information about paternal ancestry as well, since ethnicity is a combination of maternal and paternal ancestry. See Section 5 of [1] for an explanation, but for intuition, note that men could e.g., select women that have mtDNA bases in common with them, which would over time cause overlap to grow between maternal and paternal mtDNA through selection, rather than heredity, which is impossible with mtDNA, since it is inherited directly from the mother to the child, with little and possibly no mutation at all.

I’m now in the process of applying these methods generally to uncover the origins of humanity, and I believe, I just solved the problem. To begin, we have to accept the astonishing fact that many living human beings are nearly perfect matches to archaic humans. Specifically, the Roma and Papuans (and some others) are a 95% match to Heidelbergensis, the Finns and some Jews (i.e., both Sephardic and Ashkenazi) are a 70% match to Denisovans, and many populations contain people that are a 95% match to Neanderthals. See [1] generally. These percentages are given by (x) the number of matching bases divided by (y) the full mtDNA genome length (around 17,000 bases), after making use of a simple, global alignment. See Section 1.3 of [1]. Again, because the predictions are so accurate, you simply cannot argue with the methods, as they are plainly more precise than haplogroups, which instead should produce an accuracy around chance, and moreover, generally cross national boundaries (e.g., Sweden and Norway are combined into one haplogroup below).  See Section 7.1 of [1], and the map below, courtesy of Wikipedia. That is, the methods in [1] are plainly superior to traditional heredity analysis, since they can predict ethnicity at the national level, distinguishing between, e.g., Swedes, Norwegians, and Finns, despite using only mtDNA, and as a consequence, the heredity analysis should also be superior to haplogroups.

I’ve applied the techniques presented in [1] generally, with the goal of discovering the origins of humanity, and I’ve come to the conclusion that all of us descend from Denisovans. This follows from the simple fact that Neanderthals and Heidelbergensis both have a meaningful relationship to Denisovans, whereas Neanderthals and Heidelbergensis have no real meaningful relationship to each other. This is consistent with the hypothesis that both species descend from Denisovans. See Section 6.1 of [1]. I’ve also managed to assemble a fairly detailed portrait of the migration patterns of human beings globally, which is discussed below.

The Peopling of the Pacific

I recently discovered that the people of Hawaii have only minimal connections to archaic humans. Specifically, for the Hawaiians, at or above 33% of the genome, there is no relationship to the Neanderthals, Heidelbergensis, or Denisovans. Remarkably, the same is true of the Ancient Egyptians. Moreover, the Ancient Egyptians and Hawaiians have 99.7% of their genomes in common, and even at 30% of the genome (where you would expect imprecise matches), they have very similar distributions of matching ethnicities. This suggests the astonishing possibility that the Ancient Egyptians either settled Hawaii, or both the Hawaiians and Ancient Egyptians descend from the same people. I have only two Ancient Egyptian genomes, and one Hawaiian genome, but the distributions are very similar, and so I don’t think you can ignore the possibility that the Ancient Egyptians settled at least parts of the Pacific.

In any case, both populations plainly did not mate with archaic humans, in any appreciable amount, since they have so few bases in common with archaic humans. See Section 5 of [1]. For intuition, again, selection can cause two distinct mtDNA lines to converge into a new third genome, which would cause, e.g., homo sapiens to have many bases in common with archaic humans, which is the case with, e.g., many Finns, that have more bases in common with Denisovans than they should without selection. In this case, as you can see in the chart above, which shows the differences between the Hawaiians and Ancient Egyptians at 30% of the genome, the Hawaiians are closer to the Roma populations (e.g., the Iberian Roma, Russians, and Papuans) than the Ancient Egyptians are. Note that IB stands for Iberian Roma, and all of the applicable acronyms can be found at the end of [1]. The chart above is constructed by fixing a threshold match percentage, in this case 30%, and then calculating a normalized percentage within each population that are a match to e.g., the Ancient Egyptians. So, e.g., if one Norwegian is a 30% match to at least one Ancient Egyptian, then a counter is incremented for the Norwegian population, and this is done for every genome in the dataset. Those counters are then normalized to [0, 1]. The chart above shows the differences between the match distributions for the Ancient Egyptians and Hawaiians, producing a chart over the interval [-1, 1].

One sensible hypothesis is that the Hawaiians mated at least somewhat with the Papuans, causing them to converge slightly to the Roma lineage. They’re also closer to the Javanese and the people of the Solomon Islands than the Ancient Egyptians are. This makes perfect sense, since the people of Hawaii presumably came from somewhere in Asia, initially settled islands closer to Asia (e.g., Java, the Solomon Islands, and Papua), and only eventually spread to the deep Pacific. Keep in mind, the charts above were generated using a 30% match, and as a consequence, this relationship is not very strong, and instead highlights even subtle differences between the Ancient Egyptians and Hawaiians. You’ll also note that the Ancient Egyptians are somewhat closer to the Thai. Putting it all together, one sensible hypothesis is that some Thai people sailed further into the Pacific, mated with people that were already living in Java, the Solomon Islands, and Papua, and eventually formed an isolated and new people in Hawaii.

If this is true, which is consistent with the mtDNA of the Ancient Egyptians and Hawaiians, then the people of Java, the Solomon Islands, and Papua, should all be ancient and possibly archaic people, since they would have already been in the Pacific under this hypothesis. This is consistent with the fact that the Javanese and Solomon Islands people are a 95% match to some Neanderthals, suggesting the astonishing possibility that Neanderthals knew how to sail over large distances. Similarly, the people of Papua are a 96% match to Heidelbergensis, suggesting the more general thesis, that archaic humans knew how to sail. The net picture would be that the Ancient Egyptians (or their close relatives) avoided mating with archaic humans as a general matter, prior to traveling to the Pacific, and then presumably could not avoid doing so once there, eventually settling Hawaii with somewhat more archaic mtDNA than their Egyptian relatives. This is also consistent with the clear preference for avoiding archaic humans in populations such as the Icelandic, Munda, Basque, and Igbo.

The obvious question is, how did these people get to Hawaii? Unlike Papua, Java, and the Solomon Islands, Hawaii is completely isolated, and extremely far from Asia. Moreover, because the Hawaiians have no appreciable relationship to archaic humans, they must be some of the earliest humans. Logic dictates that they were probably the first humans that learned to sail, at least over distances this large, allowing them to completely avoid archaic humans in remote locations like Hawaii. Further, as these are remote islands, that are impossible to get to without a boat, it follows that the original settlers would almost certainly have had sophisticated seafaring abilities, possibly even telescopes. To understand why, just keep in mind human visibility is extremely limited, and if you simply sail out into the open Pacific, you will have no drinking water, other than what you bring with you, and as a result, any navigational errors will quickly lead to death, in just a few days. As a consequence, they could not have simply stumbled upon these islands, and instead, must have known where the islands were in advance. It’s possible they followed migratory birds, but again, birds can travel much faster than a boat, at times, and as such, if you lose the birds, you might again find yourself dead. Moreover, some birds can travel thousands of miles without rest, implying that again, unless you know where the birds are going beforehand, you could end up in the open Pacific, and therefore dead. It is instead more sensible to assume that people capable of building giant pyramids that stand to this day, were also capable of fabricating telescopes, which, because they’re presumably made of glass, and probably small, might not survive thousands of years, possibly longer, depending upon when these people actually showed up.

The Migration-Back Hypothesis

The Icelandic people, who are also geographically isolated, have no relationship with archaic humans at or above 33% of the genome. However, this is not limited to geographically isolated people, specifically, the Basque, Igbo, and Munda people, who are all closely related to each other, and the Thai, have no relationship to archaic humans at or above 33% of the genome. In contrast, the Norwegians have no relationship with archaic humans at or above 96% of the genome, and for all percentages below that, there is a non-zero relationship to Heidelbergensis. Note that this does not mean that all Norwegians are a 96% match to Heidelbergensis, and instead means that at least some Norwegians are a 96% match to Heidelbergensis. This suggests the general premise that some isolated peoples (whether geographically or culturally) have managed to avoid mating with archaic humans. However, Iceland was relatively recently populated by Nordic people around 1,000 AD, and Iceland has no indigenous people. Because the Norwegians are Nordic, just like the Icelandic, and all of these populations apparently avoided archaic humans generally, when compared to others, it follows that the migration to Iceland, by the Nordic people, could have been at least partially motivated by a desire to remain genetically isolated from archaic humans.

As a general matter, Scandinavia presents spectacular evidence for the hypothesis that some Asians migrated back from Asia, to Europe and Africa (i.e., the migration-back hypothesis). Specifically, the Swedes and Igbo are close to the Munda of India, whereas the Norwegians and Nigerians generally, are close to the Thai and the Munda. This is obviously consistent with a migration-back from Asia, in this case, with two distinct groups, making basically the same journey back, splitting into a Northern European group (the Swedes and Icelandic, on one hand, and Norwegians on the other) and an African group (the Igbo and Nigerians generally, respectively).

The obvious question is, how is it that completely morphologically distinct people are all so closely related to each other? In particular, some Norwegians and Nigerians are a 99.7% match, and many are a 99.0% match, and therefore nearly identical on the maternal line. This is completely contrary to common intuition, which is that morphologically distinct people, should have major differences in their genetics. You can argue that because we’re looking only to mtDNA, that the picture is limited, and this is undoubtedly the case. However, as noted, the methods in [1] are able to predict ethnicity with 80% accuracy, and as a consequence, it’s not rational to ignore such a high match count between populations. One sensible hypothesis is that both groups descend from a common set of ancestors, ultimately from Africa, that migrated to Asia, and then migrated back, splitting into two groups, one moving to Scandinavia, the other moving to Nigeria. I would wager that this migration occurred prior to the development of modern human appearance, and that we were anatomically modern, but still perhaps even without complexion altogether, for the simple reason that we might not have lost our body hair. This would over time, allow the two populations to develop distinct appearances, without changing mtDNA at all. Note that mtDNA can remain stable for thousands of years, and as such, the histories we’re considering are in the tens of thousands of years, and possibly longer. For the Norwegians, it would be much easier to avoid mating with other people than it would have been for the Nigerians, for the simple reason that Norway is geographically isolated, but the Basque and Igbo (also from Nigeria) show us that it is possible. Moreover, as noted, the Norwegians do have an appreciable relationship to archaic humans, whereas the Basque and Igbo do not, suggesting that cultural isolation might be a more powerful factor in avoiding archaic humans. In any case, the overall conclusion, is that some Europeans, Africans, and Asians have ancient relationships, that could predate the modern superficial distinctions between human beings, all of which is consistent with a migration-back hypothesis.

The Overall Migration History of Humanity

Putting the peopling of the Pacific in the context of the migration-back hypothesis, it seems likely that Neanderthals and Heidelbergensis had already learned to sail and settled somewhat remote locations like Papua and Java. Sometime afterwards, homo sapiens travelled North East, from Africa to Central Asia, specifically, somewhere near Kazakhstan. See Section 6.1 of [1]. Then, some of those homo sapiens travelled back to Africa (e.g., the Ancient Egyptians), whereas others travelled further East, eventually into the Pacific. This would explain the otherwise inexplicable relationships between e.g., the Ancient Egyptians and Hawaiians, and the Scandinavians, Africans, and Asians generally. That is, the migration-back hypothesis, and the theory of the peopling of the Pacific above, together form a fairly complete portrait of the macroscopic history of humanity.

Who Were the Vikings?

Although it might seem tangential to the bigger picture of history presented above, this exact same analysis, using the same populations, can be applied to the case of the Vikings, revealing a perfectly sensible answer as to who they were, that is consistent with not only genetics, but archeological evidence, historical evidence, linguistic evidence, and common sense. The answer is in my opinion, that they were a subset of the Scandinavian people that lived primarily (at least at some point) in South East Sweden, with ancient connections to the Finns. The basic intuition for this hypothesis follows from the distribution of Rune Stones, about half of which are located in Sweden. Within Scandinavia itself, Sweden has about 2000 Rune Stones, whereas Denmark has about 250, and Norway has 50. You can see in the map below, courtesy of Wikipedia, that the distribution of Rune Stones in Sweden is concentrated in South East Sweden. This is of course close to Finland, and moreover, the genetic evidence I’ll present also suggests an ancient connection to modern day Finns.

As noted above, the Jews (i.e., both Ashkenazi and Sephardic) are also related to the Denisovans. This does not imply that the Vikings were Jews, though you can’t ignore the obvious fact that the Danes, Finns, Irish, and Jews, are all closely related to the Denisovans (see the chart below).

As a matter of religion (as opposed to genetics) modern Finns are generally not Jewish (they are predominantly Christian), and moreover, in the past, they practiced a form of Paganism, not Judaism. That said, the geography of the Rune Stones suggests at least the possibility of a unique people, and moreover, there appears to be a genuine connection between the Canaanite religions and languages, and the Vikings. Specifically, the Vikings had a god named Odin, whose son was Baldr, and the Canaanites had a god named El or Adon, whose son was Baal. Moreover, there are strange similarities between the Phoenician alphabet, the Runic Alphabet, and an Ancient Finnish Alphabet known as Karelian, which is shown below, courtesy of Wikipedia. Finnish is an Uralic language, and it’s certainly not accepted theory that Phoenician is Uralic, though that’s not the point in any case. The point is instead, it seems at least plausible that the Vikings borrowed culture and language from the Middle East.

Finally, there’s at least one example in Viking art, of what might be a Hamsa (the hand with an eye in it, bottom left of center, in the image below), and possibly a Phoenician-style eye (the two eyes, one in the figure’s head, the other external, suggesting a spirt or deity). You can also see the resemblance between Karelian and Phonecian, and the scripts in the image below. That said, the Vikings were extremely well-travelled, and certainly adopted religious symbols from other cultures, in particular, Buddha. As a consequence, I don’t think we can read too much into the art, though the alphabet is plainly reminiscent of Phoenician and Ancient Finnish, which when coupled with the apparent overlap in deities, suggests a bona fide connection to the Middle East, that defined a unique group of people in Scandinavia. The image below is of a Viking artifact found in Funen, Denmark, courtesy of Wikipedia.

All of that said, none of this evidence is as compelling as the genetics itself. Specifically, as noted above, selection by one group with respect to another, can cause the two groups to converge genetically. Despite the fact that a larger portion of the Ashkenazi population is a 70% match to the Denisovans (see the chart above), it turns out that the non-Denisovan Finns are closer to the Denisovans than the non-Denisovan Ashkenazi. That is, if you look at the Finns and the Ashkenazi that are not a 70% match with the Denisovans (i.e., every individual that does not contribute to the chart above), you find that these Finns have more bases in common with the Denisovans than the non-Denisovan Ashkenazi, but the difference is slight, with an average of about 11 more bases. This is consistent with a relationship between Finns and Denisovans that is somewhat more ancient than that which is between the Ashkenazi and the Denisovans. That is, the Denisovans lived in Finland for a very long time, and as a consequence, the mtDNA of Denisovans converged significantly with the local population, and slightly more than that of the Ashkenazi. Counterintuitively, despite the fact that there are fewer living Denisovan matches in Sweden and Norway (again, see the chart above), the match between non-Denisovan Norwegians and Swedes is even stronger than the match with the Finns. This suggests more intense selection for Denisovan mtDNA, causing Norwegians (with 76 more bases in common than the Ashkenazi) and Swedes (173 more bases in common) to be even closer to Denisovans, despite having a much smaller truly Denisovan population than Finland and the Ashkenazi.

This is the intuition for the hypothesis that the Vikings were actually related to ancient Finns, and not Ancient Swedes, despite the location of the Rune Stones in Sweden. Specifically, present-day Finland has the largest percentage-wise population of Denisovans in Scandinavia (see the chart above), and so it is sensible to assume that the Denisovans in the rest of Scandinavia originated in Finland. Moreover, Denisovan remains are generally found in Asia, and not anywhere else. Common sense suggests that Denisovans migrated West from Asia to Finland, and some of them moved on to other areas in Scandinavia and elsewhere (possibly e.g., Estonia, given the language groups). Moreover, there are probably not many living people related to Denisovans in Russia (see the chart above), despite Denisovan remains in Asia generally, suggesting that the Denisovans fled West to Finland, and beyond.

Because the Vikings settled Iceland and Dublin, we should find a similar relationship to the Denisovans there. Specifically, if the Vikings were at least part Denisovan, then we should find Denisovans in Iceland and Ireland, and moreover, among those that are not a 70% match to Denisovan, we should find evidence of selection for Denisovan mtDNA. This is exactly the case, as the Irish have a significant Denisovan population (see the chart above), and moreover, though I have only one Icelandic genome, and one genome from Dublin, they are a 99.7% match to each other. Moreover, both exhibit not only strong selection for Denisovan mtDNA, but the strongest among the Scandinavians (with about 250 more bases in common with Denisovans than Ashkenazi), which is consistent with the hypothesis that the Vikings were related to modern day Finns, and therefore significantly Denisovan.

Finally, there is some genetic evidence that the connections between the Vikings and the Middle East are genetic, and not merely the result of, e.g., trade between the Middle East and the Vikings, which definitely happened. Specifically, the Dublin genome is a 99.87% match to a very large number of Sephardic Jews, and a decent number of Pashtuns. They’re also a match to the Ukrainians, but this is not surprising, given interactions between the Vikings and Ukrainians. Finally, they’re also a decent match to the Swedes, Ashkenazi, Germans, and Scotts, and while you might question the connection to the Ashkenazi, the obvious truth is that Ashkenazi Jews are very close to Northern Europeans generally. Taken as a whole, this is probably the right distribution.

The same is true to a marginally lesser extent of the Icelandic genome, which is a 99.75% match to the same populations, though this genome is a match (albeit at a lower threshold) to more Scandinavians. All of this would make perfect sense, if at least some of the Vikings were Canaanites. Specifically, if they were Phoenician, then this would explain basically everything, including their ability to build ships and sail large distances, and perhaps even the timing. The Phoenicians were conquered by the Romans around 64 BC, and the Vikings came to fruition about 1,000 years later, which leaves plenty of time. Putting it all together, I’d wager that a group of people from the Middle East somehow found their way to Scandinavia, and this set the spark to the flame that became the Vikings, and eventually modern Scandinavia.

The Code and the Dataset

All of the code you need to run these examples is linked to in [1], and the dataset is here.