Can Modern Cave Art Tell Us About Prehistoric Cave Art?
Depiction of Woolly Rhinos at Chauvet Cave, Ardèche, France (Patilpv24, SS BY-SA)
One of the most familiar and well-regarded aspects of Paleolithic human history has to be the rich collection of cave art that has been preserved through the ages. Though there is no evidence he ever said this, artist Pablo Picasso has often been quoted as arguing that "we have invented nothing" when faced with prehistoric art (Bahn, 2005). Despite the apocryphal nature of this statement, there is no denying that most reactions to this super-ancient art have been in awe, wonder, and astonishment.
But what is the nature of prehistoric cave art? Why did people paint such images? Thankfully, there is a rich ethnographic record of living people around the world who produce cave and rock art of their own, so perhaps they may have something to say? Or, perhaps, not?
For the purposes of this article, I will mainly be focusing on prehistoric cave art from the European Paleolithic, which is primarily centered around northern Spain and southern France. Prehistoric cave and rock art has been found all over the world, from the Sahara to Australia to the American Southwest and the Amazon, but for a good amount of this there is a living tradition of Indigenous peoples who have maintained connections to this art (Silberman, et al., 2012). With prehistoric European art, however, these connections have long since been severed by continuous population movements and changes in lifeways, beliefs, and practices. Just the existence of such art in Europe alone wasn't even publicly known before 1879 (David, 2017), so we have to rely on comparative methods from ethnography as well as the ongoing research of archaeologists.
The Range of Paleolithic European Cave Art
Depiction of horses and bison from Cueva de Ekain, País Vasco/Basque Country, Spain (Xabier Eskisabel, CC BY-SA 3.0)
"Rock art, sensu stricto, denotes any form of artistic activity on rock" (Silberman, et al. 2012). Much of the cave art I'm interested in here are pictographs, petroglyphs, and engravings: images painted, carved, or incised on the rock surfaces of caves.
Generally, most prehistoric art can be attributed to Homo sapiens, although arguments have been made that Neanderthals and other members of Homo practiced some forms of it (Hoffmann, et al. 2018; Capín, 2025). Paint-making kits utilizing ochre and incised lines date to between 100,000 and 55,000 years ago at South African sites, and it seems highly likely that the practice of making cave art emerged in African Homo sapiens, who carried the practice into Eurasia, Oceania, and the Americas (David, 2017). So far, the oldest definitive rock art in Eurasia comes from Indonesia on Sulawesi at 51,200 years old (Oktaviana, et al. 2014).
In Europe, Homo sapiens was making art by 37,000 years ago at the celebrated Chauvet Cave at Ardèche, France. In fact, dating work at the site revealed two phases of artistic activity, from 37-33.5 KYA (KYA = thousand years ago) and from 31-29 KYA until rock-falls began to close the entrance (David, 2017). This is the site showcased on the 2010 Werner Herzog film Cave of Forgotten Dreams: walls flowing with black charcoal depictions of horses, rhinos, and cave lions with an almost animated quality of movement, light, and shadow techniques. The Arcy-sur-Cure caves and rockshelters of Burgundy, France date to between 28-27 KYA and showcase images of multiple Ice Age species which were revealed through infrared and ultraviolet photography, as the pieces had been encased in a calcite coating since their creation (David, 2017).
Most other cave art in Europe dates to after 27 KYA and lasts in various regions until the tail-end of the Pleistocene Epoch around 11,700 years ago. Lascaux, in Dordogne, France, will be immediately familiar to students of the Ice Age, with its "Hall of the Bulls" depicting enormous aurochs, deer, and horses as well as the mysterious image of a bird-headed person and an eviscerated bison. Altamira, this time in Cantabria, Spain, features polychrome images (in red ochre and black charcoal) of many bison and more abstract forms. The Cougnac caves of Dordogne, France - like many sites- have separate chambers with a seeming focus on specific animals like mammals or birds, which were determined to have been painted by two separate human cultures separated by 10,000 years (David, 2017). The controversial "Sorcerer", a human figure with deer features, hails from caves in the Pyrenees and is accompanied by bison-people and bird-people. Though often underappreciated in some surveys of prehistoric art, at several sites like Laussel and Angles-sur-Anglin are depictions of human profiles and full bodies, at times with sex organs or even sex positions highlighted (Guthrie, 2005; White, 2003). It should also be noted that the cave art we have is often accompanied by mysterious symbols. Animals or objects may be accompanied by patterns of dots, lines, or geometric shapes, and some caves from the earliest times contain several non-figurative signs that look for-all-the-world like written letters.
Nearly all of these sites have been recovered in regions where the paintings and engravings have been safe from wind, rain, and other elemental forces; i.e. caves, whether close to the surface or a mile underground. It seems highly likely that there was visible art elsewhere on the landscape - as it is in other more recent sites - but it could also be that such art was only meant for deeper recesses, to be viewed in certain contexts, hence its ubiquity in such locations (White, 2003). It will always be difficult to test such hypotheses.
All such art is associated with recognizable archaeological cultures, representing widespread societies of Homo sapiens throughout the history of Paleolithic Europe: the Aurignacian, Gravettian, Solutrean, Magdalenian, and Epigravettian, and sometimes individual walls were painted by distinct peoples of these groups across thousands of years. What is interesting to note about these cultures is that research suggests they were not uniformly hunter-gatherers of a nomadic type. While the earliest European Sapiens seem to have lived in a similar fashion to nomadic bands, the scale of some of the recovered cave art and its associated context suggest varied levels of social organization, a factor I'll return to (Condemi & Savatier, 2019).
The Range of Modern Cave Art
First Nations Australian rock art from the Kimberley region (Claire Taylor, CC BY-SA 2.0)
There have been ethnographically-documented living peoples who create art on cave or rock walls, and today anthropologists continue to work with their descendants to understand and preserve these traditions.
Aboriginal or First Nations peoples in Australia and Tasmania have maintained distinct living rock art traditions stretching back thousands of years. In the Cape York Peninsula of eastern Australia, dating techniques show that by 6,000 years ago a number of depictions of figurative beings: one area boasted human-animal hybrids or therianthropes, another featured moths, one had shown spirit-beings, and so on. Comparisons between these and the recorded distribution of linguistic groups in the region during the early years of colonialism has suggested to anthropologists that these different artistic traditions may represent specific clan totems and, thus, territories for certain nations (David, 2017). In contrast to the relatively small region in western Europe where cave art is most prominent, Australia spans a vast geographic area and encompasses disparate peoples and rock art styles, and so cannot be viewed simplistically as just "First Nations Art" (White, 2003).
As in Europe, rock art showcases extant and long-extinct animals, including thylacines, kangaroos, giant snakes, and the marsupial-lion Thylacoleo. In Arnhem Land, on the Northern territory, is a type of X-ray style which showcases the interior anatomy of humans and other animals, including the skeleton and certain organs. The Bradshaw or Gwion Gwion style of the Kimberly Region of western Australia features a number of different depictions of human beings in various garbs and holding different objects. Artists utilized red hematite, white kaolin clay, and other natural pigments to create these works, and sometimes they even used their own blood (White, 2003; David, 2017).
Another region of the world with an on-going rock art tradition into recent times is among the Bushmen or San of southern Africa, including sites such as Brandberg in Namibia and Drakensberg in South Africa. The art here is, as elsewhere, of a wide range of styles, with some depictions of animals being hyper-realistic down to successful species identification, while others are more abstract or exaggerated. There are images of humans too, some with elaborate dress or in scenes of foraging, dance, and battle. The oldest, from Apollo 11 Cave in Namibia, date to 30 KYA (David, 2017). In more recent times, anthropologists have detected roughly two phases of art in southern Africa: prior to the arrival of Bantu-speaking agriculturalists, images are usually painted in just two colors and the style is more "restful", whereas after the arrival of new peoples the style grows "almost frantic" with increases scenes of conflict and animals depicted with less realism (Severin, 1973).
For Native Americans across the northern and southern continents, there are living traditions and connections to prehistoric rock art, with North America in particular having pieces dating back 7,000 years (Silberman, et al. 2012). In the American Southwest, we see representations of ceremonial and religious motifs. On the Great Plains, Indigenous scholars speak of "Biographical Art" which depicts historic events and figures. One remarkable find at the site of Naj Tunich cave in Guatemala preserves art from the Classic Maya Period, including glyphs, which - thanks to modern scholarship and collaboration with living Maya peoples - we can understand (David, 2017).
What Do Rock Artists Say?
Depiction of hoofed mammals at Tsodilo Hills, Botswana (Oliver Vass, CC BY-SA 3.0)
As you may have gathered, different rock art from around the world means different things to their respective artists, though some commonalities that can be drawn between sites across regions.
For some First Nations Australians, rock art conveyed spiritual meanings. Baldwin Spencer and Frank Gillen, working in 1899 among the Arrernte people, learned that artists (always men) sought out sacred places for their paintings which were associated with totems, and that only they could visit them during the context of rituals (Gunn, 1999). Other art was less restrictive, and recorded both supernatural beings & stories and natural phenomena & events, including the arrival of European ships (White, 2003). A common instance across much of Australian art is the Dreaming or Dreamtime, a shared story of creation, and demarcate the paths or songlines that ancestral beings took during these early days.
Much of Bushman art also depicts historical events and the goings on of everyday life. Different ethnic groups are drawn in their own unique styles, emphasizing distinctions between communities (Severin, 1973). There is a spiritual element to these images as well. Ethnographic accounts show that much of this rock art was produced by shamanistic artists undergoing ritual trances, depicting elements of San cosmology and beliefs. Specific animals, like the eland antelope, are given prominence for just such reasons: the eland is a symbol of the transformative state of shamans. Some of the figurative images are accompanied by geometric shapes and patterns which are thought to been the visualization of images seen under entoptic or chemical influences (White, 2003). Likewise, among Amerindian peoples, some of the rock art is said to have been produced under spiritually-induced trances, while others are, aforementioned, accounts of history (Silberman, et al. 2012).
Can we say, then, that some of the cave art produced during the European Paleolithic was a result of religious experiences for spiritual customs, or that others were made to record the world around them?
What Do Anthropologists Say?
Depiction of extinct hyena and leopard at Chauvet Cave, Ardèche, France (Carla Hufstedler, Public Domain)
The earliest research done by European scientists on cave art during the late 1800s was rather dismissive of these pieces: they were seen as lazy doodles with no aesthetic meaning by peoples with increased leisure time. By the 1900s, the understanding had now shifted that the art was a kind of sympathetic magic or fertility magic, made during rituals to control nature for successful hunts of the particular animals being depicted. By this time, ethnographic research was informing researchers, and there was much emphasis on cave art being primarily spiritual in quality. However, early anthropologists had been partially misguided; they saw First Nations Australians and the other hunter-gatherer peoples they studied as primitives in a simple state-of-nature. There were religious elements to much of the art they produced, it was true, but in their Eurocentric-framework scientists argued that their beliefs were simple and only concerned with the hunt (David, 2017).
In fact, since that time, it has fallen out of practice to rely too heavily on modern forager peoples to inform on prehistoric cave art and the people who produced them, just because their shared modes of food-production were through hunting and gathering. A quick sweep through the ethnographic record reveals diverse hunter-gatherer lifeways with innumerable particularities in religious modes and spiritual beliefs, besides, and just because one group painted animals to symbolize totems or creation myths, doesn't necessarily mean that prehistoric Europeans did the same.
I've already mentioned that Paleolithic cultures do not appear to have all been nomadic hunter-gatherers, so just like living forager groups, their social organization varied in important ways. For example, it has been proposed that the Aurignacians shared a sense of common origin in the way that ethnographically-documented tribal societes have, unlike smaller-scale nomadic bands, whereas the Gravettians and Magdalenians may have been semi-sedentary and gathered into small villages supporting even larger numbers of people than tribal groups. Such levels of organization suggest the possibility that cave art of such complexity as seen in Chauvet or Lascaux was able to be produced through specialized labor, by people who were not focused on providing food but instead could hone their artistic abilities over time (Condemi & Savatier, 2019). The takeaway here is that there was no "average prehistoric European", even at the earliest stages of Sapiens' presence there, so we shouldn't expect super similar cultures producing art under the same customs.
Nowadays, anthropologists take a holistic approach to understanding cave art, one that factors in the available archaeological evidence as well as ethnographic inferences, but without relying too heavily on one or the other.
When studying cave art of the Magdalenian or Epigravettian cultures, researchers note several factors. Images were painted or engraved in areas that were not used as living spaces; they are often found in caves that were quite dangerous to traverse - in an age of torches and lamps of oil or fat, those who navigated them knew what they were doing; patterns of wear on stalactites and acoustic-studies in some of the caves suggest that sound seemed to have played a role in art production; the in-situ natural features of caves were often used to help give art-subjects form and motion; by all accounts, depictions of hunting or wounded animals are rare to nonexistent; and, notably, what physical remains of human activity like hand or footprints are present show that all sexes were using the caves and creating these pieces, not just males (White, 2003). Such data has helped refute several classic hypotheses over the last century that you may have read about in vintage paleo-books.
As we've seen previously, some caves were visited by people belonging to distinct cultures and living across thousands of years. 10,000 years span the time of activity at the Cougnac caves of Dordogne, France. This is almost as long as the full history of humans during the Holocene Epoch; twice as long as the entire written record. There is no justifiable way to argue that the artists throughout all that time made art with the exact same meaning.
Several studies over the years have, thus, developed unique explanations for different cave sites, all with fair probability and not mutually-exclusive. As accounted by Jo Marchant in The Human Cosmos, some researchers like Norbert Aujoulat and Michael Rappenglück argued that the cave art of Lascaux may have had astronomical functions. Their analyses show that certain animals were painted in a specific order and style that represents seasonal changes associated with mating, and it is hypothesized that the large aurochs in the "Hall of the Bulls" may be a depiction of the constellation Taurus, complete with dots representing the Pleiades star cluster (Marchant, 2020).
More recent work by Bennett Bacon and colleagues draw attention to the non-figurative letter-like symbols that often accompany animal pieces, some of which date back to the dawn of European Homo sapiens. Their research indicates that there is a communicative function to these symbols, which conveyed calendrical information to various animals and their behavior as shown in the caves. In other words, a form of early writing (Bacon, 2023). In fact, at cave sites around the world, similar symbols have been found with strong similarities in regions as far apart as North America and Australia; it is implied, then, that early Homo sapiens developed this symbolic ability in Africa and brought it into Eurasia and beyond (George, 2016).
Some prehistoric cave art in Europe depicted the world as it was through an understanding of natural history. Others may have been related to spiritual matters, the particulars of which are lost to time. Some may have been produced by specialized artists, others may have been activities the whole group participated in. Others still may have completely different functions. There is no one-size-fits-all approach to studying prehistoric cave art, and modern rock artists do not have all the answers. Like much of human history, even as far back as 30,000 years ago, it's complicated.
Book References
Silvana Condemi & François Savatier - A Pocket History of Human Evolution (The Experiment, 2019)
Bruno David - Cave Art (“World of Art” series, Thames & Hudson, 2017)
R. Dave Guthrie - The Nature of Paleolithic Art (The University of Chicago Press, 2005)
Jo Marchant - The Human Cosmos: Civilization and the Stars (Penguin Random House LLC, 2020)
Timothy Severin - The Horizon Book of Vanishing Primitive Man (American Heritage Publishing Co, 1973)
Neil Asher Silberman, et al. - The Oxford Companion To Archaeology: 2nd Edition (Oxford University Press, 2012)
Randall White - Prehistoric Art: The Symbolic Journey of Humankind (Harry N. Abrams, Inc., 2003)
Paper and Article Citations
Bennett Bacon, et al. 2023. An Upper Palaeolithic Proto-writing System and Phenological Calendar (Cambridge Archaeological Journal)
Paul Bahn, 2005. A Lot of Bull? Pablo Picasso and Ice Age cave art (MUNIBE: Antropologia-Arkeologia)
Miriam García Capín, 2025. Neanderthal cave art? A proposal from cognitive archaeology (Journal of Archaeological Science: Reports)
Alison George, 2016. "Code hidden in Stone Age art may be the root of human writing" (NewScientist)
Robert G. Gunn, 1999. Spencer and Gillen's contribution to Australian rock-art studies. (Academia)
Dirk. L. Hoffmann, et al. 2018. U-Th dating of carbonate crusts reveals Neandertal origin of Iberian cave art (Science)
Adhi Agus Oktaviana, et al. 2024. Narrative cave art in Indonesia by 51,200 years ago (Nature)
Race: Understanding Human Diversity - Part 3 (Finale)
Representations of the Out-of-Africa Expansion for Homo sapiens; A shows a map of early Eurasian individuals, B breaks down these movements temporally, showing known gene flow with other human species (from Bergström, et al. 2021)
In the previous post, I discussed a number of examples of human physical and genetic traits and how they came to be. Features like skin color, hair texture, skull shape, etc were all either the products of adaptation to environment or genetic drift. In prehistory, human numbers were small and it was easier for mutations to take hold and spread across demes and populations. That said, many of these traits are underpinned by multiple genes, and there is no evidence that they all necessarily correlate with each other nor that they can serve as specific markers for human races.
The most accurate way to understand how our current human diversity came to be is through genomics, and the best way to see that is by tracking the migrations and expansions of people as far back as we can. In the traditional understanding of human diversity, biological anthropologists argued that races originated from ancestral stocks that spread across the earth and, while they were not necessarily immune to interbreeding, nonetheless remained distinct until the last 500 years or so when colonialism "connected the world together".
The last few decades of modern genetics has shattered this view. Humanity has been mobile and promiscuous from the very beginning, with the boundaries between groups often considered "races" to have been broken time and time again. We now understand that unique physical traits not only are spread by migration, but that migrating groups change as well following settlement onto new lands.
As a recent review paper explains: "Generally, the evidence suggests that selective pressures and adaptations occur at the individual, family, and community levels. Consequently, across levels there are negotiations, interactions, and feedbacks that shape migration outcomes and the trajectory of evolutionary change" (Bell, 2023).
So let's track the journey of humanity and show what modern research has to say about where we all came from...
Species Origins
A computer model showing evidence of deep connections and separations predating the origin of Homo sapiens (from Cousins, et al. 2025)
Our species, Homo sapiens, shares a recent common ancestor with two extinct species, the Neanderthals (Homo neanderthalensis) and the Denisovans (currently unclassified, perhaps Homo longi?). Morphological and genetic studies are not in ready agreement as to when these species diverged from each other; it could be as recent as ~550 thousand years ago or over a million years ago (Liu, et al. 2021; Feng, et al. 2025). It is also unclear as to where this divergence occurred. It may very well have been in Africa, but alternative locations have been suggested for Southwest Eurasia (Reich, 2018).
The Middle Pleistocene fossil record in these regions nonetheless records a great diversity in human remains. Individuals from sites like Jebel Irhoud, Omo, Florisbad, Herto, and Skhul show a mosaic of features between earlier hominins and modern Homo sapiens (Mounier & Lahr, 2019). This has suggested to many paleoanthropologists in the last few years that we may need to model the origin of Homo sapiens as "multiregional", like a braided-stream of ancestry where populations diverged, merged, and diverged again, with some of these events eventually giving rise to the last common ancestor of modern humans as we would recognize them (Bergström, et al. 2021). That is to say, humans with a globular cranium, a bony chin, and a narrow, less-flared pelvis. Computer modeling utilizing sequenced genomes has indicated a possibility that our direct ancestors may have intermingled at least twice (Cousins, et al. 2025) and constituted "weakly structured" stem-groups which did not exist as "well-defined and stable populations over hundreds of thousands of years" (Ragsdale, et al. 2023). These groups also seem to include the ancestors of the Neanderthals and Denisovans, further demonstrating how closely-related we are to these relatives of ours.
Homo sapiens in Africa
Two charts showing the degree of relatedness and admixture between Indigenous African peoples (from Fan, et al. 2023)
Once our species emerged between 300 and 200 thousand years ago, fossil and genetic evidence shows that we had a widespread distribution across the African continent. Multiple times had populations moved, met, admixed (interbred), and went their separate ways, so that it has become difficult to pinpoint the "roots" of ancestry, as is expected from multiregional evolution.
Nonetheless, there do appear to be some connections that can be made between the ancestors of Indigenous African groups. The Khoesan-speaking peoples of Southern Africa (e.g. the Ju|’hoansi and the !Xoo, whose language includes "click-consonants") appear to share significant amounts of ancestry with Central African rainforest peoples (e.g. the Mbuti and the Aka), with evidence of divergence between 285 and 150 thousand years ago (Fan, et al. 2023). This makes their ancestors part of a "sister-clade", a term from cladistics designating two closely-related groups. That said, these groups did not remain isolated for very long, quite the contrary! For example, Khoesan-speaking groups in East Africa (e.g. the Hadzabe and the Sandawe) appear to share significant amounts of ancestry with Southern African Khoesan, 71% between the Hadzabe and SA Khoesan and 38% between the Sandawe and SA Khoesan; the remaining ancestry in these groups is shared with other East African populations (Fan, et al. 2023). There is also evidence that the ancestors of the South African Khoesan may have originally emerged in East Africa and subsequently expanded their range southward, all the while contributing geneflow to their northern relatives within the last 12 thousand years.
Genetic modelling of African ancestries, with an emphasis on West African groups (from Lipson, et al. 2020)
By 200 thousand years ago at least three more ancestral groups were present across Africa alongside the Khoesan-speaking and Central African groups (Liu, et al. 2021).
The third group is a mysterious genetically-unsampled population that has been detected in the ancestries of both living West Africans and an ancient 4,500 year old individual from Mota, Ethiopia. Archaeologists have yet to have unearthed or extracted ancientDNA from an individual who hailed from this group, nor is there an exact idea of where these people originally lived (Llorente, et al. 2015).
The fourth group represents the ancestral populations of Northern Africa. These peoples seem to have shared ties with both the ancestors of living West and East African peoples, and have bestowed a significant amount of DNA to a 15 thousand-year-old individual who was excavated from the site of Taforalt in Morocco (Loosdrecht, et al. 2018). North Africa has been a bit of a blind-spot in ancientDNA studies, but a growing picture is emerging for a long-standing regional continuity between the earliest Homo sapiens in the region and the late Pleistocene peoples of the Iberomaurusian culture (Bergmann, et al. 2022; Röding, et al. 2022).
The fifth group is of strong relevance as it contributed the majority of the remaining ancestry for living Indigenous African peoples as well as the group that emerged in Eurasia ~80 thousand years ago (Liu, et al. 2021; Lipson, et al. 2020). Nonetheless, it should be clear that Africa was home to a great diversity of early peoples which evolved over time through a fluid, mosaic process in which some groups emerged, admixed, and either stayed around or became subsumed by other groups. It's a fool's errand to try to find races here.
This fifth group appears to have undergone an evolutionary radiation between 80 and 60 thousand years ago, likely in East Africa (Lipson, et al. 2020). One population spread westward and gave rise to the majority of ancestry in living Niger-Congo-speaking peoples, and appears to have effectively subsumed or pushed-out earlier populations related to both the Central African rainforest peoples and the 8-3,000 year old individuals at Shum Laka, Cameroon (Lipson, et al. 2020). From there, these groups underwent several divergences and by 6-4,000 years ago the Bantu-speaking agriculturalists had emerged in present-day Cameroon. This group would go-on to spread and settle across much of Africa south of the Sahara, today making up ~30% of the continent's population (Fortes-Lima, et al. 2023). Another population stayed in East Africa and gave rise to various agro-pastoralist (farming & herding) groups that spoke Nilo-Saharan and Afroasiatic languages: these two populations, in turn, seem to have diverged from each other between 35 and 22 thousand years ago (Fan, et al. 2023). Throughout these movements, older populations responded in various ways, through coexistence, conflict, genetic admixture, or "voting with their feet".
"Out of Africa"
Highly-simplified representation of ancestral groups following the ~80 KYA 'Out of Africa' expansion (from Vallini, et al. 2024)
Related to this fifth group is a population which has subsequently gone on to populate the rest of the planet. Known to ancientDNA researchers as "Non-Africans", they would have remained African peoples or at least have been close-neighbors when they first emerged. This is evidenced by the signature of a currently-unknown population known as Basal Eurasians, who represent the oldest known genetic split from the Non-African lineage (Reich, 2018). It is unclear just where the Basal Eurasians emerged, and when, but their genes have been found in both ancient and present-day Europeans and peoples in the Middle East and North Africa, and this hints that they were a widespread and long-lived group.
Around 80 thousand years ago there is greater evidence of a sustained presence of Homo sapiens in Southwest Asia, suggesting that an Out of Africa (henceforth OoA) population expansion has begun. Genetics picks up a significant movement of peoples between 70 and 60 thousand years ago, but it isn't until about 45 thousand years ago that large waves of people began to spread further into Europe and Asia. Various studies have proposed a "hub" in either the Arabian Peninsula or the Iranian Plateau where the OoA population settled for several thousand years, gaining genetic mutations which provided adaptational benefits for the new, cooler environment (Tobler, et al. 2023; Vallini, et al. 2024).
From there, we find traces of a smaller population expansion in the remains of two individuals, one found at Zlatý kůň, Czechia and another at Ranis, Germany, both dated to over 45 thousand years ago. This group of people does not seem to have contributed ancestry to anyone in Europe today and may represent a short-lived population in the region (Vallini, et al. 2022). Their genetic signature contains evidence of recent interbreeding with Neanderthals, and in fact has been tied to a significant admixture event that occurred between 49 and 45 thousand years ago with the OoA group and a Neanderthal population (Sümer, et al. 2024). This event has been forever sealed in the genomes of most humans today who are not of Indigenous African ancestry: as much as 2–3% of the genome comes from Neanderthals. This is an area of active research, and there is evidence of both positive and negative selection for various genes shared with us from this species of human (Iasi, et al. 2024).
While there is evidence from the archaeological record of other - sometimes substantially older - groups of Homo sapiens in Eurasia and Oceania (e.g. at Fuyan Cave in China ~100 KYA and Madjedbebe in Australia ~65 KYA), these have not yet been tied to any genetic lineages and may represent groups which contributed little if any ancestry to living peoples (Bergström, et al. 2021). As well, earlier research which seemed to find genetic evidence of an older pre-OoA movement of peoples that reached New Guinea has since been called into question (Mondal, et al. 2025).
West Eurasia
Genetic evidence of three separate sweeps into Eurasia between 45 and 38 thousand years ago (from Vallini, et al. 2022)
By 45 thousand years ago, there was a significant split from the OoA population into two groups. One population ventured eastwards towards Southeast Asia, while the other migrated westwards towards Europe. Both groups utilized Initial Upper Palaeolithic technologies, consisting of core, blade, and flaking stone tools constructed from Levallois techniques (Bellwood, 2022). The ancientDNA from several individuals of this group have been uncovered, including from Ust’-Ishim from Russia ~45 thousand years ago, Oase1 from Romania ~40 thousand years ago, Bacho Kiro from Bulgaria ~46-43 thousand years ago, and Tianyuan from China ~40 thousand years ago. They are all quite closely related to each other, even separated by such distances, suggesting that admixture between Western and Eastern Eurasian groups was maintained (Hajdinjak, et al. 2021). However, though the members of the Eastern groups survived on their own for successive generations, members of the Western groups would eventually be subsumed by a second wave of migrants around 38 thousand years ago (Vallini, et al. 2022). This Upper Paleolithic expansion, the first to feature blade tools, characterized what ancientDNA researchers call the European Early Modern Humans throughout western and eastern Europe and the Ancient North Eurasians in Siberia. We'll look at the former first.
The genetic history of glacial and post-glacial Europe (from Posth, et al. 2023)
A recent papers has elucidated the subsequent genetic history of European populations through the Paleolithic or Forager Age (Posth, et al. 2023). In essence, following the Upper Paleolithic expansion ~38 thousand years ago, Europe was populated by two groups of differing ancestries, represented by the western "GoyetQ116-1" population and the eastern "Kostenki" population. By 33 thousand years ago, these groups had begun admixing and gave rise to the "Fournol" population which utilized the Gravettian culture, well-known for their portable Venus figurines. Through the Last Glacial Maximum between 19-17 thousand years ago, a new ancestry emerged in southeast Europe which shared ancestry with the older Bacho Kiro peoples: the "Villabruna cluster". These peoples then met and admixed with descendants of the Gravettians which survived the intense cold by retreated southward to the Mediterranean. Eventually, by the end of the glacial period around 11 thousand years ago, the hunter-gatherer population of Europe had effectively homogenized through constant admixture both within and without. There was a Western Hunter Gatherer or "Oberkassel" ancestry that can be distinguished from an Eastern Hunter Gatherer or "Sidelkino" ancestry which originated from the Ancent North Eurasians (Posth, et al. 2023).
Spread of Ancient North Eurasian ancestry (naturalearthdata.com CC BY-SA 4.0)
The Ancient North Eurasians diverged from the Early Europeans around 39 thousand years ago, and are well-represented in the archaeological record by the peoples of the Mal'ta–Buret' culture, who inhabited the lands west of Lake Baikal in Siberia around 24 thousand years ago (Reich, 2018). Clearly, the nomadic movements of the Ancient North Eurasians were widespread and successful. Though as a population they no longer exist - surviving as late as 4,000 years ago in the Tarim Basin (Zhang, et al. 2021) - they seem to have frequently interbred with almost every group they met and thus contributed significant ancestry to Native Americans, Paleosiberians, Ugaric-speaking peoples, and the descendants of the Indo-European-speaking Yamnaya culture as far apart as Spain and India. Yet another example of the common threads uniting seemingly disparate groups.
Farmers and Riders in the West
As mentioned previously, genetic studies have been few and far between in North Africa as well as Southwest Asia (the Middle East), but recent work has began to fill in some gaps in our knowledge.
A 2024 paper using modern genomes has discovered a series of "soft-splits" or non-rigid genetic divergences for the ancestral populations of the Amazigh and Arab peoples (Serradell, et al. 2024). In essence, both groups belong to the Western Eurasian group and are close relatives of living Europeans but throughout their histories have incorporated extensive geneflow from both Africa and Europe over the last 18 thousand years or so. The Amazigh peoples were the first to diverge and constitute a return to the African continent where they today remain across much of North Africa; incidentally, it is generally understood that the Amazigh have some relation towards the ancestry of the Guanches of the Canary Islands (Serrano, et al. 2023). The Arabs, in contrast, share a closer relationship with European peoples and diverged around 8.6 thousand years ago. The Islamic expansions of the early Postclassical Period would, in turn, spread Arabic ancestry across Africa and Eurasia.
AncientDNA from the late glacial and postglacial periods seem to add further evidence that Southwest Asia was a hub of genetic diversity, with evidence of several groups living in relative isolation prior to the origins of agriculture around 11 thousand years ago (Reich, 2018). In particular, populations living in Anatolia (modern-day Türkiye), the Zagros Mountains of Iran, and the Levant (e.g. Palestine & Syria) share differing levels of ancestry between each other. Ancient Anatolians show evidence of admixture with Early European Hunter Gatherers, while early peoples of the Zagros show more admixture with Caucasus Hunter Gatherers (Chataigner, et al. 2024), and Levantine peoples show high levels of shared ancestry with East and North Africans.
As agricultural practices began to take root, these groups began to come into contact with each other so that, for example, farmers in Iran constituted ancestry from both Anatolian and Iranian foragers (Shinde, et al. 2019). As well, farmers began to spread to new lands in such numbers that they would significantly change the genetic landscapes wherever they went. Anatolian farmers began to spread into Europe around 9,000 to 8,500 thousand years ago, and over a period of just 3,000 years they had near-fully spread across Europe. These people, now known as "Early European Farmers", would not only end the lifeways of the older foraging groups but effectively admix them out of existence through a genetic sweep (Tsoupas, et al. 2025). According to geneticist David Reich: "Today, Early European Farmer ancestry remains widespread throughout Europe, ranging from about 60% near the Mediterranean Sea (with a peak of 65% in the island of Sardinia) and diminishing northwards to about 10% in northern Scandinavia. According to more recent studies the highest Early European Farmer ancestry found in modern Europeans ranges from 67% to over 80% in modern Sardinians, Italians, Greeks and Iberians, with the lowest ancestry found in modern Europeans ranging from 35% to 40% in modern Finns, Lithuanians and Latvians" (Reich, 2018).
One model for the spread of the Western Steppe Herders (Koba-chan, CC BY-SA 3.0)
Then, as if that weren't enough, another genetic sweep would overtake Europe. Around 5,300 years ago, the Yamnaya culture emerged on the Pontic–Caspian steppe from a combination of Caucasus, Eastern European forager, Ancient North Eurasian, and Southwest Asian farmer admixture (Lazaridis, et al. 2022). They are genetically and archaeologically well-documented as kurgan (burial-mound)-building bronze age horse-back riders and they are widely viewed as the originators of the Indo-European language family (Reich, 2018). Around 4,900 years ago, these "Western Steppe Herders" began to expand across Eurasia, moving westward into Europe, eastward into Siberia, and southward into present-day Pakistan and India. Genetic evidence suggests that, at least for Europe, the sweep commenced over a thousand-year period (Allentoft, et al. 2024), and that it involved a complex process which effectively subsumed most of the populations Early European Farmers in all but a few regions: a widely-studied example are the Basques, who have maintained genetic isolation since the European Iron Age and speak a non-Indo-European language (Flores-Bello, et al. 2021).
Similar stories played out across Africa and Eurasia. Levantine agriculturalists contributed genetic ancestry to Northern and Eastern African groups over 3,000 years ago, some of which was picked up by Khoesan pastoralists who migrated with their herds into Southern Africa (Liu, et al. 2021). One farming population distantly related to ancient Iranians spread into South Asia and contributed substantial ancestry to the founders of the Indus Valley cities (Shinde, et al. 2019). Likewise with Europe, present-day India and Pakistan experienced several genetic incursions over millennia (Narasimhan, et al. 2019; Reich, 2018). It's earliest Eastern Eurasian ancestry will be illuminated more in the next section, but it is worth mentioning that prior to 9,000 years ago the subcontinent was primarily inhabited by hunter-gatherer groups known to ancientDNA researchers as "Ancient Ancestral South Indians". When the Iranian-related farmers arrived they quickly spread agriculture throughout the region, and some researchers suspect that the Dravidian-speaking peoples derive much of their ancestry from admixture with the Indigenous groups there (Bellwood, 2022): genetically, they are known as "Ancestral South Indians". The Iranian-related farmers also spread northward into the Hindu Kush, where they eventually came into contact with the expanding Western Steppe Herders: these, in turn, are known as "Ancestral North Indians". From there, between 4,000 and 3,000 years ago, these two ancestral groups converged, with the Ancestral North Indians contributing anywhere from 39–71% ancestry in living South Asian groups today (Reich, et al. 2009).
East Eurasia
Simplified overview of the earliest diverging populations in Eurasia (from Yang, 2022)
Let's back-track a bit to the OoA movements around 45 thousand years ago. Parallel with the western Eurasian expansion, there was an eastern Eurasian movement of Initial Upper Paleolithic peoples as well. Unlike the former, the peoples of the east were not swept over genetically by a second wave but instead diversified into at least three lineages: the aforementioned Ancient Ancestral South Indians, the Australasians, and the Ancient East and Southeast Asians (Yang, 2022).
As in Southwest Asia and Europe, where Homo sapiens encountered and admixed with Neanderthals, so too did Homo sapiens in Eastern Eurasia encounter Denisovans. In contrast, however, there appears to be evidence of multiple admixture events at different times and with several highly-divergent Denisovan populations (Ongaro & Huerta-Sanchez, 2024). One event likely occurred in Southeast Asia with the ancestors of Aboriginal Australians, Negritos (a term for often small-statured Indigenous peoples in Island Southeast Asia), and the peoples of New Guinea and the Melanesian Islands, where today all these groups possess around 5% Denisovan DNA in their genome. Other events occurred on the mainland with the ancestors of South Asians, East Asians, and Native Americas, who possess 0.2% Denisovan DNA in their genomes. The contrast between the two can be explained by successive genetic sweeps with new Homo sapiens groups which have smothered the Denisovan signal.
The genetic signal for the Ancient Ancestral South Indians is very small and has only been recovered from a few individuals throughout South Asia who lived between 5,000 and 1,500 years ago (Yang, 2022). Nevertheless, they appear to have had a widespread presence on the subcontinent prior to the arrival of agriculturalists. Genetic comparisons with some isolated ethnic groups have been made, showing that the Adivasi of Sri Lanka share over half their genomes with AASI ancestry (Aragon, et al. 2025), while the Andamanese Islanders - once thought to represent members of this group - are now known to be only distant relations (Yang, 2022).
The Australasian group will be discussed in the next section, but by far the group with the largest and most complex population history are the Ancient East and Southeast Asians, who prior to 40 thousand years ago were widespread across the mainland and represented two genetic lineages.
Across Southeast Asia - which, during glacial periods, was not a series of islands but an extension of the mainland - was the "Hòabìnhian" group. These peoples utilized flaked cobble tools and inhabited regions as far apart at Yunnan, China and Laos; as well they share a common ancestry with the Andamanese Islanders. Further north along a band stretchin from South Asia to mainland Southeast Asia was a "Yunnan" group which represented the primary ancestral population of the Austroasiatic-speaking peoples like the Vietnamese and Khmer (Tagore, et al. 2021). Still further north was the "Tianyuan" group, well represented by individuals recovered from northern China 40-33 thousand years ago (specifically AR33K and Tianyuan). These groups formed the majority of Eastern Eurasian genetic diversity, alongside the eastern-most Ancient North Eurasian populations known to geneticists as "Ancient Northern Siberians" and represented by the 31 thousand-year-old individuals from the Yana River in Russia.
Even before the end of the Last Glacial Maximum, these different ancestries were admixing and diversifying. An individual found at Salkhit, Mongolia was found to share ancestry with both the Tianyuan group and Ancient Northern Siberians (Mao, et al. 2021). From 39 thousand years ago, the Tianyuan group began to diverge into a number of different populations, with one group crossing into the Japanese archipelago by at least 25 thousand years ago to become the ancestors of the Jōmon people, and another more mysterious group entering the Tibetan Plateau by 21 thousand years ago (Zhao, et al. 2009). From there, the genetic trail shows a large-split between "Northern East Asians" and "Southern East Asians" between 28 and 22 thousand years ago (Mao, et al. 2021). These source populations would effectively contribute the majority of subsequent ancestry to living East and Southeast Asian peoples, and eventually subsume many of the older groups through genetic sweeps much as had occurred in Europe.
A key admixture event occurred around 25 to 20 thousand years ago between some Ancient Northern Siberians and a group of Northern East Asians which gave rise to the "Ancient Paleosiberians", who are represented by two individuals: Kolyma1 and Ust-Kyakhta-3, who lived between 14 and 9,000 years ago. This group would give rise to both "Modern Paleosiberians" and the Amerindians or Native Americans (Yang, 2022). I'll consider the history of Indigenous Americans in a later section, but the Modern Paleosiberians will be of interest here. Over time, this group spread over a wide-range of northern Eurasia, contributing ancestry to the Chukotko-Kamchatkan-speaking peoples of far-eastern Siberia (e.g. the Chukchi and the Koryak), the Nivkh or Gilyak peoples of Sakhalin Island and mainland Russia, the Yeniseian-speaking peoples of central Siberia, and even the Uralic-speaking peoples, who spread far westward from Yakutia into northeast Europe around 4,500 years ago, today represented by the Finns, Hungarians, Estonians, and Sámi (Zeng, et al. 2025). Much of the earlier European hunter-gatherer ancestry in Finno-Scandinavia would be subsumed by this last group. Likewise in Siberia, the Paleosiberians would effectively replace the earlier Ancient Northern Siberian peoples (Sikora, et al. 2019).
Farmers and Riders in the East
Eastern Eurasia over the last 15,000 years (from Yang, 2022)
There is a very complex history of genetic exchange and migrations following the end of the last Ice Age. Both the Northern and Southern East Asians experienced continued diversification as their populations swept over the region.
By 14 thousand years ago, one group of Northern East Asians emerged in the Amur region between present-day Russia and China, and was found to have contributed to the genomes of the Ancient Paleosiberians as well. This Amur group (sometimes referred to as "Neosiberians") interacted widely with other populations in Mongolia and the Central Asian steppes from 10 thousand years ago, and underwent a large-scale population expansion which settled across much of Siberia and eventually overtook the Paleosiberian peoples from much of their lands (Wong, et al. 2017; Sikora, et al. 2019). It is widely believed that this genetic expansion was accompanied by the spread of the Altaic or Transeurasian language family, including speakers of the Tungusic (e.g. the Evenks and Manchus), Mongolic, and Turkic languages. Interactions with Indo-European-speaking groups would facilitate the introduction of horses and allow for further migrations such as those documented in recorded history.
In the Yellow River region of China, another group of Northern East Asians with close affinities to the Amur group was settling down into village communities. By 10 thousand years ago they had developed millet-based agriculture and were likely speaking a proto-form of Sino-Tibetan. As farmers have often done, they interacted widely with neighboring peoples and expanded their populations far and wide. There is evidence that one group migrated westward into the Tibetan Plateau around 5,800 years ago, acquiring an adaptive gene towards life in high mountains which had been distantly inherited from Denisovans (Xiong, et al. 2025). We know from the archaeological record that farming also originated along the Yangtze River region, south of the Yellow River, where the people there developed rice-based agriculture: until recently ancientDNA from this region was unknown. A recent 2025 paper found that these peoples represented a distinct but closely-related group to the Yellow River populations, hinting at past admixture events (Xiong, et al. 2025). The Han Chinese, today the largest ethnic group in the region, descend from admixture between groups along the lower Yellow River and the northern Yangtze River (Reich, 2018).
It has been argued that the Korean and Japanese languages also belong to the Transeurasian family, but the ancestries of these peoples seem to be as complex as others in East Asia. The modern Japanese, for example, have been shown to constitute a tripartate ancestry from the Indigenous Jōmon people, a mysterious Northern East Asian "West Liao" people which adopted agriculture and entered the archipelago from the Korean Peninsula, and a later agriculturalist influx related to the Yellow River group (Cooke, et al. 2021). Modern Koreans show close affinity to both the West Liao people and the Amur group (Sun, et al. 2023).
Some of the ancestry of the Yangtze River farmers could be traced to the Southern East Asian populations, which until now have remained undiscussed. Between 8,000 and 6,000 years ago, a number of groups had emerged from this ancestral population, including the "Guangxi/Longlin" group in southern China which is not detected in any of the living peoples in that region today (represented by the "Red Deer Cave People"), and the "Fujian" group which seems to have also lived ancestrally in southern China but played a far more long-lasting role in world history.
The spread of rice farmers into Southeast Asia (Obsidian Soul, Public Domain)
The peopling of Southeast Asia involved a multitude of southward population expansions involving hunter-gatherers and agriculturalists. The first movements involved groups related to the ancestors of Indigenous peoples in Australia and Melanesia, these being the aformentioned Hòabìnhian peoples. Some ethnic groups today, like the Punan of Borneo, show evidence of genetic continuity with these first groups (Kusuma, et al. 2023), while others like the Negritos of Malaysia show connections to the Hòabìnhian as well as Northern East Asian farmers (Aghakhanian, et al. 2022). The peoples of the Philippines consist of ancestry from at least four main movements of people into the islands, some within the last 15 thousand years (Larena, et al. 2021).
In fact, it seems that within the last 5,000 years or so, there was a large influx of agriculturalists which seems to have largely overtaken much of the original hunter gatherer populations in Southeast Asia, isolating many of them - in the case of, say, the Negritos - or fragmenting their communities - in the case of the Austroasiatic-speaking peoples (Yang, 2022; Ma, et al. 2024). The majority of this ancestry seems to descend primarily from the admixed Yellow/Yangtze farmers and the Fijian group (thus, constituting both Northern and Southern East Asian ancestries). Linguistically, these movements have been linked with the origin and spread of the Kra-Dai (e.g. Tai and Lao-speakers), Hmong-Mien, and Austronesian language families (Yang, et al. 2020).
Australia and the Pacific Islands
The peopling of Australia and Melanesia, with representative archaeological sites and hypothetical movements (from Tobler, et al. 2017)
A period of time between 49 and 45 thousand years marks the presence of Homo sapiens on the continent of Sahul, a landmass consisting of Australia, Tasmania, and New Guinea which periodically emerged during periods of low-sea levels during the last Ice Age. There are a number of earlier sites known - some going back as far as 65 thousand years - but it is unclear how accurate some of these dates are or whether they represent genetically unsampled groups which did not contribute to living Indigenous populations.
Evidence from the DNA of Aboriginal Australians confirms a continuous presence on the continent since the first settlement, following a single rapid migration from Southeast Asia that quickly reached all parts of the landmass (Tobler, et al. 2017). Even while Sahul was in existence, geneflow between Australian and New Guinean groups seems to have ceased between 40 and 25 thousand years ago - corresponding to the formation of the now-extinct Super-Lake Carpentaria - and from that time the two regions experienced marked lineage diversification (Malaspinas, et al. 2016). One group, associated with the Pama–Nyungan language family, appears to have emerged and spread across much of Australia within the last 10 thousand years or so.
In New Guinea, populations maintained genetic continuity as well, while also expanding into neighboring islands. The Bismark and Solomon Islands were settled between 43 and 39 thousand years ago, while one westward expansion settled on the island of Sulawesi and admixed with a Hòabìnhian-related group, becoming the ancestors of the Toalean peoples (Carlhoff, et al. 2021).
Movements of South Pacific wayfinders (from Ioannidis, et al. 2021)
Much more recent in time was the settlement of the remaining Pacific islands. Genetic studies have, by now, seemingly cemented an understanding that the very distant origins of the Polynesians lie in Taiwan and mainland China. Austronesian-speaking peoples inhabited southeastern coastal China and Taiwan between 6,000 and 5,000 years ago, consisting of genetic ancestry from Yangtze River farmers and Taiwanese hunter gatherers (Xiong, et al. 2025). Between 5,000 and 4,000 years ago, this population began a continuous southern expansion through the Philippines, Malaysia, and Indonesia. Some of the groups settled there and show evidence of admixture with Negritos and Austroasiatic-speaking peoples (Lipson, et al. 2014), while others sailed through the Indian Ocean and settled on Madagascar and the East African coast.
One group seems to have largely bypassed New Guinea (by then having flourishing agricultural communities in the highlands) and settled on the Bismark and Solomon Islands around 3,600 years ago. From there, the archaeological record tells us, emerged the Lapita Culture, which pioneered the design for a double-hulled outrigger canoe which could traverse more open waters (Bellwood, 2022). Just 100 years later, groups had managed to travel north to the Mariana Islands, (Pugach, et al. 2020). Between 3,200 and 2,200 years ago other members of the Lapita traveled further east, reaching as far as Fiji, Tonga, and Samoa, after which further voyages ceased for over 1,500 years: it was during this period that a recognizable Polynesian population formed. There is genetic evidence of distant contact with the peoples of New Guinea, and it appears that some individuals from that region traveled to the South Pacific on later voyages (Reich, 2018).
Beginning in the 800s AD - as the European Middle Ages were underway - there was a second great pulse of Polynesian expansion across the rest of the Pacific islands (Ioannidis, et al. 2021). The Southern Cook Islands, the Australis, the Society Islands, the Tuamotu Archipelago, and the Marquesas were all settled over a period of a few hundred years. The distant island chains of Hawaiʻi, Rapa Nui (Easter Island), and Aotearoa (New Zealand) were all settled by the 1250s. In all this time, the ingenuity of the wayfinders ensured continuious contact between several island groups, as evidenced, for example, by the Tuʻi Tonga Empire whose influence spread as far as the Tuvaluan Archipelago by the 1500s (a distance of over 1,300 kilometers). There is even evidence of contact with South America: around 1200, a group of Polynesian wayfinders reached a region roughly in present-day Colombia, admixed somehow with the Amerindians there, and transferred this ancestry to the South Pacific, long before the rest of the far eastern islands had been discovered (Ioannidis, et al. 2020).
The Americas
The first peopling of the Americas (from Willerslev & Meltzer, 2021)
As mentioned earlier, the Indigenous peoples of the Americas can ultimately trace their ancestry to an admixture event between 25 and 20 thousand years ago between Ancient Northern Siberians and a group of Northern East Asians (Yang, 2022). The region of Beringia was dry cold land around this time and seems to have supported a small but genetically-rich group of people by 18 thousand years ago, which had begun to rapidly diverge into a number of lineages which became the founding populations of the Amerindians (Moreno-Mayar, et al. 2018). One of these, termed the "Ancient Beringians" (represented by two 11,500-year-old children from Alaska), seem to have around until at least 9,000 years ago but may have been genetically subsumed by other peoples. Three other groups, currently mysterious but present in the genomes of some Central and South American Indigenous groups, seem to have also diverged around this time: "Unsampled Population A" was detected in a few living Mixe peoples, "Unsampled Population A2" was detected in northern and central Mexico (Villa-Islas, et al. 2023), while "Population Y" has been found across Amazonia and the Andes in individuals up to 10 thousand years old, but curiously retains genetic signals from both the ancient Tianyuan group and the ancestors of Australasian peoples (Ferraz, et al. 2023). There has been much controversy about Population Y, with some researchers speculating an alternative model that this group represents an earlier entry into the Americas (Raff, 2022). Further research has revealed more deeply-divergent lineages: a 5,600-year-old individual from Big Bar Lake, British Columbia and sampled DNA from historic & living members of the Blackfoot Confederacy date to this period of rapid divergence (First Rider, et al. 2024).
At present, all other Amerindian peoples sampled descend from two sister-lineages that emerged on the North American continent between 17,500 and 14,600 years ago (Moreno-Mayar, et al. 2018). One, dubbed the "Northern Native Americans" were the primary source of ancestry to the Algonquian, Salishan, Tsimshian, and Na-Dené-speaking peoples, while the "Southern Native Americans" are represented across both American continents, including in individuals associated with the widespread Paleolithic Clovis culture (Raff, 2022).
The spread of peoples across the Americas during the end of the last Ice Age proceeded with a rapidity matched by earlier times and places. Tierra del Fuego, the southern tip of South America, was settled twice by 14 thousand years ago by groups who became adapted to marine and terrestrial resources (Balentine, et al. 2022). There is even evidence of not only north-to-south migrations but south-to-north migrations as well, with genetic evidence of an Atlantic coastal expansion which settled in northeastern Brazil (Dos Santos, et al. 2022).
Movements of peoples within the Americas over the last 11,000 years (from Willerslev & Meltzer, 2021)
Some groups in the Americas retained a long-standing regional continuity in their homelands: a recent study found that the people of central Mexico - the region of Teotihuacan and the Aztec Triple Alliance - share the same genetic ancestry from pre-Hispanic to modern times with no evidence of outside genetic infusion (Villa-Islas, et al. 2023). Other groups show evidence of admixture and further population expansions: Members of the "Unsampled Population A" group expanded from Central America through much of South America from 9,000 years ago, subsuming the earlier Clovis-related genetic signatures (Raff, 2022). DNA from peoples of the California Channel Islands has been found in Andean individuals, pointing to an admixture event after 5,000 years ago (Posth, et al. 2018). The genetic blueprint of the Maya region seems to have been influenced by the movement of a group related to the Chibchan-speaking peoples of southern Central America and present-day Colombia after 5,600 years ago (Kennett, et al. 2022). In the Andes, the site of Machu Picchu has revealed that the Inka Empire consisted of a retainer community of peoples from as far as Amazonia (Salazar, et al. 2023). The peopling of the Caribbean likewise proceeded in multiple waves, with earlier hunter gatherer groups who arrived 5,000 years ago eventually being subsumed by agriculturalists - the ancestral Taíno - from northern Amazonia over the last 2,800 years (Raff, 2022).
Even after Beringia was swallowed up by rising sea-levels, the region remained a hot-spot of human diversity and expansion. There is evidence of geneflow from Modern Paleosiberian groups into North America from 11,500 years ago (Moreno-Mayar, et al. 2018), and vice-versa there was consistent Native American geneflow into northeast Asia over the last 5,000 years which contributed "non-negligible amounts" of ancestry into the ancestors of the Koryak, Chuckhi, and Itelmen peoples (Wang, et al. 2023). The peopling of the North American Arctic proceeded in a step-wise manner: around 5,500 years ago there was movement of "Paleo-Inuit" peoples from Kamchatka that made it as far as Greenland by 4,400 years ago; later on, the people of the Thule culture or "Neo-Inuit" emerged in Alaska between 4,900 and 2,700 years ago, through an admixture event with Modern Paleosiberians. This group began to spread over the Arctic from around the 1200s AD and reached Greenland after a thousand years, in that time sweeping over the genetic signal of the earlier Paleo-Inuit peoples: the living Inuit, Yup'ik, and Unangax̂ (Aleut) peoples are primarily descended from the Neo-Inuit expansion, while the DNA of the Paleo-Inuit survives in admixed form with both Unangax̂ & Na-Dené-speakers in North America and modern Paleosiberians (Raff, 2022).
So far from being an isolated landmass, the Americas remained connected in important ways to the northeast of Eurasia as well as the Pacific islands.
Wrapping-Up
If you made it this far, I offer my sincerest thanks.
As you can probably guess, the story of human migrations did not end in prehistory. It still continued well throughout the Postclassical World and on to the modern age of colonialism and the industrial revolution. It involved the creation of new ethnic groups - Latin Americans being a prime example - while also spreading different populations far and wide over the planet through every means. In the 21st Century, Homo sapiens has continued to be a monotypic, geographically-diverse but near genetically-identical species. Despite all these past popultion expansions stretching back hundreds of thousands of years, the data still stares us in the face: on average, humans are more genetically similar within populations than between populations, with the DNA underpinning supposed "racial characteristics" being miniscule proportionally and non-correlated with each other. Race is not a scientifically accurate way to understand human biological diversity.
I hope these three articles have proviced some insights into how human diversity is understood today, and changes the way you think about "biological race". Humans have a rich and varied past of continuity, population expansions, and admixture that is better understood on its own terms, rather than that be pigeon-holed into a few discrete racial roups. Doing so brings the science of biological anthropology in line with how biologists study the rest of life on Earth, and that benefits everyone.
Book References
Peter Bellwood. The Five Million Year Odyssey (Princeton University Press, 2022)
Jennifer Raff. Origin: A Genetic History of the Americas (Twelve, Grand Central Publishing, 2022)
David Reich. Who We Are and How We Got Here (Oxford University Press, 2018)
Paper & Article Citations
Farhang Aghakhanian, et al. 2022. Sequence analyses of Malaysian Indigenous communities reveal historical admixture between Hoabinhian hunter-gatherers and Neolithic farmers (Scientific Reports)
Morten E. Allentoft, et al. 2024. Population genomics of post-glacial western Eurasia (Nature)
Jose A. Urban Aragon, et al. 2025. Population histories of the Indigenous Adivasi and Sinhalese from Sri Lanka using whole genomes (Current Biology)
Elena Arciero, et al. 2025. Whole-genome sequences provide insights into the formation and adaptation of human populations in the Himalayas (Current Biology)
Christina M. Balentine, et al. 2022. Evaluating population histories in Patagonia and Tierra del Fuego, Chile, using ancient mitochondrial and Y-chromosomal DNA (American Journal of Biological Anthropology)
Adrian Viliami Bell, 2023. Selection and adaptation in human migration (Evolutionary Anthropology)
Inga Bergmann, et al. 2022. The relevance of late MSA mandibles on the emergence of modern morphology in Northern Africa (Nature Scientific Reports)
Anders Bergström, et al. 2021. Origins of modern human ancestry (Nature)
Selina Carlhoff, et al. 2021. Genome of a middle Holocene hunter-gatherer from Wallacea (Nature)
Christine Chataigner, 2024. The South Caucasus from the Upper Palaeolithic to the Neolithic: Intersection of the genetic and archaeological data (Quaternary Science Reviews)
Niall P. Cooke, et al. 2021. Ancient genomics reveals tripartite origins of Japanese populations (Science Advances)
Trevor Cousins, et al. 2025. A structured coalescent model reveals deep ancestral structure shared by all modern humans (Nature Genetics)
Andre Luiz Campelo Dos Santos, et al. 2022. Genomic evidence for ancient human migration routes along South America's Atlantic coast (Proc Biol Sci)
Shaohua Fan, et al. 2023. Whole-genome sequencing reveals a complex African population demographic history and signatures of local adaptation (CellPress)
Xiaobo Feng, et al. 2025. The phylogenetic position of the Yunxian cranium elucidates the origin of Homo longi and the Denisovans (Science)
Tiago Ferraz, et al. 2023. Genomic history of coastal societies from eastern South America (Nature Ecology and Evolution)
Dorothy First Rider, et al. 2024. Genomic analyses correspond with deep persistence of peoples of Blackfoot Confederacy from glacial times (ScienceAdvances)
André Flores-Bello, et al. 2021. Genetic origins, singularity, and heterogeneity of Basques (Cell Press)
Cesar A Fortes-Lima, et al. 2023. The genetic legacy of the expansion of Bantu-speaking peoples in Africa (Nature Portfolio)
Haechan Gill, et al. 2024. Reconstructing the Genetic Relationship between Ancient and Present-Day Siberian Populations (Genome Biology and Evolution)
Mateja Hajdinjak, et al. 2021. Initial Upper Palaeolithic humans in Europe had recent Neanderthal ancestry (Nature Portfolio)
Leonardo N. M. Iasi, et al. 2024. Neanderthal ancestry through time: Insights from genomes of ancient and present-day humans (Science)
Alexander G. Ioannidis, et al. 2021. Paths and timings of the peopling of Polynesia inferred from genomic networks (Nature)
Alexander G. Ioannidis, et al. 2020. Native American gene flow into Polynesia predating Easter Island settlement (Nature)
Douglas J. Kennett, et al. 2022. South-to-north migration preceded the advent of intensive farming in the Maya region (Nature Communications)
Pradiptajati Kusuma, et al. 2023. Deep ancestry of Bornean hunter-gatherers supports long-term local ancestry dynamics (Cell Rep)
Maximilian Larena, et al. 2021. Multiple migrations to the Philippines during the last 50,000 years (PNAS)
Iosif Lazaridis, et al. 2022. The genetic history of the Southern Arc: A bridge between West Asia and Europe (Science)
Mark Lipson, et al. 2020. Ancient West African foragers in the context of African population history (Nature)
Mark Lipson, et al. 2014. Reconstructing Austronesian population history in Island Southeast Asia (Nature Communications)
Yichen Liu, et al. 2021. Insights into human history from the first decade of ancient human genomics (Science)
M. Gallego Llorente, et al. 2015. Ancient Ethiopian genome reveals extensive Eurasian admixture in Eastern Africa (Science)
Marieke van de Loosdrecht, et al. 2018. Pleistocene North African genomes link Near Eastern and sub-Saharan African human populations (Science)
Minmin Ma, et al. 2024. Forager-farmer transition at the crossroads of East and Southeast Asia 4900 years ago (Science Bulletin)
Anna-Sapfo Malaspinas, et al. 2016. A genomic history of Aboriginal Australia (Nature)
Xiaowei Mao, et al. 2021. The deep population history of northern East Asia from the Late Pleistocene to the Holocene (CellPress)
Mayukh Mondal, et al. 2025. Resolving out of Africa event for Papua New Guinean population using neural network (Nature Communications)
J. Víctor Moreno-Mayar, et al. 2018. Terminal Pleistocene Alaskan genome reveals first founding population of Native Americans (Nature)
Aurélien Mounier & Marta Mirazón Lahr, 2019. Deciphering African late middle Pleistocene hominin diversity and the origin of our species (Nature Communications)
Vagheesh M. Narasimhan, et al. 2019. The formation of human populations in South and Central Asia (Science)
Linda Ongaro & Emilia Huerta-Sanchez, 2024. A history of multiple Denisovan introgression events in modern humans (Nature Genetics)
Cosimo Posth, et al. 2023. Palaeogenomics of Upper Palaeolithic to Neolithic European hunter-gatherers (Nature)
Cosimo Posth, et al. 2018. Reconstructing the Deep Population History of Central and South America (CellPress)
Irina Pugach, et al. 2020. Ancient DNA from Guam and the peopling of the Pacific (PNAS)
Aaron P. Ragsdale, et al. 2023. A weakly structured stem for human origins in Africa (Nature)
Carolin Röding, et al. 2022. Mugharet el'Aliya: Affinities of an enigmatic north African Aterian maxillary fragment (American Journal of Biological Anthropology)
Lucy Salazar, et al. 2023. Insights into the genetic histories and lifeways of Machu Picchu’s occupants (Science Advances)
Jose M. Serradell, et al. 2024. Modelling the demographic history of human North African genomes points to a recent soft split divergence between populations (Genome Biology)
Javier G. Serrano, et al. 2023. The genomic history of the indigenous people of the Canary Islands (Nature Communications)
Vasant Shinde, et al. 2019. An Ancient Harappan Genome Lacks Ancestry from Steppe Pastoralists or Iranian Farmers (CellPress)
Martin Sikora, et al. 2019. The population history of northeastern Siberia since the Pleistocene (Nature)
Arev P. Sümer, et al. 2024. Earliest modern human genomes constrain timing of Neanderthal admixture (Nature)
Na Sun, et al. 2023. The genetic structure and admixture of Manchus and Koreans in northeast China (Annals of Human Biology)
Debashree Tagore, et al. 2021. Insights into the demographic history of Asia from common ancestry and admixture in the genomic landscape of present-day Austroasiatic speakers (BMC Biol)
Ray Tobler, et al. 2023. The role of genetic selection and climatic factors in the dispersal of anatomically modern humans out of Africa (PNAS)
Ray Tobler, et al. 2017. Aboriginal mitogenomes reveal 50,000 years of regionalism in Australia (Nature)
Alexandros Tsoupas, et al. 2025. Local increases in admixture with hunter-gatherers followed the initial expansion of Neolithic farmers across continental Europe (ScienceAdvances)
Leonardo Vallini, et al. 2024. The Persian plateau served as hub for Homo sapiens after the main out of Africa dispersal (Nature Communications)
Leonardo Vallini, et al. 2022. Genetics and Material Culture Support Repeated Expansions into Paleolithic Eurasia from a Population Hub Out of Africa (Genome Biol Evol)
Viridiana Villa-Islas, et al. 2023. Demographic history and genetic structure in pre-Hispanic Central Mexico (Science)
Ke Wang, et al. 2023. Middle Holocene Siberian genomes reveal highly connected gene pools throughout North Asia (Current Biology)
Eske Willerslev & David J. Meltzer, 2021. Peopling of the Americas as inferred from ancient genomics (Nature)
Jianxue Xiong, et al. 2025. The genomic history of East Asian Middle Neolithic millet- and rice-agricultural populations (Cell Genomics)
Melinda A. Yang, 2022. A genetic history of migration, diversification, and admixture in Asia (Human Population Genetics and Genomics)
Melinda A. Yang, et al. 2020. Ancient DNA indicates human population shifts and admixture in northern and southern China (Science)
Tian Chen Zeng, et al. 2025. Ancient DNA reveals the prehistory of the Uralic and Yeniseian peoples (Nature)
Fan Zhang, et al. 2021. The genomic origins of the Bronze Age Tarim Basin mummies (Nature)
Mian Zhao, et al. 2009. Mitochondrial genome evidence reveals successful Late Paleolithic settlement on the Tibetan Plateau (PNAS)
Visualization of different conceptual models of race (from here)
In the previous post, I outlined the history of racial anthropology and explained what current research as to say about the diversity of humans. In brief, Homo sapiens is a monotypic species with great regional variation, but this variation cannot be quantified into discrete races and can only be understood clinally. There have not been very great geographical or biological barriers between populations and demes to have given rise to different subspecies. Genetically, humans were remarkably similar to each other, more similar than the members of other large mammalian groups are to each other. When quantified, the concept of biological race falls flat.
Still, that regional variation is important and cannot be ignored. What about the great range of skin colors, hair types, skull morphologies, and other features of our bodies? How can we explain all this, if we're so similar?
The following is a rundown of various physical and genetic traits that have been used over the centuries in attempts to classify races, and were viewed by some as essential markers of race. There has been a large body of modern research on these characteristics, as you will see below:
Skin Color
Map of Indigenous skin colors in relation to latitude, based on the work of anthropologist Renato Biasutti (Samuel, CC BY-SA 3.0)
Perhaps the most visible marker of race, skin pigmentation has been so important in delineating human groups that even today we speak of "white people" and "Black people" as if these are discrete categories with biological meaning behind them.
In the early days of modern research, anthropologists noticed a correlation between latitude & climate and the maps of Indigenous skin colors around the world. It was argued that skin color was an adaptive feature in response to ultraviolet radiation and how it affects vitamin D synthesis, the breakdown of folates, and other aspects of healthy functioning (Jablonski & Chaplin, 2000). African populations vary in dark skin-pigmentation and, so the hypothesis went, as one subset expanded into Eurasia from roughly ~80 thousand years ago they encountered new latitudes and environmental conditions unfamiliar to their continent of origin at the time. Dark skin protects from intense UV radiation but by doing so it prohibits essential production of vitamin D that is needed - for example - for aid in pregnancy and milk production. In regions where there is little UV radiation, then, humans evolved lighter skin pigmentation to allow for vital vitamin functions. As well, over time, females of all populations developed slightly lighter skin from males, ensuring more healthy pregnancies.
Seems simple enough, but further research has called into question some things in this model. For example, among African groups alone, skin pigmentation is highly diverse, and all the genes associated with lighter colors in, say, Europe and East Asia, not only originated in Africa but predate Homo sapiens by several hundred thousand years (Crawford, et al. 2017; Feng, et al. 2021). As well, for groups living in Australia or the Indian Subcontinent, they directly inherited their dark skin from Africa too and these variants also originated before our species evolved. Thus, the genomes of the small groups who left Africa ~80 KYA contained all the alleles for such vastly different skin colors within them, but those for lighter skins may not even have necessarily been selected for in response to new conditions. In a review by Andrea Hanel and Carsten Carlberg, ancient DNA recovered from the remains of Paleolithic and Neolithic (the old & new stone ages, respectively) Europeans shows that, up until the last 5,000 years or so, they were still primarily dark-skinned to some degree. How could this be if dark skin would inhibit certain health functions in the bitter-cold, low-UV landscapes of Ice Age Europe? It seems that there were adaptations to other proteins that managed to allowed humans to gain the vitamin D and folate-breakdowns necessary for life even if their skin couldn't do the job (Hanel & Carlberg, 2020).
It seems, then, that skin color is a very fluid and ever-shifting condition for our species. Ancestrally, the genus Homo varied in pigmentation as our bodies shifted from fur to thinner (but no less dense) hair on our practically naked bodies. There were many different alleles for both darker or lighter skin that fluctuated over tens of thousands of years. In fact, the lighter skin of modern Europeans and East Asians seem to have been remarkably recent developments; for Europeans within the last 5,000 years, for East Asians within the last 7,500 years (Zhang, et al. 2022). Paleoart depictions of Ice Age peoples in both these regions should, therefore, probably be more dark-skinned than is usually imagined.
On a related note, I found this article very enlightening in regards to the myth of "skin thickness" between "races".
Hair Color & Texture
Map of Indigenous hair colors (translated from ecologix, CC BY-SA 4.0)
So skin color seems to be related to natural selection. Research by Nina Jablonski and George Chaplin seem to indicate that hair color diversity has more to do with genetic drift than natural selection. This is a far more random process in which the genes in founding populations, which have split off from larger and more genetically-diverse groups, take root and spread widely among their descendants.
A look at the distribution of hair colors above is a case in point: ancestrally Homo sapiens appears to have had jet-black hair and this is our default condition, but over time as certain groups with hair-color mutations arose in parts of the world, those traits became more fixed (Jablonski & Chaplin, 2017). The more isolated some groups were, the more likely such highly-derived hair colors would remain, hence the red hair in certain Northern European peoples or the blond hair in Solomon Islanders, which are regulated by specific alleles. In fact, hair color can near accurately be predicted from genetic material because of how few genes regulate it (Branicki, et al. 2022). It should be noted, as well, that hair color as a trait can change within a person's lifetime (Kumar, et al. 2018).
Hair texture and fiber shape is highly reflective of historic human movements, with research on the morphology of hair in different human sample groups shows a greater diversity within African populations than elsewhere in the world (Franbourg, et al. 2003; Lasisi, et al. 2016). This is an area of active research in desparate need of better research tools (Lasisi, et al. 2021), but we have learned that hair texture does appear to be influenced by natural selection. The tight curly hair typical of many African populations statistically played the best role in reducing heat gain and water loss from intense solar radiation and may have been the ancestral condition for our species (Lasisi, et al. 2023).
Eye Color
Prehistoric spread of skin, hair, & eye colors in Western Eurasia, from Hanel & Carlberg, 2020
Both hair and eye color do not appear to correlate with geographic distribution nor levels of UV radiation, and seem to be better explained as products of genetic drift than strong natural selection (Jablonski & Chaplin, 2017). Dark brown eyes seem to have been the standard for the earliest members of our species, and as populations moved around the world, a select few mutations created the great host of different eye colors we see today (Hanel & Carlberg, 2020). In the earliest European groups, for example, a gene called OCA2 which codes for iris de-pigmentation is responsible for the blue-eyed condition, which, in turn, was shuffled as other groups entered Europe in successive years; the first agriculturalists to enter Europe from Southwest Asia, as one instance, introduced genes for brown eyes (Fu, et al. 2016).
Skull Shape
Principle Component Analysis (PCA) of a global sample size of human skulls, plotted by degrees of physical similarity (from Matsumura, et al. 2022)
The history of biological anthropology is full of research based on craniometry, or the study of skull morphology. Texts abound with descriptions and classifications of "dolichocephalic" and "brachycephalic" peoples and these were features used to define human races. At first glance it seems obvious that the shape of the skull would be key to "figuring out a person's race" but recent studies incorporated genetics and the fossil record have demonstrated that skull shape is only part of the story.
For starters, all living human skulls share more similarities with each other than they do to other extinct human species like Neanderthals, and as well, all living and fossil Homo sapiens skulls are more closely allied physically than they are to other humans (Relenthford, 2024). Previous attempts to find a million-year-old deep-ancestry for modern races within different fossils of Homo erectus have not stood the test of time, as the genetic signature of all living humans shows a significantly more recent ancestry than this.
Secondly, the genetics factors behind skull shape have been found to be highly complex. A recent 2023 paper found that the physiology of the cranial vault (the upper part of the skull which supports the brain) was supported by 30 different areas of the genome, and related primarily to health and development (Goovaerts, et al. 2023). Taken in full, the human cranium is underpinned by a significant amount of genetic variation that seems to be constrained in its ability to undergo evolutionary change by its connection to other genes which affect development (Martínez-Abadías, et al. 2009). That said, certain areas of the skull show more of a tendency to change and be inherited than others, like the nasal and orbital cavities and the cheek bones, and these seem to be related to environmental factors; for example, there is a marked level of evolutionary convergence in the skulls of peoples living in the freezing-cold north of Europe, North America, and Northeast Asia (Hubbe, et al. 2009). Likewise, nasal projection in certain human populations seems to be related to dry and humid conditions (Carey & Steegmann, 1981). Interestingly, the rise of agriculture in several parts of the world does not seem to have had as much of an affect on skull shape: when the skulls of foragers and farmers were compared in one study, the differences between them were significantly small and this may reflect the long pre-agricultural history of cooking, grinding, and pounding food that hunter-gatherers had first perfected over tens of thousands of years (Katz, et al. 2017).
Much like hair texture, skull shape shows a greater morphological variation within Africa, which decreases statistically the further away from the continent you go. This is to be expected, considering the recent ~80 KYA population expansion from Africa which involved successively less-genetically diverse groups the further away they migrated (Matsumura, et al. 2022; Cramon-Taubadel, 2014). Even so, as the PCA analysis above demonstrates, there is still a significant overlap in skull shape between different global populations.
Taken together, global research indicates that the total variation in skull morphology was dictated by neutral evolution (which confers no effects on reproductive fitness) while certain aspects of the skull like the nose and cheeks are affected by evolutionary adaptation to climate, and that this variation is easily inherited through genetic drift.
As a relevant aside, a 2002 study found an opposite correlation of genetic variation between craniometrics and skin color: "roughly 13% of the total diversity is among regions, 6% among local populations within regions, and 81% within local populations" for skull shape, "88% of total variation among regions, 3% among local populations within regions, and 9% within local populations" regarding skin color (Relenthford, 2002).
Dentition
Shovel vs non-shovel shaped incisors (edited from dozentist, CC BY-SA 4.0)
The human jaw does not appear to be affected by neutral evolution but instead is directly correlated to diet and the physical pressures of chewing (Harvati, et al. 2024). This is in direct contrast with the cranium, which as aforementioned showed little distinctions between foragers and farmers. Sexual dimorphism also plays a role in the shape of the mandible, once again demonstrating a human characteristic affected by multiple factors.
As far as teeth are concerned, they do appear to evolve neutrally and therefore can be linked to genetic mutations and specific population movements (Rathmaan, et al. 2023). This has led to some fascinating research regarding the historical clues left in human dentition in certain parts of the world, much of it through the work of biological anthropologists Richard Scott and Christy Turner.
Some populations in East Asia show a condition called sinodonty, where the back of their incisor teeth are shaped like shovels (see above). This trait has been traced back both genetically and in the archaeological record about 35 thousand years ago, and it appears to have been inherited by the ancestors of Amerindian peoples in a derived "hyper-sinodont" form. Other groups in Eastern Eurasia show sundadonty, which lacks shoveling, especially among Southeast Asians, Polynesians, and the historic Jōmon of Japan (Aboriginal Australians and the people of New Guinea show neither condition). These distinctions in dentition clearly are reflective of successive population movements in prehistory that can be detected in human remains:
"As with virtually all human biological variation, however, dental traits do not show presence/absence patterns that enable clear differentiation between biogeographic groups. Rather, we see varying degrees of affinity that support the inference that human biological variation exhibits a gradual change in trait frequencies across populations" (Scott, et al. 2023).
Blood Type
Splits and changes of red blood cell alleles over time between different human species (from Mazières, et al. 2025)
Red blood cell types have also historically been used in studies on race in the pre-genomic age. Recent research has focused more on DNA sequencing and full-scale genome sequencing, but in recent years interest as re-emerged on blood types.
The distribution of the A, B, and O blood types is best explained primarily through genetic drift and positive evolutionary selection (unlike neutral, positive selection confers beneficial reproductive success). Thus, when blood types are mapped onto the continents, there are very few exclusive correlations with human populations.
A recent 2025 study found evidence that the change in blood groups can be traced almost neatly to the recent Eurasian expansion of ~80 thousand years ago (Mazières, et al. 2025). Both Indigenous Africans and our nearest fossil relatives the Neanderthals and Denisovans share the same blood groups, pointing to an inherited trait from our common ancestor well beyond 400,000 years ago. As a few small groups emerged from Africa 80 KYA and experienced a rapid change in blood types in the millennia following this movement.
Genetic Traits Related to Diet & Health
Some traits discovered through genetic sequencing have been argued to be specific "racial markers" for certain groups. Given, at this point, we're trying to understand human diversity through the lense of a modern evolutionary understanding, it's better to look at these traits and markers as evidence of past adaptations and movements rather than as a "checklist" of traits for any one race.
Take, for example, variant 370A of the gene EDAR. This encodes for a protein responsible for the development of skin, hair, teeth, and sweat glands in the embryo. Around 35 thousand years ago, evidence of this variant change emerged in Eastern Eurasia and spread across the region by 19 thousand years ago, where it was subsequently picked up by both the ancestors of Native Americans and the early Polynesian wayfinders. In the descendants of all these groups, EDAR 370 A is associated with thick hair, an increase in the density of sweat glands, and sinodonty (Kamberov, et al. 2013; Mao, et al. 2021; Zhang, et al. 2022). Why did this trait emerge and spread so quickly during this time. Research is ongoing but it is notable that during the Last Ice Age East Eurasia experienced a significantly humid phase, so having thicker hair and a greater ability to sweat would have been beneficial for hunter-gatherers living in such a dry region.
Other traits are related to agricultural developments within the last 12 thousand years or so. For example, an allele variant of the gene ADH1B has been found to have emerged within East Asian groups that domesticated rice, which was often fermented into alcoholic beverages: though this process conferred nutritional benefits at the time, in recent centuries it has also been correlated with increased negative affects of alcohol consumption (Peng, et al. 2010). Lactase persistence - or the ability to drink milk into adulthood - has been shown to correlate with a deep history of cattle domestication and reliance on milk products. Contrary to the "whimsy" of certain white-supremacist groups, while lactase persistence is highly prevalant in European populations, it is also found widely in African and southern Eurasian groups (Itan, et al. 2010). Drinking milk as an adult doesn't make you "racially superior".
Another common misconception is that the "sickle-cells are a Black trait". In brief, sickle-cells are caused by a mutation that changes the shape of red blood cells from circles to crescent "sickle" shapes. This change goes some way towards the prevention of malaria as the Plasmodium parasite which causes it cannot survive for long in a sickle-celled blood stream. That said, in the modern age of increased lifespans, having sickle-cells for many years can lead to anemias and other diseases that destroy the body's ability to function. Recent genomic research has shown a single origin of the sickle-cell mutation as far back as 7,300 years ago to the Holocene Wet Phase, a period of humid, wetland conditions across the Sahara, Middle East, Southern Asia, and the Mediterranean (Shriner & Rotimi, 2018). This mutation subsequently spread across Africa and into Eurasia. So although it is true that people of African and African American descent show a high likelihood of having the sickle-cell mutation, specific cases of it merely correlate with areas that currently or historically harbor malaria parasites. South Asians, Greeks, Turks, Arabs, and other descendant populations can have the mutation too.
What about Intelligence?
I suppose a brief word must be said about IQ and intelligence.
A popular hot-button topic among academics and cranks alike is the question of whether human races differ cognitively. Much ink has been spilled on this topic - personally, I found Angela Saini's 2019 book Superior to handle the subject most completely & honestly - and there are obvious political stakes depending on who you ask.
Attempts have been made by some researchers to find specific genetic markers of intelligence, but have always come to dead-ends. Geneticist Bruce Lahn comes to mind: his work arguing that gene variants for brain size recently evolved in some human groups verses others have since failed to be replicated and debunked (Mekel-Bobrov, et al. 2007; Timpson, et al. 2007). Curiously, such individuals conducting such research always seem to have ulterior motivations.
Then there is the matter of IQ or intelligence quotients, which provide a score of cognitive intelligence based on specific tests. Previous IQ studies have been used to "prove" that certain human groups are smarter than others, but upon close examination what such IQ tests show says more about socio-political status and environmental health and stability than anything about "biological race". Poor and disenfranchised groups are often struggling with basic survival and nutrition, which affects brain health and development and thus shaped the IQ scores they may recieve. As well, IQ tests may say something about the ability of an individual to perform mathematical, scientific, or academic problems well, but will say nothing about a person's ability to perform socio-cultural or artistic functions: different "intelligences" are needed for each of these. In total, IQ gives only part of the status of an individual and betrays an ability to celebrate what makes each person whole and unique.
In any case, as the quote below from Hampshire, et al. 2012 beautifully spells out, a multitude of factors go into human intelligence, on genetic, environmental, nutritional, and social fronts. There is nothing specific in the brain in regards to intelligence that can be accurately measured and thus traced genetically through some human descent groups verses others.
In Summary
If this brief and in-exhaustive look at the modern science of human physical and genetic variation is a bit confusing, admittedly that is the point. In evolutionary biology - especially cladistics - we are so used to thinking of the acquisition of defining evolutionary traits as very straight forward. Tigers and house cats have retractable claws, but bears don't; tigers, house cats, and bears have shearing carnassial teeth, but horses don't; tigers, house cats, bears, and horses have a complex placenta, but wombats don't; and so on.
When studying human variation in the present and over time, we simply cannot look at traits in the same way. Physical traits that appear to be concrete like skin color, hair texture, and skull shape, are influenced by so many, often non-correlating factors, and our flexibility in adapting to new environments and interbreeding with each other means that such traits will never stay distinct within populations and demes for long.
Instead of thinking about human variation in terms of racial groups, which creates misleading simplistic binary models, it is better and more scientifically accurate to consider the totality of the data and understand human variation as the complex historic process is truly was.
And in the last part of this series, I hope to do just that and trace the origin and spread of Homo sapiens to where it stands today, showing a far more accurate way to understand human diversity.
Book References
Gavin Evans. Skin Deep: Dispelling the Science of Race (Oneworld Books, 2019)
Angela Saini. Superior: the Return of Race Science (Beacon Press, 2019)
G. Richard Scott & Christy G. Turner. The Anthropology of Modern Human Teeth (Cambridge University Press, 2000)
Paper & Article Citations
Wojciech Branicki, et al. 2011. Model-based prediction of human hair color using DNA variants (Human Genetics)
J W. Carey & A T. Steegmann Jr, 1981. Human nasal protrusion, latitude, and climate (American Journal of Physical Anthropology)
Noreen von Cramon-Taubadel, 2014. Evolutionary insights into global patterns of human cranial diversity: population history, climatic and dietary effects (Journal of Anthropological Sciences)
Nicholas G. Crawford, et al. 2017. Loci associated with skin pigmentation identified in African populations (Science)
Yuanqing Feng, et al. 2021. Evolutionary genetics of skin pigmentation in African populations (Human Molecular Genetics)
A. Franbourg, et al. 2003. Current Research on Ethnic Hair (J Am Acad Dermatol)
Qiaomei Fu, et al. 2016. The genetic history of Ice Age Europe (Nature)
Seppe Goovaerts, et al. 2023. Joint multi-ancestry and admixed GWAS reveals the complex genetics behind human cranial vault shape (Nature Communications)
Andrea Hanel & Carsten Carlberg, 2020. Skin colour and vitamin D: An update (Experimental Dermatology)
Katerina Harvati, et al. 2024. Comparative 3D Shape Analysis of the Iwo Eleru Mandible, Nigeria (PaleoAnthropology)
Mark Hubbe, et al. 2009. Climate Signatures in the Morphological Differentiation of Worldwide Modern Human Populations (The Anatomical Record)
Yuval Itan, et al. 2010. A worldwide correlation of lactase persistence phenotype and genotypes (BMC Evolutionary Biology)
Nina Jablonski & George Chaplin, 2017. The colours of humanity: the evolution of pigmentation in the human lineage (Philos Trans R Soc Lond B Biol Sci)
Nina Jablonski & George Chaplin, 2000. The evolution of human skin coloration (Journal of Human Evolution)
Yana G. Kamberov, et al. 2013. Modeling Recent Human Evolution in Mice by Expression of a Selected EDAR Variant (Cell)
David C. Katz, et al. 2017. Changes in human skull morphology across the agricultural transition are consistent with softer diets in preindustrial farming groups (PNAS)
Anagha Bangalore Kumar, et al. 2018. Premature Graying of Hair: Review with Updates (International Journal of Trichology)
Tina Lasisi, et al. 2023. Human scalp hair as a thermoregulatory adaptation (PNAS)
Tina Lasisi, et al. 2021. High-throughput phenotyping methods for quantifying hair fiber morphology (Nature Scientific Reports)
Tina Lasisi, et al. 2016. Quantifying variation in human scalp hair fiber shape and pigmentation (American Journal of Biological Anthropology)
Xiaowei Mao, et al. 2021. The deep population history of northern East Asia from the Late Pleistocene to the Holocene (Cell)
Neus Martínez-Abadías, et al. 2009. Heritability of human cranial dimensions: comparing the evolvability of different cranial regions (Journal of Anatomy)
Hirofumi Matsumura, et al. 2022. Global patterns of the cranial form of modern human populations described by analysis of a 3D surface homologous model (Nature Scientific Reports)
Stéphane Mazières, et al. 2025. Rapid change in red cell blood group systems after the main Out of Africa of Homo sapiens (Nature Scientific Reports)
N. Mekel-Bobrov, et al. 2007. The ongoing adaptive evolution of ASPM and Microcephalin is not explained by increased intelligence (Human Molecular Genetics)
Luca Pagani, et al. 2016. Genomic analyses inform on migration events during the peopling of Eurasia (Nature)
Yi Peng, et al. 2010. The ADH1B Arg47His polymorphism in East Asian populations and expansion of rice domestication in history (BMC Evolutionary Biology)
Frédéric B. Piel, et al. 2020. Global distribution of the sickle cell gene and geographical confirmation of the malaria hypothesis (Nature Communications)
Hannes Rathmann, et al. 2023. Inferring human neutral genetic variation from craniodental phenotypes (PNAS NEXUS)
John H. Relethford, 2023. Craniometric variation and the ancestry of modern humans (American Journal of Biological Anthropology)
John H. Relethford, 2002. Apportionment of global human genetic diversity based on craniometrics and skin color (American Journal of Physical Anthropology)
Jon Riddell, et al. 2020. Characterisation of a second gain of function EDAR variant, encoding EDAR380R, in East Asia (European Journal of Human Genetics)
Nakisa B. Sadeghi & Adewole S. Adamson, 2023. The Problematic Legacy of Skin-Thickness Measurement in Race-Based Dermatology Research (JAMA Dermatology)
G. Richard Scott, et al. 2023. Peopling of the Americas: A new approach to assessing dental morphological variation in Asian and Native American populations (American Journal of Biological Anthropology)
G. Richard Scott, et al. 2016. Sinodonty, Sundadonty, and the Beringian Standstill model: Issues of timing and migrations into the New World (Quaternary International)
Daniel Shriner & Charles N. Rotimi, 2018. Whole-Genome-Sequence-Based Haplotypes Reveal Single Origin of the Sickle Allele during the Holocene Wet Phase (The American Journal of Human Genetics)
Pontus Skoglund & Iain Mathieson, 2018. Ancient Genomics of Modern Humans: The First Decade (Annual Review of Genomics and Human Genetics)
Nicholas Timpson, et al. 2007. Comment on Papers by Evans et al. and Mekel-Bobrov et al. on Evidence for Positive Selection of MCPH1 and ASPM (Science)
Xiaoming Zhang, et al. 2022. A Late Pleistocene human genome from Southwest China (Current Biology)
A small sample of our species (Eric Haynes, CC BY-NC-ND 3.0)
The following is adapted from the extensive notes for what would have been Episode 4 of my Through Time and Clades lecture series "Humanity, a Prologue - Season 2". I still find the material I intended to cover very important and felt it would make for a good series of blog posts.
This first post will outline a brief history of the study of race and human diversity in anthropology, and then ask whether Homo sapiens can be scientifically classified in such a way. I intend the second post to outline how human diversity is understood today, and the third post will trace the evolution of human diversity through time, primarily through genetics and aDNA.
What is Race?
A race is social classification unit of humans based on physical characteristics that are considered significant as distinguishing traits. Put another way, "race is a worldview and social classification that divides humans into groups based on their appearance and assumed ancestry, and that has been used to establish social hierarchies" (Graves & Goodman, 2022). Therefore, the idea of human races is tied with the history of human social relations, as we shall soon see below.
However, the word race has become confused in the public perception of anthropology. There are assumptions of synonymy between terms like "race", "ancestry", "ethnicity", etc, but it must be stressed that each of these terms have distinct meanings.
For the benefit of readers, let's go over some of these other terms, as they will be used again throughout this series of articles.
Ethnicity - and ethnic groups - are explicitly cultural definitions. Belonging to a specific ethnicity means identifying with a group of people who share specific norms, mores, customs, values, traditions, beliefs, and habits (Kottak, 2008). Ethnic groups can be in essence religiously-based, or linguistically-based, or tied to a specific country or region, and are considered by others and themselves as belonging to the same culture and having the same history. Sometimes, but not exclusively, they can be based on race.
Ancestry is somewhat trickier to define. We think of our ancestors as the people who came before us (and a few folks may even suspect that all of their ancestors belong to a particular race) and it is a fact that the further back you go eventually you will reach the last common ancestral population of our species Homo sapiens. However, the essence of ancestry is complex and multifaceted.
Take, for example, the thought-experiment outlined by Steve Olson in his 2002 book Mapping Human History: think about your biological parents, now think about their parents. That gives you four grandparents. Now think about their parents. And keep going. You are essentially doubling your ancestral pool as you go back down the generations. Mathematically, you will eventually reach a point where the total number of ancestors alive at any given time will far exceed the known global human population (for example: if you take a generation to mean 20 years, then 30 generations ago you should have over a billion ancestors... but estimates of the human population in 1400 don't reach much beyond 375 million people!).
That means, according to statistician Joseph Chang, that when you reach a certain point (by his calculations just 800 years ago) there are two possibilities: any given person alive at the time is either your ancestor and everyone else's ancestor, or they have left no living descendants (Chang, 1999). Subsequent work has elaborated and tweaked this model - the common-ancestral age estimate has been placed somewhere 3,600 years ago - but the core findings hold true: "we see families as discrete units in our lifetimes, which they are. But they're fluid and continuous over longer periods beyond our view, and our family trees sprawl in all directions" (Rutherford, 2017). Given what we know of world history from beyond the modern colonial era, people got around fairly easily and were certainly not overwhelmingly limited by geography.
For the purposes of genetics research, scientists have used the term "ancestry" to mean different things in regards to the sample size or the period of time being studied. For the purposes of these articles, I will be discussing genetic ancestry as referring to the recovered DNA signature in descent groups that are inherited from older related or ancestral populations.
That brings us now to more technical terminology that you might come across in scientific literature. According to Philip W. Hedrick, populations are defined in terms of mating propensity, while demes are a group of interbreeding individuals that exist together in time and space, from whom mates would normally be chosen (Hedrick, 2011). These then are historical terms with no absolute boundaries: much like the rest of life, human populations and demes change over time and acquire new characteristics. This is especially true in the case of admixture or introgression, terms designating when two or more temporal populations intermix genetically, which is all the more clear in our aforementioned discussion of ancestry. As a rule, the larger the sampled population, the more frequent will be the record of admixture.
Thus, we need to be precise when using such terminology. Populations and ethnicities are not races, and race is not the same as ancestry. For a more rigorous explanation of these distinctions, statistical geneticist Sasha Gusev has written on the subject here.
A Brief History of Racial Anthropology
So where did our understanding of race and human diversity come from? People have no doubt recognized differences and similarities amongst each other since the earliest times, but so far as the evidence tells us, ancient peoples put significantly less emphasis on race as a biological concept than modern peoples have done. The Greek philosophers Aristotle & Hippocrates II spoke at length about environmental determinism and believed that climate and latitude could produce "better" or "lesser" peoples (Quinn, 2024). Other Greeks, and later Roman authors (including Christians), would embrace these ideas and use them to justify slave practices and colorism (Kendi, 2016). However, the construction of biological races as immutable categories was not in the minds of the ancients, despite an awareness of human physical variation (Graves & Goodman, 2022). They were more concerned with cultural differences and superiority, and a person - regardless of their country of origin - could assimilate and "make themselves" into, say, a Greek or Roman (Sussman, 2014).
The first inklings of biological-racialism began to show towards the end of the post-classical or medieval world. The horrifying practices of the Spanish Inquisition were at once racist and antisemitic in that they singled-out specific groups (primarily the Jewish people but also Romani & Christianized Muslims) and sought to remove them from society (Sussman, 2014). Far from being simply a religious conflict, the Inquisition specifically classified people based on "impurity of blood" and forbid the assimilation of groups that had been the norm since classical times. This mentality transferred with the Iberian colonists as they reached the Western Hemisphere and encountered Amerindian peoples, as well as during their earlier expeditions along the western African coasts.
It is particularly telling that the origin of racist ideas should have commenced during the trans-Atlantic colonial era. In previous times, explorers in Eurasia and North Africa tended to travel on foot or over caravans. As we will see, human variation is clinal with no prominent breaks in sequence: a person traveling from western Europe to China would meet people who each time would look gradually different but still retain mostly similar features. Now consider a Spanish or British sailor traveling to Central America or southern Africa: their frame of reference would now be disparate and opposite ends of a spectrum, the differences in physical features on people being necessarily enhanced by having missed the clinal variation between them (Graves & Goodman, 2022).
In any case, such differences made racist thoughts, laws, and practices all the more easier to justify, and it wasn't long until cultural and physical descriptions merged into racist categories: one of the first was the French scholar François Bernier in 1684 whose "new division of the earth" recognized 4-5 races of people and argued that white Europeans were the original blueprint of humanity. Carl Linnaeus - the founder of our modern biological classification - elaborated this scheme with his own 5-race system in 1758. He agreed with the elevation of Europeans and specifically defined the races in terms of both physical traits and cultural norms which to him were treated objectively but are so very clearly prejudiced and ignorant upon basic reading (Kendi, 2016). Johann Friedrich Blumenbach, a contemporary of Linnaeus, devised a rival 5-race system in 1795 and was the first to coin the term "Caucasian" to refer to Europeans and related peoples in southern Asia. At the same time, the bodies of disenfranchised and Indigenous peoples were being studied and dissected to support these schemes. This practice was carried on well into the 1800s. The American doctor Samuel George Morton had amassed an enormous collection of human skulls for the purposes of measurement and comparison, beginning in 1839. He concluded that the skulls of Caucasians had the highest cranial capacity and thus were the most intelligent of humans. It is interesting to consider that, while paleontologist Stephen Jay Gould had reviewed and criticized Morton's work in 1981 to much surprising controversy, in Morton's own time a German scholar Friedrich Tiedemann performed much the same experiment and could not replicate the results, subsequently seeking to challenge Morton's findings in the name of antiracism (Kendi, 2016).
From the earliest colonial times to the peak of pre-Darwinian evolutionary thought, there was an on-going discourse among Europeans about whether the human races descended from different ancestors (the polygenesis model) or whether all humans descend from one ancestor (the monogenesis model). Such arguments played key roles in the justification of slavery and other racist practices among the European nations at home and abroad. After Charles Darwin outlined natural selection in print in 1859, he eventually penned his research on human origins in another book The Descent of Man and Selection in Relation to Sex by 1871. Darwin's analysis was at once enlightened and prejudiced: he agreed with the monogenesis model and pointed out problems in racial classification from earlier authors, but he argued that the races were ancient, suspected that they came about due to sexual selection, and that each had differing levels of cognition. He agreed with his contemporaries that European peoples were the most civilized and lightly implied that the subjugation and genocide of "primitive" "savages" was an outcome of natural selection (DeSilva et al. 2021). From almost immediately this moment on, the study of human races gained an evolutionary veneer, and many scholars who misunderstood the principles of natural selection used its language to justify human hierarchies.
In one direction, this developed into a persistent and sinister ideology. In 1855, French royalist diplomat Arthur de Gobineau published An Essay on the Inequality of the Human Races in which he argued over four volumes that the physical and mental characters of different human races had remained consistent from the beginning and would remain so into the future. Among the Caucasians, Gobineau highlighted the "Aryans" (a term for the proto-Indo-European-speaking groups which lived across Central Asia) were the most advanced and responsible for everything good and civilized in the world (Kendi, 2016). Several authors took Gobineau's work and expanded upon it, including economist William Z. Ripley, whose 1899 book The Races of Europe was treated as an authoritative work in racial anthropological circles (Sussman, 2014). In the end, a direct line of descent can be traced from Gobineau & his peers to the foundations of both the American eugenics movement and the ideology of Nazism.
Long before the early 1900s, when modern anthropology came of age, there had always been antiracist critiques of the mainstream academics and scholars who purported their ideas about race. From John Woolman to Frederick Douglass, it would be unfair to say that the study of human diversity was always upheld by bigots. It would also be unfair to downplay the role of racism and its prevalence in all forms across the history of anthropology.
The work of Franz Boas and his students did much to shape the course of the field. As members of the American School of Cultural Anthropology, they approached the study of humanity from a holistic and non-essentialist approach, rejecting ideas of evolutionary hierarchy and white supremacy (Erickson & Murphy, 2008). Though not always free from cultural biases, Boas himself studied American immigrants of many backgrounds and found that some traits (like skull shape) said more about health and environment than "racial purity". Critically, in a 1894 paper , Boas said the following: "Historical events appear to have been much more potent in leading races to civilization than faculty, and it follows that achievements of races do not warrant us to assume that one race is more highly gifted than the other" (Sussman, 2014). Further work by Robert Lowie, Alfred Kroeber, & Margaret Mead did much to dismantle the scientific paradigm that a person's race determined their culture or psychology.
There were still attempts to classify human races, however, with the most notable and widely-cited attempt by Carleton S. Coon in his 1962 The Origins of Races. This work adapted an earlier scientific racial nomenclature for much of the later 20th Century, describing "Caucasoids" & "Mongoloids", for example. Such terms, frustratingly, remain in use in some circles to this day. Coon utilized older arguments from early paleoanthropology to argue that human races developed from deep in the past, from various populations of Homo erectus rather than from a single Homo sapiens ancestor. In many ways, Coon was challenging a growing consensus among younger anthropologists that "biological race" as a concept should be abandoned in regards to humans. One of Boas' students, Ashley Montagu, had written extensively on human evolution from the view of the growing modern synthesis since the 1940s, and he played a key role in outlining the UNESCO statements on race in the 1960s (Sussman, 2014). And the work of Sherwood Washburn - notably a 1951 paper - pushed for a complete overhaul of physical anthropology and a rejection of the idea of racial classification.
By the end of the 20th Century, racial anthropology had undergone a significant metamorphosis. The increasing study of human cultural and physical diversity, the advent of genetic studies, and the processing power of computer modeling had through the world's largest wrench into the study of "race". There was also a greater awareness of the societal damage done by racial anthropology and the need to seriously study the phenomena earlier scholars under modern scientific advances: "conformity to political correctness was not the cause of these changes; rather awareness of the uses of race in colonialism, slavery, segregation, and in the holocaust stimulated re-examination of the race concept using the new genetic data that was accumulated throughout the 20th century" (Leiberman, et al. 2003).
Today, many researchers have brought to light a proper understanding of human diversity in a modern evolutionary context. And their tools and data sets are even greater than at the turn of the century: we can now peer into the human genome and sequence DNA and proteins that are thousands to millions of years old. Much of this new and current research will be explored in these articles.
How Are Subspecies Recognized?
So why are humans so diverse? We're mammals right? And we recognize that many mammalian species can be divided into subspecies, so why not us? And if we can be, wouldn't these subspecies be considered races?
If the question of "what is a species" is controversial in biology and paleontology, then the question of "what is a subspecies" is probably more-so. Since the early days of Linnaean taxonomy, naturalists have flooded the literature with trinomial names, often using little justification beyond differences in, say, fur color or relative size. In 1942, biologist Ernst Mayr defined a subspecies (which he synonymized with "geographic race") as "a geographically localized subdivision of the species, which differs genetically and taxonomically from other subdivisions of the species". Other later authors have provided mathematical rules based on genetics; for example, Susan Haig & colleagues wrote that "the only quantitative subspecies definition we found was the 75% rule that states a subspecies is valid if 75% or more of a population is separable from all (or >99% of ) members of the overlapping population" (Haig, et al. 2006).
In their 2019 book concerning wolf conservation, Joseph Travis and colleagues provide more multifaceted criteria. They argued for combining a wealth of 1) morphological and fossil evidence, 2) genetic evidence, and 3) ecological and behavioral evidence to support a proper and valid taxonomic division of a subspecies, with "the general view being that subspecies are groups of actually or potentially interbreeding populations that are phylogenetically distinguishable from, but reproductively compatible with, other such groups" (Travis, et al. 2019).
One could argue that all this talk of subspecies is moot. Stephen Jay Gould argued in his essay "Why We Should Not Name Human Races - a Biological View" that the use of subspecies as a classifier inhibits and oversimplifies our understanding of regional evolution. Using case studies from island land snails and house sparrows, he decried "shall our approach to such variation be that of a cataloger? Shall we artificially partition such a dynamic and continuous pattern into distinct units with formal names? Would it not be better to map this variability objectively without imposing upon it the subjective criteria for formal subdivision that any taxonomist must use in naming subspecies?" (Gould, 1977).
Regardless, there is at least a set of guidelines that can be followed to determine whether humans can be classified as subspecies. Such processes have already been used to mold a proper understanding of living mammalian diversity, and this research is perpetually changing. For example, it was once proposed that tigers (Panthera tigris) constituted about nine subspecies, but arguments have been made to condense this number to just two. In 2023 alone, one paper favored the two-subspecies proposal while another favored the old nine-subspecies model (Wang, et al. 2023; Sun, et al. 2023). Giraffes have gone through a similar back-and-forth, between proposals that the giraffe is one species with nine subspecies and that the giraffe represents multiple species: the most recent analysis favors four species, each with a few subspecies (Kargopoulos, et al. 2024).
Far from being exempt, much research has been done on the question of human genetic diversity and whether this matches what we in other large mammals.
How Does Human Diversity Quantify?
Morphological and fossil evidence had been used in the past to classify humans into subspecies and races, but recent work on a more complete hominin fossil record shows without question that all Homo sapiens remains share more in common with each other than with other human species like Neanderthals or Homo erectus. There are characteristic features like our globular skull, prominent chins, and gracile rib-cage and pelvis which clearly define our species whenever remains have been found on every continent. There is regional variation in space but also time, and these changes to skeletal morphology have been shown as having more to do with environment, culture, and sheer chance than with relatedness to a specific racial type. Paleolithic humans have historically been difficult to classify with modern "racial groups", and there has never been enough time and distance to keep any one population isolated for so long as to develop distinct traits. We're really good at moving around and mating.
The difficulties in objectively studying differences based on morphology have let researchers to turn to genetics. One of the most famous studies was conducted by Richard Lewontin in 1972. He sampled populations from around the world and quantified the genetic diversity within and between proposed racial and ethnic groups. Here's Lewontin in his own words:
"The results are quite remarkable. The mean population of the total species diversity that is contained within populations is 85.4%, with a maximum of 99.7% for the Xm gene, and a minimum of 63.6% for Duffy. Less than 15% of all human genetic diversity is accounted for by differences between human groups! Moreover, the difference between populations within a race accounts for an additional 8.3%, so that only 6.3% is accounted for by racial classification" (Lewontin, 1972).
Lewontin's work was widely shared and referenced in the years since its publication, but there have also been subsequent studies that have essentially validated its primary findings: that there are more genetic differences between human groups within proposed races than between proposed races, and that human genetic diversity as a whole is pretty small. Just some examples: Barbujani, et al. 1997 reported 84.5% of genetic diversity found within populations, 5% is found between populations within a “race”, 8-11.7% between “races”; Rosenberg, et al. 2002 reported 3-95% of genetic diversity found within populations, 2.4% is found between populations within a “race”, 3.6% between “races”; and Li, et al. 2008 found 89.9% of genetic diversity within populations, 2.1% is found between populations within a “race”, 9% between “races”.
Furthermore, work has failed to find evidence of specific genetic variants which are "private to geographic regions (excluding individuals with likely recent admixture from other regions)" (Bergström, et al. 2020). It is quite apparent that Homo sapiens had a very recent ancestry and descended from a fairly small gene pool which has expressed itself in the remarkably similar genetic blueprint between all of us: "63% of variants common in at least one region are also globally widespread, in the sense of being found across all five regions. This number rises to 82% for variants common in at least one region outside of Africa." (Biddanda, et al. 2020).
How human genetic diversity compares to other large mammals (Templeton, 1998)
When all of this is compared to other large mammals, the genetic variation shrinks even more. A 1998 paper by Alan R. Templeton used a fixation index on a sample of mammal species to see where they stood on a spectrum between equally-shared genetic diversity within but not between populations and fixed genetic diversity between but not within populations. The results showed that global human genetic diversity was lesser than, say, the population of impalas in Kenya or all the populations of wolves across the Northern Hemisphere. Even among our fellow great apes, we find that the total diversity of humans is dwarfed by that found in regional populations of gorillas or chimpanzees (Gagneux, et al. 1999). Again, such results corroborate the findings of Lewontin and his successors.
How human genetic diversity compares to other great apes (Gagneux, et al. 1999)
Lastly, using ecology or behavior to classify humans into subspecies just does not work, because adaptive traits can vary widely between human groups, even within the same environments, and are also often determined by cultural or historical factors. So if you wanted to classify humans in this way, you would find that different "races" have adapted to the challenges of different climatic or environmental conditions inconsistently. Sometimes cultural solutions would overlap, but never always. Biological human ancestry does not parallel with behavioral ecology in the same way it does with other organisms.
In Essence...
Taken together, morphological, genetic, and ecological data have repeatedly shown that, statistically, these are not great enough to warrant classification into subspecies, let alone races. At no time has a human group been isolated geographically for so long that it is morphologically and genetically distinct from the rest of the world population. Human diversity is best understood as clinal, showing continuity between groups with little in the way of significant geographic barriers that are not broken somewhere in the world. Noah Rosenberg's aforementioned 2002 study utilized a software called STRUCTURE which placed genetic similarities into clusters which could then be divided into any given number the researcher provides. In this particular study, the majority of the clusters centered on five geographic groups that one could say support a division of humanity into five races, but (as Adam Rutherford eloquently explained in his 2017 book A Brief History of Everyone Who Ever Lived) this would be succumbing to the sort of past subjectivity that painted racial anthropology: "look for clusters, and you'll find clusters". Rather than plugging these results into preexisting racial classification systems, wouldn't it be better to ask new questions? "Why do we see these groupings, especially if the rest of the genome, in fact the majority of the genome, does not show such regional variation?"
STRUCTURE clusters found across global human samples (Rosenberg, et al. 2002)
It must be stressed, then, that biological anthropologists today do recognize that there is clear regional variation between human groups. While we are all so similar genetically, we are not clones. The study of this diversity has opened up exciting avenues of research which have revealed key insights into the last 300,000 years of human evolution. But it is also with a strong and firm hand that it must also be stressed that race is an inaccurate, unhelpful, and meaningless way to understand all this diversity. In much the same way that most paleontologists and evolutionary biologists have done away with the essentialist Linnaean schemes that have plagued animal classification, so to have proper anthropologists abandoned "race science". As we've clearly seen at the start of this post, race science was clearly born of racism, and this is a fact that must always be emphasized whenever someone tries to argue about "race realism" or "human biodiversity": nearly of the time, arguments about the validity of race as a biological concept have racist roots, no matter how deep.
Representations of clinal human variation. A showcases a simple cline and the effect selective sampling can do in suggesting distinct genetic clusters; B shows global genetic diversity represented by continental-scale gradients (Maglo, et al. 2016)
For all this diversity, human beings are still importantly similar, moreso than many other large mammals. So much so that we can be considered a monotypic species: biologically we are Homo sapiens and that's as specific as we can get. Human subgroups do not fall under the criteria used to classify subspecies of mammals, nor any other category like breed (see Norton, et al. 2019). Though human physical & genetic diversity is real, racial classification does not reflect this; nor can a cladogram fully represent the evolution of human diversity over time, so full of admixture as it is. Our genetic past is a tangled knotted bush, not a neat pedigree-tree. Human genetic variation is statistically greater within populations than between populations, even when factoring in geographic clustering. All humans, everywhere, are closely related members of the same species.
In the next post, I'm going to specifically address the latest research into human diversity. What does the science say about our range of skin colors, hair shapes, skull dimensions, dentition, blood type, and other features of our anatomy? What about disease? Or IQ and intelligence? Some of the answers may surprise you.
Book References
Jeremy DeSilva, et al. A Most Interesting Problem: What Darwin's Descent of Man Got Right and Wrong about Human Evolution (Princeton University Press, 2021)
Paul A. Erickson & Liam D. Murphy. A History of Anthropological Theory (University of Toronto Press, 2008)
Alan H. Goodman & Joseph L. Graves Jr. Racism, Not Race: Answers to Frequently Asked Questions (Columbia University Press, 2022)
Stephen Jay Gould. The Mismeasure of Man - 2nd Edition (W, W, Norton & Company, 1996)
Stephen Jay Gould. Ever Since Darwin: Reflections in Natural History (W. W. Norton & Company, 1977)
Philip W. Hedrick. Genetics of Populations - 4th Edition (Jones & Bartlett Learning, 2011)
Ibram X. Kendi. Stamped from the Beginning: The Definitive History of Racist Ideas in America (Bold Type Books, 2016)
Conrad Phillip Kottak. Cultural Anthropology - 12th Edition (McGraw Hill, 2008)
Ernst Mayr. Systematics and the Origin of Species (Columbia University Press, 1942)
Steve Olson. Mapping Human History (Houghton Mifflin Company, 2002)
Jonathan Pritchard. An Owner's Guide to the Human Genome (Standford University, 2024 & ongoing)
Josephine Quinn. How the World Made the West (Bloomsbury Publishing, 2024)
Adam Rutherford. A Brief History of Everyone Who Ever Lived (The Experiment, 2017)
Robert W. Sussman. The Myth of Race (Harvard University Press, 2014)
Joseph Travis, et al. Evaluating the Taxonomic Status of the Mexican Gray Wolf and the Red Wolf (National Academies Press, 2019)
Paper & Article Citations
Guido Barbujani, et al. 1997. An apportionment of human DNA diversity (PNAS)
Anders Bergström, et al. 2020. Insights into human genetic variation and population history from 929 diverse genomes (Science)
Arjun Biddanda, et al. 2020. A variant-centric perspective on geographic patterns of human allele frequency variation (Elife)
Joseph T. Chang, 1999. Recent common ancestors of all present-day individuals (Advances in Applied Probability)
Pascal Gagneux, et al. 1999. Mitochondrial sequences show diverse evolutionary histories of African hominoids (PNAS)
Sasha Gusev, 2024. A molecular genetics perspective on the heritability of human behavior and group differences (Gusev Lab)
Susan Haig, et al. 2006. Taxonomic Considerations in Listing Subspecies Under the U.S. Endangered Species Act (Conservation Biology)
Nikolaos Kargopoulos, et al. 2024. Heads up–Four Giraffa species have distinct cranial morphology (PLoS One)
Richard Lewontin, 1972. The Apportionment of Human Diversity (Evolutionary Biology)
Jun Z Li, et al. 2008. Worldwide human relationships inferred from genome-wide patterns of variation (Science)
Leonard Lieberman, et al. 2003. The Decline of Race in American Physical Anthropology (Anthropological Review)
Koffi N Maglo, et al. 2016. Population Genomics and the Statistical Values of Race: An Interdisciplinary Perspective on the Biological Classification of Human Populations and Implications for Clinical Genetic Epidemiological Research (Fronteirs in Genetics)
Heather L. Norton, et al. 2019. Human races are not like dog breeds: refuting a racist analogy (Evolution: Education and Outreach)
Noah A. Rosenberg, et al. 2002. Genetic structure of human populations (Science)
Xin Sun, et al. 2023. Ancient DNA reveals genetic admixture in China during tiger evolution (Nature Ecology and Evolution)
Alan R Templeton, 2013. Biological races in humans (Stud Hist Philos Biol Biomed Sci)
Alan R. Templeton, 1998. Human Races: A Genetic and Evolutionary Perspective (American Anthropologist)
Chen Wang, et al. 2023. Population genomic analysis provides evidence of the past success and future potential of South China tiger captive conservation (BMC Biol)
Sherwood Washburn, 1951. The New Physical Anthropology (Section of Anthropology: Transactions of the New York Academy of Sciences)
Competing hypotheses for some stem-eutherian relationships, from Brady, et al. 2024
Placental mammals - clade Placentalia - constitute the vast majority of mammals alive today, with well over 6,400 species (according to the Mammal Diversity Database). The last few decades of molecular genomics have revealed four major lineages of placentals that have encompassed nearly all of this diversity:
Xenarthra - a small group of American mammals lacking many skeletal features in other placentals (e.g. stirrup-shaped ear bones), including sloths, anteaters, & armadillos
Afrotheria - emerging from Africa, a collection of bizarre forms including elephants, hyraxes, sea cows, tenrecs, golden moles, sengis, & aardvarks
Euarchontoglires - the largest group, whose members evolved an appendix (convergently with herbivorous marsupials), including rodents, rabbits, treeshrews, colugos, & primates
Laurasiatheria - a clade of vastly disparate morphologies, including shrews, moles, hedgehogs, bats, pangolins, carnivorans, odd-toed & even-toed hoofed mammals (the latter including whales)
It is well understood that laurasiatheres and euarchontoglirans are sister-groups united under Boreoeutheria, while it is less clear but generally thought that xenarthrans and afrotherians are united under Atlantogenata. Both of these larger clades correspond with ancient continental divisions, the boreoeutheres having evolved in the Laurasia; the atlantogenatans in Gondwana.
When looking at the fossil record quite a number of groups have been revealed which, though they are clearly related to living placentals, their classification among them has been quite contentious. Indeed, some of these may not even be placentals at all!
A distinction needs to be made between stem and crown groups.
Cladogram diagram (Philcha, CC BY-SA 3.0)
In modern evolutionary biology, researchers try to classify organisms based on shared common ancestry: this is the basis of cladistics or phylogenetics, and thus the representations of these relationships are called cladograms or phylogenies. A clade is the total group of organisms descended from a single common ancestor, constituting a monophyletic group. The diagram above is a representative cladogram, and the branching groups surrounded in blue and red are clades. But what about the yellow? Clearly, it includes a common ancestor and its descendants, but not all of them are accounted for. This is known as a paraphyletic group.
As well, the clades in red consist of groups which survive to the present day (that is, they are extant). But all the groups within the yellow space are extinct, and have left no living descendants. Sure, they may be part of a lineage that is alive today, but those particular branching points are gone.
So, groups which are allied to living groups but no longer survive are referred to as stem groups. A clade in which the last common ancestor of two descendant lines which both survive into the modern day are referred to as crown groups.
The clade Placentalia is a crown group, because all its descendants are around today. But there are also a number of related forms (stem groups) which are more closely related to placental mammals than to the nearest living group (being the marsupials), so all these mammals + Placentalia constitute a larger clade called Eutheria.
It's very difficult to know what the first eutherians were like and whether they were similar physiologically to placental mammals in, say, having a complex placenta with a trophloblast layer around the embryo (Rose, 2006). These don't fossilize easily, but bones and teeth tend to do better! It has been observed, for example, that eutherians tend to possess three molar teeth, two roots on the upper canines, and other dental distinctions (Bi, et al. 2018). Luckily for paleomammalogists, teeth are more readily preserved in the fossil record, but unluckily that means other aspects of the skeleton are not so well known. Concerns have thus been raised that some proposed relationships based on dental traits may say more about convergent evolution than common descent (Prothero, 2017).
Prior to the genetics revolution, evolutionary relationships were based on skeletal similarities that could be convincingly shown to be homologous or shared from a common ancestor, and for a good while a decent family tree of mammals was being built. Once it was demonstrated that molecular sequencing could reveal relationships, it revolutionized the field. This has meant a new classification for mammals, but also that now many fossil groups would need to be reconciled within these clades. And that hasn't always been easy. The more data is used, the more researchers have found conflicting results that some specific mammals are either stem placentals, nested within crown placentals but outside living clades (making them a new type of stem group in respect to the survivors), or belong within crown groups. It can get more than a little confusing...
In the most recent case, a 2022 paper revealed an entire branch of stem placentals that was named Tamirtheria and consisted of a selection of small mammals including leptictids, zalambdalestids, and asioryctitheres (Velazco, et al. 2022). Later research a few years hence questioned this, noticing that the way anatomical characters are ordered or unordered in a cladistic analysis can alter the branching relationships in the final result (Brady, et al. 2024). This meant that, depending which arrangement of characters is used, the monophyly of Tamirtheria could either be supported or not, and in the team's analysis, they rarely recovered the tamirtheres as a natural group. Instead, they found them as a branching system of stem groups leading up to Placentalia. So, for the moment, the recognition of this newly named clade is controversial, and more comprehensive work would be needed to sort out these groups.
Listed below are the handful of eutherian mammal lineages that have been, to the best of the most recent studies, the best candidates for stem-placentals. They are listed roughly in order of their divergence respective of the crown, but it must be understood that this sequence is not set in stone. There could be new research tomorrow that upends some or all of this, but for the moment this is what the current paleontological evidence tells us.
Parade of Stem-Placental Mammals
Early Cretaceous eutherians Duristodon (on branch) and Duristotherium (running) (Mark Witton, CC BY 2.0)
Until fairly recently, paleontologists have recognized three species that were claimed as the earliest eutherians. There was Juramaia sinensis from the Late Jurassic and Eomaia scansoria & Acristatherium yanensis of the Early Cretaceous, both found in the deposits of Liaoning, China. Contemporanous to the latter two was Sinodelphys szalayi, then recognized as one of the earliest metatherians (the total group which includes marsupials), and these fossils seemed to align well with the then-current molecular data showing a divergence of the daughter therian clades around 160 million years ago in the Late Jurassic (Luo, et al. 2011).
However, subsequent studies have cast doubts on all of this. Acristatherium & Juramaia may not be a eutherians but a stem-therians (Sweetman, et al. 2017), and Juramaia might not even be Jurassic in age but Early Cretaceous instead (King & Beck, 2020). Eomaia is more contentious: some analyses have argued it is a stem-therian (O'Leary, et al. 2013), and some argue it - as well as Acristatherium & Juramaia - are eutherians after all (Wang & Wang, 2023). In contrast, Sinodelphys may be a eutherian and not a metatherian (Bi, et al. 2018, Wang & Wang, 2023)!
A Jurassic origin for eutherians is still on the table somewhat, with other analyses showing a diversity displayed among other fossil mammals that suggests a pre-Cretaceous origin (Sweetman, et al. 2017). In particular, more confidently-placed early eutherians include Duristotherium & Duristodon from the Early Cretaceous of the Purbeck Group of southern Britain, Montanalestes from the Early Cretaceous of the Cloverly Formation in the central USA (Cifelli, 1999), & Cokotherium and Ambolestes from the Early Cretaceous of Liaoning, China (Bi, et al. 2018, Wang, et al. 2022).
Ambolestes and Cokotherium in particular are known from complete skeletons which preserve the middle ear bones, the hyoid bone (which anchors the tongue), and hair and body impressions. The phylogenetic analysis of Cokotherium supported the conclusion that it and many of these aforementioned mammals are the earliest diverging eutherians (Wang, et al. 2022). The mounting evidence from these more complete fossils and their recognition of Sinodelphys as a close relation has lend support to an understanding that early eutherians were far more metatherian-like than is generally appreciated: the anatomies of the sister therian clades almost blend into each other (Rose, 2006).
The clearest example of this is the presence of epipubic bones in more complete fossils of early eutherians, like zhelestids & zalambdalestids. While the epipubis has now been found extending deep into mammaliaform history, it has been traditional argued that in marsupials it functions as a support for the pouch. This has been called into question, as it now seems these bones function in a greater reproductive and locomotory sense (Novacek, et al. 1997). Prior to the late 1990s, the loss of epipubic bones has been treated as a eutherian trait "related to the evolution of prolonged gestation, which would not require prolonged external attachment of altricial young" (Novacek, et al. 1997). With these bones confirmed in the earliest eutherians, it now means that it cannot be used as a sole trait in determining a placental or placental-relative from other mammals.
Epipubic bones of a living kangaroo (edited, from Pierre-Yves Beaudouin, CC BY-SA 4.0)
Then we find other genera from the Early Cretaceous that ally more closely to placentals than these earlier forms, but the certainties of their relationships end there. These include Prokennalestes from Khovoor, Mongolia (Lopatin & Averianov, 2018), Murtoilestes from Transbaikalia, Russia (Averianov & Skutschas, 2001), and Bobolestes from Uzbekistan. These show the early eutherians were still primarily Laurasian in range at this time.
There is a rather enigmatic clade called Adapisoriculidae that has been argued in the past to have been related to either lipotyphlans (shrews, moles, & hedgehogs) or archontans (primates, colugos, & treeshrews), but recent studies show them to be early-diverging eutherians distantly related to placentals (Manz, et al. 2015). Fossils, primarily teeth and jaw fragments have been found from the Late Cretaceous to the Eocene Epoch in India, western Europe, and north Africa. They seem to have been roughly shrew-sized animals and little is known of their ecology (Rose, 2006). The most curious information regards Deccanolestes, whose presence in the Late Cretaceous of India adds to a growing discourse about whether land connections existed between the wider world and the subcontinent when it was an isolated landmass (Prasad, et al. 1994).
Another early group was the Zhelestidae which is known from more complete remains from the Late Cretaceous of Eurasia, Africa, and North America. They were generally mice-to-rat-sized mammals, and were primarily found living on "wetter, low coastal plain settings" (Nessov, et al. 1998). Zhelestids and potentially related forms show evidence of specialized cheek teeth (including five upper premolars) and jaws pointing to adaptations towards herbivory or at least greater omnivory (Kemp, 2005). Previously this has led to proposals that they have something to do with living ungulates or hoofed mammals (Prothero, 2017), but recent work supports a stem placental placement (Gheerbrant & Teodori, 2021; Archibald & Averianov, 2013).
Several studies have converged on the presence of a larger stem clade consisting of at least three major groups (Brady, et al. 2024; Velazco, et al. 2022; Halliday, et al. 2015).
The first of these to diverge is Cimolesta, a historically important group which has encompassed many varieties of Cretaceous and Paleogene mammals over many decades of classification. For example, the 1997 McKenna/Bell classification scheme saw Cimolesta as a ordinal-group consisting of pantodonts, tillodonts, taeniodonts, apatotheres, pantolestids, and pangolins; this order was also argued to be related to carnivorans and creodonts (Rose, 2006). Subsequent work has confirmed a relationship between pangolins and carnivorans, but the majority of these other groups have remained very contentious. Pantodonts and tillodonts may be related to ungulates (Halliday, et al. 2015; Bertrand, et al. 2023), while apatotheres may belong to Euarchontoglires (Silcox, et al. 2010); this would make these forms crown placentals, but as always these results are bound to be controversial.
In recent years the outline of the remaining cimolestans has gotten more streamlined. We can confidently say that the group is Late Cretaceous in origin, based on the remains of two clades: Cimolestidae and Pantolesta. Cimolestids - including Cimolestes proper, Batodon, and Maelestes - were small climbing carnivorous forms with shearing & slicing "incipiently carnassial" teeth that preyed across the Americas, eastern Eurasia, and north Africa (Halliday, et al. 2015). Rudolph Zallinger famously illustrated a Cimolestes on the day he signed his The Age of Reptiles mural for the Yale Peabody Museum in 1947. An interesting choice, considering this genus is also known from Paleocene and Eocene deposits, meaning that these mammals survived the bolide impact that ended the non-avian dinosaurs.
Otter-like pantolestid Palaeosinopa (ДиБгд, CC BY 4.0)
Pantolestans were far more diverse and ranged until the Oligocene Epoch across North America, Africa, and Eurasia. They seem to be united in their dental morphology, in which their four large premolars and three molar teeth with low crowns and rounded cusps (Rose, 2006). Such toughened chompers suggest a diet of toughened foods (what is known as durophagy), and indeed some fossils preserve worn-down cheek teeth. At least three lineages are included. The Pantolestidae had broad wide heads and a crest at the back of the skull which provided extra support for neck muscles. Based on the fossils of Buxolestes, Pantolestes, and Palaeosinopa, they appear to have been water-loving omnivores that feasted on mollusks, fish, and other freshwater organisms and may have resided in water-side burrows just like river otters. Postcranial remains complete this resemblance by showing the presence of flexible limbs and a strong, propulsive tail. The Paroxyclaenidae are solely Eurasian, and the best representative genus was Kopidodon (Rose, 2006). Built like a beefy-raccoon, these mammals had broad faces and walked on their soles, and the flexible limbs and long-bushy tail (known from impressions found at the famous Messel Pit in Germany) suggest that they were arboreal. Lastly, the Pentacodontidae are known primarily from skull material, but a few postcrania have been found (Rose, 2006). They seem like more robustly-built versions of pantolestids, and it has been speculated that the genus Bisonalveus was venomous based on the presence of grooves in the canine teeth, though this has been seriously questioned (Folinsbee, et al. 2007).
Taeniodont Wortmania (TheExplainer, CC BY-SA 4.0)
Of immediate interest is the status of the Taeniodonta. They appear to be closely related to the Cimolestidae through the genera Procerberus & Alveugena (Rook & Hunter, 2013; Bertrand, et al. 2022). Ranging from the Late Cretaceous to Eocene of North America and Europe, taeniodonts were a sort-of cross between an opossum, a beaver, a panda bear, and an armadillo (sans armor). They had massive, powerful jaws with enlarged incisor & canine teeth and simple cheek teeth that got smaller the further back in the mouth you go (Rose, 2006). The front teeth appear to have been increasingly specialized across successive genera like Wortmania, Psittacotherium, Ectoganus, and Stylinodon, and point to adaptiations for chiseling and rooting plant materials from leaves to twigs (Prothero, 2017). As well, they could dig up roots or pull down branches with ease thanks to long forelimbs tipped with curved, clawed fingers. The name taeniodont means "ribbon tooth" in Greek and was coined after an enamel band that grew on the incisors of Stylinodon, which allowed for self-sharpening.
Seen as relatives to the cimolestans are two sister groups: Asioryctitheria and Zalambdalestidae (Brady, et al. 2024, Velazco, et al. 2022, Halliday, et al. 2015).
The clade Asioryctitheria - as the name suggests - spanned the late-Early Cretaceous to Late Cretaceous Epochs of Eastern Eurasia, though this group seems to have died out prior the K-Pg Extinction Event (Kemp, 2005). They were tiny animals, with roughly 1-inch long skulls that somewhat resembled those of treeshrews (Carrol, 1988). The foot bones were built in such a way that the astragalus (a bone in the ankle joint) always touches the ground when these mammals are moving about. This is a small group, only consisting of a few genera known primarily from skull material and mostly incomplete skeletons: Ukhaatherium & Kennalestes, both from Mongolia, are some of the most well-known and are frequent representatives of the clade in morphological phylogenies.
Zalambdalestes (Smokeybjb, CC BY-SA 3.0)
The clade Zalambdalestidae is also only known from the Cretaceous Period of East Eurasia, though these forms grew a bit larger, with skulls roughly 2 inches long and a body the size of a rat (Kemp, 2005). These mammals were far from being rat-like, however. Recent studies on Zalambdalestes proper have revealed a powerful and agile predator. Fossils reveal an almost rabbit-like locomotion provided by a flexible spine and long limbs tipped with elongated toes (Chen & Wilson, 2015). Zalambdalestes likely bounded after insects and other small animals, catching them with quick-firing neck and lengthy, flattened lower incisor teeth; it may even have been somewhat spiny like living hedgehogs (Arnold, et al. 2024). Such a speedy-lifestyle would be beneficial for dodging attacks by small predatory dinosaurs, like the contemporaneous Velociraptor.
The remaining stem-placentals have been usually recovered as especially close to Placentalia (Brady, et al. 2024; Bertrand, et al. 2022; Halliday, et al. 2015), but as always such results do not exactly agree with each other.
Leptictidium (DagdaMor, CC BY 3.0)
The Leptictida are probably the most familiar of all the stem-placentals, thanks to the depiction of the European Leptictidium in BBC's Walking With Beasts (which likely earned its representation on, of all things, Nickelodeon's Jimmy Neutron series). This clade lived from the latest Cretaceous, through the K-Pg, and into the Paleogene Period, across North America and Eurasia. The earliest forms were Gypsonictops & Sikuomys (both North American), which are known from fragmentary but fairly common fossils of jaws and teeth (Kemp, 2005). Later forms from the Paleogene are far more complete and show animals with elongated snouts and thin, lenghty hindlimbs; the third premolar teeth have also been lost (Carroll, 1988). Included are the genera Prodiacodon, Palaeictops, Pseudorhyncocyon, Leptictis, and the aformentioned Leptictidium. The latter has been the subject of some discourse regarding the nature of its bipedal anatomy: the jury is still out on whether this animal ran on its hindlimbs like a theropod dinosaur or hopped like the mammalian jerboa (Rose, 2006). Regardless, the forelimbs were far from weak, and their robust-build suggests a use in digging. Research on the inner ear of a number of genera show that Leptictidium was fairly agile and fast moving, while Leptictis & Palaeictops were less so (Ruf, et al. 2016). Otherwise, like many of the stem-placentals, leptictids were fairly generalized skeletally compared to living forms.
One clade, the Didymoconidae of Paleogene Central Asia, has proven quite frustrating to classify and previous work has sought relations with groups as disparate as mesonychids, lipotyphlans, creodonts, and arctocyonids; the most recent work using material from newly discovered fossils loosely suggests that the late tooth-eruption and other dental traits point to a close relationship with the Leptictida (Morio & Nagel, 2002). Otherwise, the didymoconids are fairly mysterious as a group. We mostly have teeth and jaws - some of which are complete - show mammals with broad heads and deep jaws, sometimes studded with long canine teeth (Rose, 2006). What little postcrania there is (representing the genera Didymoconus & Ardynictis) include forelimb bones pointing to fossorial or burrowing behaviors.
Lastly, we come to the Palaeoryctidae, from nearly global deposits of the Late Cretaceous to the Eocene Epoch. It is unclear exactly how they related to other stem-placentals: they appear to be a closely allied stem branch to Placentalia (Bertrand, et al. 2022), but there is a unresolved question about whether they are close relatives of leptictids or not (Wible & Bertrand, 2024). Quite small, short-snouted, and shrew-like, palaeoryctids are among the smallest fossil mammals known, with the molar teeth of Palaeoryctes minimus being only a millimeter in length (Rose, 2006). They were likely shrew-like in habits too, having shearing & piercing dentition suggesting a specialized diet of insects and other invertebrates (Kemp, 2005).
Natural History
All of the earliest eutherians were small opossum- or mouse-like mammals that could rest comfortably in your palm. They scampered and scurried in the foliage and through the trees, snapping up insects and other small animals while avoiding being prey to smaller dinosaurs. They ranged across North America and Eurasia during the earliest Cretaceous (and likely earlier) and seemingly gave rise to a wide variety of forms that were globally widespread by the later Cretaceous.
By the end of the Cretaceous, several lineages survived the destruction, including the cimolestans, leptictids, and adapisoriculids, but many of the older more plesiomorphic groups died out (coinciding with the diversification of multituberculates). Analysis of Paleocene sites immediately following the bolide impact suggest that mammals recovered so well that they had almost recovered to their Cretaceous body-sizes by 100,000 years hence; by 700,000 years hence large and diverse bodyplans had originated in the increasingly warm and tropical environment (Lyson, et al. 2019). In fact, they may have done a little too-well, as scans of fossil skulls show that in the rapid increase of scale and proportions, mammalian brain sizes actually decreased relative to the growth (Bertrand, et al. 2022). This is a trend seen across all eutherian groups, including forms related to living placentals.
And what about the placentals anyway? Where do they fit into the story? Genetic analyses calibrated with fossils point to a Late Cretaceous origin for Placentalia, alongside all the other stem groups, but the various daughter groups outlined at the very start of this article do not seem to have emerged until during or after the K-Pg boundary, 66 million years ago (Carlistle, et al. 2023). Such recent work illuminates a decades-long discourse between the "Long Fuse Model" and the "Soft Explosive Model": that is, whether placental mammal groups had all originated in the Cretaceous and remained cladistically static until they began to diversity in the Paleogene (the Long Fuse) or whether the common-ancestral group remained static only to then diversity into the main branches in the Paleogene (the Soft Explosive). It seems like both models explain this pattern in the fossil record and molecular trees.
Recognizing stem-placentals has helped clear this up. You will have by now recalled how many of the listed clades had been allied in the past with living placental groups like ungulates, carnivorans, and lipotyphlans. This mattered because if it was true that they all represented Cretaceous placentals, then the entire family tree would necessarily have to extend to that time. Thus, by the discovery that all these fossils represent now-extinct stem lineages, suddently the divergence-data makes more sense.
In the end, these stem-clades lost out in the evolutionary story. By all accounts they do not appear to have been successful in the long run despite their diverse adaptations. By the Eocene and Oligocene epochs, the ancestors of living placental mammals had begun to expand around the world, growing larger brains with enlarged sensory regions and having behaviors that could respond to the increasingly complex and crowded global environment (Bertrand, et al. 2022). Somehow, the stem-placentals could not keep up, and today there are no clearly recognizable descendants of any of them, as much as paleontologists have tried to find them.
Sure, there are no more taeniodonts, zhelestids, or palaeoryctids running around, but the morphological and ecological blueprints of these groups live on in the over 6,000 different mammals we see today. Perhaps that is legacy enough?
Book References
Robert L. Carroll. Vertebrate Paleontology and Evolution (WH Freeman, 1988)
T. S. Kemp. The Origin and Evolution of Mammals (Oxford University Press, 2005)
Donald R. Prothero. The Princeton Field Guide to Prehistoric Mammals (Princeton University Press, 2017)
Kenneth D. Rose. The Beginning of the Age of Mammals (The Johns Hopkins University Press, 2006)
Paper & Website Citations
J. David Archibald & Alexander O. Averianov, 2013. Phylogenetic analysis, taxonomic revision, and dental ontogeny of the Cretaceous Zhelestidae (Mammalia: Eutheria) (Zoological Journal of the Linnean Society)
Patrick Arnold, et al. 2024. The Late Cretaceous eutherian Zalambdalestes reveals unique axis and complex evolution of the mammalian neck (Science Bulletin)
Alexander O. Averianov & Pavel P. Skutschas, 2001. A new genus of eutherian mammal from the Early Cretaceous of Transbaikalia, Russia (Acta Paleontologica)
Ornella C. Bertrand, et al. 2023. The virtual brain endocast of Trogosus (Mammalia, Tillodontia) and its relevance in understanding the extinction of archaic placental mammals (Journal of Anatomy)
Ornella C. Bertrand, et al. 2022. Brawn before brains in placental mammals after the end-Cretaceous extinction (Science)
Shundong Bi, et al. 2018. An Early Cretaceous eutherian and the placental–marsupial dichotomy (Nature)
Peggy L. Brady, et al. 2024. The effects of ordered multistate morphological characters on phylogenetic analyses of eutherian mammals (Journal of Mammalian Evolution)
Emily Carlisle, et al. 2024. A timescale for placental mammal diversification based on Bayesian modeling of the fossil record (Current Biology)
Meng Chen & Gregory P. Wilson, 2015. A multivariate approach to infer locomotor modes in Mesozoic mammals (Paleobiology)
Richard L. Cifelli, 1999. Tribosphenic mammal from the North American Early Cretaceous (Nature)
Kaila E. Folinsbee, et al. 2007. Canine grooves: morphology, function, and relevance to venom (Journal of Vertebrate Paleontology)
Emmanuel Gheerbrant & Dominique Teodori, 2021. An enigmatic specialized new eutherian mammal from the Late Cretaceous of Western Europe (Northern Pyrenees) (Rendus Palevol)
Thomas J. D. Halliday, et al. 2015. Resolving the relationships of Paleocene placental mammals (Biol Rev Camb Philos Soc.)
Benedict King & Robin M. D. Beck, 2020. Tip dating supports novel resolutions of controversial relationships among early mammals (Proc. R. Soc. B.)
A. V. Lopatin & Alexander. O. Averianov, 2018. The stem placental mammal Prokennalestes from the Early Cretaceous of Mongolia (Paleontological Journal)
Zhe-Xi Luo, et al. 2011. A Jurassic eutherian mammal and divergence of marsupials and placentals (Nature)
T. R. Lyson, et al. 2019. Exceptional continental record of biotic recovery after the Cretaceous–Paleogene mass extinction (Science)
Mammal Diversity Database, Version 2.2 (2025). Zenodo
Carly Manz, et al. 2015. New partial skeletons of Palaeocene Nyctitheriidae and evaluation of proposed euarchontan affinities (Biology Letters)
Pieter Missiaen, et al. 2013. A new species of Archaeoryctes from the Middle Paleocene of China and the phylogenetic diversification of Didymoconidae (Geologica Belgica)
Michael Morlo & Doris Nagel, 2002. New Didymoconidae (Mammalia) from the Oligocene of Central Mongolia and first information on tooth eruption sequence of the family (Neues Jahrbuch für Geologie und Paläontologie - Abhandlungen)
Lev. A Nessov, et al. 1998. Ungulate-like mammals from the Late Cretaceous of Uzbekistan and a phylogenetic analysis of Ungulatomorpha (Bulletin of Carnegie Museum of Natural History)
Michael J. Novacek, et al. 1997. Epipubic bones in eutherian mammals from the Late Cretaceous of Mongolia (Nature)
Maureen A. O'Leary, et al. 2013. The Placental Mammal Ancestor and the Post–K-Pg Radiation of Placentals (Science)
Guntupalli V. R. Prasad, et al. 1994. Eutherian mammals from the Upper Cretaceous (Maastrichlian) Intertrappen Beds of Naskal, Andhra Pradesh, India (Journal of Vertebrate Paleontology)
Deborah L. Rook & John P. Hunter, 2013. Rooting Around the Eutherian Family Tree: the Origin and Relations of the Taeniodonta (Journal of Mammalian Evolution)
Irina Ruf, et al. 2016. Digital reconstruction of the inner ear of Leptictidium auderiense (Leptictida, Mammalia) and North American leptictids reveals new insight into leptictidan locomotor agility (Paläontologische Zeitschrift)
Mary T. Silcox, et al. 2010. Cranial anatomy of Paleocene and Eocene Labidolemur kayi (Mammalia: Apatotheria), and the relationships of the Apatemyidae to other mammals (Zoological Journal of the Linnean Society)
Steven C. Sweetman, et al. 2017. Highly derived eutherian mammals from the earliest Cretaceous of southern Britain (Acta Palaeontologica)
Paúl M. Velazco, et al. 2022. Combined data analysis of fossil and living mammals: a Paleogene sister taxon of Placentalia and the antiquity of Marsupialia (Cladistics)
Hai-Bing Wang & Yuanqing Wang, 2023. Middle ear innovation in Early Cretaceous eutherian mammals (Nature Commnications)
Hai-Bing Wang, et al. 2022. A new mammal from the Lower Cretaceous Jehol Biota and implications for eutherian evolution (Phil. Trans. R. Soc. B)
John R. Wible & Ornella C. Bertrand, 2024. Basicranial Anatomy of Leptictis haydeni Leidy, 1868 (Mammalia, Eutheria, Leptictidae) (Annals of the Carnegie Museum)
Mammalian diversity today consists of three major lineages: the egg-laying monotremes, the pouched marsupials, and the placental mammals. In deep time, there were many other groups which evolved from the same common ancestor of mammals and their relatives, but left no descendants today. Multituberculates ('multis' from hereon out) constituted one of the most successful of these groups, and even managed to make it through the great mass extinction at the end of the Cretaceous Period, from which they well and truly flourished.
Anatomy and Phylogeny
The clade name Multituberculata means "many small tubers" in Latin, and refers to one of their diagnostic features: the premolar and molar teeth are studded with multiple tubercles or cusps (Prothero, 2017). The premolars in particular have been described as "blade-like", and in some forms, like Ptilodus, the last of the these are greatly enlarged. On average, multis have three-five premolar teeth and only two molar teeth on each side of the jaws. Similar to rodents and lagomorphs (rabbits & kin), the incisor teeth were long and curved (though they did not continuously grow), and they were separated from the cheek teeth by a gap or diastema. There were no canines. This would make the dental formula 1-3:0:3-5:2.
Lower jaw or mandible from Ptilodus (James W. Gidley/Smithsonian, Public Domain)
This peculiar rodent-like dentition is mirrored in the rest of the skull, in which the eye-sockets or orbits were placed on the sides of the cranium, and the zygomatic arch or cheek bones are wide and strong to accommodate large chewing muscles. But the similarities end there: analysis of the jaw joints indicates that they chewed with the mandible moving in a forwards-&-backwards motion, rather than the side-to-side or backwards-&-forwards motion typical of mammals today (Kemp, 2005). By looking at the wear patterns on teeth, it seems that multis did not gnaw their food but rather grinded and sliced it, and so may have had a broadly omnivorous diet that included small animals and soft fruits as well as hard nuts and seeds. There is even evidence that multis chewed on bones of dead dinosaurs! (Skutschas, et al. 2025).
From a postcranial perspective, multi skeletons reveal a wide array of forms resembling modern mammals like possums and squirrels, both ground and tree-dwelling varieties. More on this diversity will be elaborated in the next section.
Multis belong to a larger clade of mammals called Allotheria, whose other members also sported rodent-like skulls and diverse bodyplans. Recent studies indicate that the multi's closest relatives were the Gondwanatheria and perhaps the Euharamiyida (Hoffmann, et al. 2020), but a minority of researchers have proposed that this entire group is polyphyletic with its members spread across the mammal family tree (King & Beck, 2020).
It has generally been argued that multis evolved from the same common ancestor of the crown mammals (that is, the ancestor from which all living mammals descend), and this has been based on studies looking at the limited sample of complete skeletal remains on hand (Hoffmann, et al. 2020). In particular, they would be more closely allied to the live-bearing mammals or Theria, than the monotremes. More recent studies, looking at tooth morphology, have argued for placement outside of crown Mammalia, but not by much (Mao, et al. 2024). All in all, this suggests that the multis and other allotherians shared enough characters with early crown mammals that, regardless of where they are on the family tree, they would have closely resembled their common ancestor in key aspects. By all accounts, the multis have been considered "a very conservative group" on the basis of their skeletal morphology (Kemp, 2005).
Phylogeny of mammals & their relatives, based on dentition (Mao, et al. 2024)
The exact relationship of multis to living mammals would shed some very fascinating light on has been discovered about their life histories. It had previously been argued that multis gave birth to live young, because the pelvis girdle was built in such a way that could only support viviparity (Kielan-Jaworowska, 1979). The bone tissues of the mammalian femoral cortex (the outside layer of the femur) contain microstructures that show a correlation between proportional size and how long the lactation period is: a 2022 paper found that the time between birth and weaning in multis was most similar to placental mammals, and not marsupials (Weaver, et al. 2022). Thus, like placentals, the young were fed on milk for a brief period of time and could move freely on their own soon after birth, whereas marsupial mammals give birth to underdeveloped young that must move to the pouch and stay latched on a teat for an extended period of time to finish their growth.
It is remarkable to consider - should further evidence show multis as outside the crown mammals - that such life histories evolved convergently between the Allotheria and Placentalia. If more research ends up confirming the older model of multis as nearest to live-bearing mammals, then it could mean that the placental-life history represents the ancestral condition for therians, but convergence would still be a consideration on the table (Weaver, et al. 2022).
Diversity
The oldest known fossils of multis date back to the Middle Jurassic Bathonian age and span two regions: the Forest Marble Formation of Oxfordshire and Dorset in the UK (genera Kermackodon, Kirtlingtonia, & Hahnotherium) and the Itat Formation of Western Siberia, Russia (Tashtykia & Tagaria), both Eurasian (Butler & Hooker, 2005; Averianov, et al. 2020). In both cases, the fossils consist mainly of teeth - which is nothing new in paleomammology - but given the dental distinctions discussed previously, an alliance to the multis was fairly easy to make. In fact, the two key observations were made on this material, A) there is a lot of overlap in structure between these multi teeth and those of the Euharamiyida clade, cementing the possibility of a link between these groups, and B) the sheer diversity in tooth morphology in relation to the euharamiyidans suggests a potential Norian Age (Late Triassic) origin of the group (Averianov, et al. 2020).
Like many fossil vertebrate groups, the most plesiomorphic (with more ancestral traits) and the most apomorphic (with more derived traits) multis have been classified into two groups: the Plagiaulacida and Cimolodonta. However, in the most recent phylogenetic analysies, this diversity is better understood with the plesiomorphic multis forming a series of lineage splits prior to the evolution of the Cimolodonta clade (Carvalho, et al. 2025).
Restoration of Cambelodon by Victor Carvalho (from Carvalho, et al. 2025)
The earliest diverging multis consist of two sister groups: the Pinheirodontidae and the Paulchoffatiidae. The oldest known fossils of these date to the Late Jurassic, while the youngest died out by the Early Cretaceous. Pinheirodonts include the newly described Cambelodon and are primarily European, found in fossil sites within the Iberian Peninsula. All we have so far from this group are isolated teeth and a partial mandible. Paulchoffatiids are also Iberian but have representatives in England and Germany. The genus Paulchoffatia proper has a particularly "ancient" dentition, with the premolar teeth being rectangular in shape and more suited for crushing food than most later multis (Kemp, 2005). As well the molar teeth bore fewer cusps than what was to come.
From there a number of genera diverge, showing that by the Late Jurassic the multis had spread to North America while land connections were still established with Europe. These forms include Ctenacodon, which coexisted with the giant dinosaurs of the Morrison Formation of the northwest US states. Like earlier forms, Ctenacodon had the low-cusp molar condition, but now the premolar teeth had gained the long, shearing shape typical of the group (Kemp, 2005). Plagiaulax, from the English Early Cretaceous, was the first multi described by modern paleontology, having been named in 1857 by Hugh Falconer (a pioneer of the field in many respects).
The clade Eobaataridae was the next to evolve and are a primarily Early Cretaceous group of Eurasia. It's at this point in the fossil record that we have gotten far more complete remains, including postcrania. The genus Sinobaatar was recovered from the prolific Yixian Formation of Liaoning, China and revealed limb bones and digits (as well as teeth) that were less derived than the later cimolodonts (Yaoming & Yuanqing, 2002). One fascinating insight from the related genus Jeholbaatar concerns the evolution of mammalian ears. In this fossil, the middle ear bones were preserved but the way they are built in relation to the mandible or lower jaw joints suggests that they evolved this condition independently of other mammals like ourselves (Wang, et al. 2019). It's likely that it had something to do with facilitating the forewards-to-backwards chewing motion aformentioned in this article.
Thus we come to the Cimolodonta or "derived multis". They evolved in the Aptian Age, towards the end of the Early Cretaceous, would flourish into a wide variety of forms, and would reach their maximum geographic distribution. One genus, Corriebaatar, even reached Australia, likely from South America via Antarctica considering its closest relative was the Argentine Ferugliotherium (Rich, et al. 2009; Rich, et al. 2022). Cimolodonts have lost their first upper incisor teeth and reduced their number of premolar teeth to four upper and two lower teeth (Kemp, 2005). It's in these groups that the shearing premolars really come into their own.
The interrelationships of the cimolodonts has yet to be properly resolved, but a number of subgroups have been made out:
Fossil skull of Cimolomys (Albert C. Silberling/Smithsonian, Public Domain)
The Cimolodontidae, Neoplagiaulacidae, & Ptilodontidae clades are sometimes considered close relatives and were common throughout the Late Cretaceous Epoch & Paleogene Period of North America and Eurasia. The ptilodonts in particular had large, arch-shaped lower premolars, and the most familiar genus (indeed one of the most familiar multis) was Ptilodus proper (Kemp, 2005). Multiple species are known, and we have some good fossils that reveal a small omnivorous mammal with a powerful, flexible (perhaps prehensile) tail and tarsal or hindfoot bones which could rotate, allowing it to maneuver through tree branches (Jenkins & Krause, 1983).
The Cimolomyidae are found in North America as well as East Eurasia (Mongolia), suggesting they utilized the Beringian corridor for dispersal. They appear to have been quite common in the Late Cretaceous, with many genera approaching the weight of an average rabbit (Williamson, et al. 2015). Representative genera include Cimolomys proper (shown above), Essonodon, & Meniscoessus.
There are a number of poorly-known multi groups, including the Eucosmodontidae and Microcosmodontidae, all of which lived from the Late Cretaceous and into the Eocene Epoch of Europe and North America.
The Kogaionidae, in contrast, has become more familiar to paleontologists. In particular, they appear to have inhabited the now lost Hațeg Island of Late Cretaceous Romainia as the predominant mammal fauna (Codrea, et al. 2016). There is variation in the known tooth morphology that suggests a diversity in form within individual species (what is known as intraspecific variation) as well as between species. The genus Barbatodon sported red iron-pigmented enamel on their incisors and cheek-teeth cusps, much like some species of soricine shrews which is indicative of a diet of hard crunchy insects like beetles (Smith & Codrea, 2015). The genus Litovoi was even found to have a domed skull, highly-tuned sensory organs, and "one of the smallest brains relative to body size of any advanced mammaliaform" (Csiki-Sava, et al. 2018).
There are two particularly large groups of multis that represent some of the last known varieties and share a common ancestry, perhaps including the kogaionids, united by self-sharpening incisor teeth (Kemp, 2005).
One clade is the Djadochtatherioidea, which was primarily from Central Eurasia but also represented in European and American fossils from the Late Cretaceous through the Paleocene Epoch. These multis had evolved modifications in their jaw muscles which stretched them far forward on the mandible, giving the group a larger and more directed bite force (Kemp, 2005). Two related genera, Bulganbaatar & Nemegtbaatar, have had differing interpretations of their gait by different teams: these animals may have walked in a parasagittal-gait like living opossums or they may have been saltatory like gerbils, hopping on their hindlimbs and using their forelimbs as shock-absorbers (Kemp, 2005). This hopping motion has been more readily demonstrated in Kryptobaatar and Catopsbaatar, however (Chen & Wilson, 2015).
The other clade is the Taeniolabidoidea, from the Late Cretaceous to the Paleocene of North America and East Asia, which sported boxy skulls and had a tendancy to grow to very large sizes. Taeniolabis proper, from the North American Paleocene, was comparible in size to a beaver and likely was a ground-dwelling herbivore (Prothero, 2017). In contrast, Lambdopsalis from the Paleocene of East Asia was a fossorial or burrowing genus which sported a flat and broad head and fused cervical or neck vertebrae - features akin to moles (Kemp, 2005). This genus is particular is quite special in that we have a fossil coprolite from some unknown carnivore which contained the remains of its hair which forms impressions around the animal's bones (Meng & Wyss, 1997). Depending on where the Allotheria lay on the mammaliamorph family tree, their presence here is at least confirmation that the common ancestor of all these groups was furry.
By the Late Cretaceous, multis were widespread and represented by over six different clades. They witnessed the end of the world as the massive bolide struck the Yucatán Peninsula and closed the Mesozoic Era. Remarkably, save for the cimolomyids, the multis survived alongside the ancestors of the monotreme, marsupial, and placental mammals (they were one of the only other mammal groups to do so, in fact, for several other lineages died out during the mass extinction). There is even evidence of species on both sides of the K-Pg Boundary (Longrich, et al. 2016). Multis not only survived, but flourished in the aftermath: they form particularly common fossil in Early Paleogene sites in Europe and North America (Prothero, 2017). In many ways, their peak diversity was reached during the initial Paleocene Epoch, and the largest representatives evolved during this time (like Taeniolabis). However, by the latest Eocene Epoch, their numbers dwindled and they all went extinct.
Extinction
What happened to the multis? How could a group that was alive through the End Cretaceous mass extinction event and flourishing in the Paleogene Period then just die out?
Traditionally, it has been argued that the newly-evolving placental mammals, especially rodents and perhaps early primate-relatives out-competed them ecologically (Prothero, 2017). They had very similar skeletal anatomies, after all, and there is an apparent correlation in timing between the rise of rodents and the decline in multis during the Paleogene of Eurasia (Krause, 1986). Close comparisons between the two groups of mammals have given differing estimates of chewing biomechanics, with multis sporting lower bite forces but greater bite speed and rodents having higher bite forces but higher stresses on the skulls (Adams, 2019). Though it is suspected that multis had a broad diet, perhaps the rodents were able to better process the same foods, given that their lengthy incisor teeth could continuously grow. Not all paleomammologists accept that competition explains everything, and some have even argued that direct testing of such a hypothesis is "usually quite impossible" (Kemp, 2005).
For one, the appearance of competition in the fossil record is not so clear cut. Surveys of known fossil records show that while in Europe and North America the decline of multis is then followed by a rise in rodents, in Eastern Eurasia the diversity of the two groups fluctuated in parallel, and that there may simply have been available niches that rodents filled at the expense of the multis (Wood, 2010). As well, in several North American sites there is an overlap of 15 million years where multis & early rodents coexisted, with the multis inhabited fairly generalized niches (Ostrander, 1984). This has led some authors to conclude that the situation is more complicated, and that a closer look at climate and environment would provide better clues.
The Earth underwent a gradual cooling and drying of the climate at the end of the Eocene Epoch as the continent of Antarctica began to grow its massive glaciers. There is evidence of associated environmental shifts and faunal turnovers on land and in the seas (as outlined in my prior post on extinction events). In a recent 2025 analysis, it was observed that multis flourished most commonly in moist, temperate forests dominated by Metasequoia, Glyptostrobus, alders, and cone nut trees, while rodents tended to favor a wider range of habitats including oak, elm, pine, & chestnut forests (Burger, 2025). There is evidence that the former habitats shrank in range as the global climate cooled, and so the extinction of the multis is implicated to have corresponded more to environmental change than direct competition by early placentals like rodents. So the rise of these placentals would simply have been more about occupying new landscapes that multis couldn't survive in, instead of directly pushing them out of their niches. It is important to remind readers that many mammal groups - not just multis - experienced turnovers and extinctions at the Eocene-Oligocene boundary.
The reality of the situation is that more research on more fossil sites & remains needs to be conducted before any one hypothesis can be confirmed or refuted. We simply have no consensus on why the multis went extinct. There can be no doubt, however, that this was a long-lived and highly diverse group of mammals that should in no way ever be considered "failures" on the evolutionary stage.
Book References
T. S. Kemp. The Origin and Evolution of Mammals (Oxford University Press, 2005)
Donald Prothero. The Princeton Field Guide to Prehistoric Mammals (Princeton University Press, 2017)
Paper Citations
Neil F. Adams, et al. 2019. Functional tests of the competitive exclusion hypothesis for multituberculate extinction (The Royal Society Publishing)
Alexander O. Averianov, et al. 2020. Multituberculate mammals from the Middle Jurassic of Western Siberia, Russia, and the origin of Multituberculata (Papers in Palaeontology)
Benjamin John Burger, 2025. Comparative spatial paleoecology: assessing niche competition between Eocene North American multituberculates and rodents regarding forest resources to elucidate the cause of multituberculate extinction (Paleobiology)
Percy M. Butler & Jerry J. Hooker, 2005. New teeth of allotherian mammals from the English Bathonian, including the earliest multituberculates (Acta Palaeontologica)
Victor F. Carvalho, et al. 2025. Cambelodon torreensis, a new pinheirodontid multituberculate from the Upper Jurassic of western Portugal (Papers in Paleontology)
Meng Chen & Gregory P. Wilson, 2015. A multivariate approach to infer locomotor modes in Mesozoic mammals (Paleobiology)
Vlad Aurel Codrea, et al. 2016. First mammal species identified from the Upper Cretaceous of the Rusca Montană Basin (Transylvania, Romania) (Comptes Rendus Palevol)
Zoltán Csiki-Sava, et al. 2018. Dome-headed, small-brained island mammal from the Late Cretaceous of Romania (PNAS)
Simone Hoffmann, et al. 2020. Phylogenetic placement of Adalatherium hui (Mammalia, Gondwanatheria) from the Late Cretaceous of Madagascar: implications for allotherian relationships (Journal of Vertebrate Paleontology)
Z. Kielan-Jaworowska, 1979. Pelvic structure and nature of reproduction in Multituberculata (Nature)
Benedict King & Robin M. D. Beck, 2020. Tip dating supports novel resolutions of controversial relationships among early mammals (Proc. R. Soc. B.)
F. A. Jenkins & David W. Krause, 1983. Adaptations for climbing in North American multituberculates (Mammalia) (Science)
David W. Krause, 1986. Competitive exclusion and taxonomic displacement in the fossil record: The case of rodents and multituberculates in North America (Vertebrates, Phylogeny, & Philosophy)
Nick R. Longrich, et al. 2016. Severe extinction and rapid recovery of mammals across the Cretaceous–Palaeogene boundary, and the effects of rarity on patterns of extinction and recovery (Journal of Evolutionary Biology)
Fangyuan Mao, et al. 2024. Jurassic shuotheriids show earliest dental diversification of mammaliaforms (Nature)
Jin Meng & André R. Wyss, 1997. Multituberculate and other mammal hair recovered from Palaeogene excreta (Nature)
Gregg E. Ostrander, 1984. The Early Oligocene (Chadronian) Raben Ranch Local Fauna, Northwest Nebraska: Multituberculata; with Comments on the Extinction of the Allotheria (Transactions of the Nebraska Academy of Sciences and Affiliated Societies)
Thomas H. Rich, et al. 2009. An Australian multituberculate and its
palaeobiogeographic implications (Acta Palaeontologica)
Thomas H. Rich, et al. 2022. Second specimen of Corriebaatar marywaltersae from the Lower Cretaceous of Australia confirms its multituberculate affinities (Acta Palaeontoloiga)
Pavel P. Skutschas, et al. 2025. Evidence of osteophagia in Mesozoic mammals: multituberculate tooth marks on a hadrosaurid maxilla from the Late Cretaceous of the Russian Far East (Historical Biology)
Thierry Smith & Vlad Codrea, 2015. Red Iron-Pigmented Tooth Enamel in a Multituberculate Mammal from the Late Cretaceous Transylvanian “Haţeg Island” (PLOS One)
Haibing Wang, et al. 2019. Cretaceous fossil reveals a new pattern in mammalian middle ear evolution (Nature)
Lucas N. Weaver, et al. 2022. Multituberculate Mammals Show Evidence of a Life History Strategy Similar to That of Placentals, Not Marsupials (The American Naturalist)
Thomas E. Williamson, et al. 2015. A new taeniolabidoid multituberculate (Mammalia) from the middle Puercan of the Nacimiento Formation, New Mexico, and a revision of taeniolabidoid systematics and phylogeny (Zoological Journal of the Linnaean Society)
Joseph D. Wood, 2010. Wood, D. Joseph (2010). The Extinction of the Multituberculates Outside North America: a Global Approach to Testing the Competition Model (Thesis - Ohio State University)
Hu Yaoming & Wang Yuanqing, 2002. Sinobaatar gen. nov.: First multituberculate from the Jehol Biota of Liaoning, Northeast China (Chinese Science Bulletin)
The first page from PALEOCENE #5. Sixty-six million years ago, our tiny primate ancestors survived an asteroid impact that killed off most life on Earth. What happened next?
Find out! Support the crowdfunding campaign: https://www.kickstarter.com/projects/keesey/paleocene-5-comic-book?ref=cescdw
Sixty-six million years ago, an asteroid impact wiped out most life on Earth. How did our ancestors survive?
[Unfortunately, as of this publication, no one has tried to claim the prize money offered (Figures 1 and 2) in the search for the Smilodon baculum at La Brea. Maybe the most iconic sabertooth did not have an ossified baculum. Maybe none of the machairodonts did. But if anyone ever does find one, I've got a crisp $100 bill that I'd be ecstatic to send them!]
‘While bats can only sense the outer shapes and textures of their targets, dolphins can peer inside theirs. If a dolphin echolocates on you, it will perceive your lungs and your skeleton. It can likely sense shrapnel in war veterans and fetuses in pregnant women. It can pick out the air-filled swim bladders that allow fish, their main prey, to control their buoyancy.
It can almost certainly tell different species apart based on the shape of those air bladders. And it can tell if a fish has something weird inside it, like a metal hook. In Hawaii, false killer whales often pluck tuna off fishing lines, and “they’ll know where the hook is inside that fish,” Aude Pacini, who studies these animals, tells me. “They can ‘see’ things that you and I would never consider unless we had an X-ray machine or an MRI scanner.”
This penetrating perception is so unusual that scientists have barely begun to consider its implications. The beaked whales, for example, are odontocetes that look dolphin-esque on the outside—but on the inside, their skulls bear a strange assortment of crests, ridges, and bumps, many of which are only found in males.
Pavel Gol’din has suggested that these structures might be the equivalent of deer antlers—showy ornaments that are used to attract mates. Such ornaments would normally protrude from the body in a visible and conspicuous way, but that’s unnecessary for animals that are living medical scanners.’
Apatemyids were a group of unique early placental mammals that lived during the first half of the Cenozoic, known from North America, Europe, and Asia. Due to their specialized anatomy their evolutionary relationships are rather murky (they were traditionally part of the convoluted mess that was "Insectivora"), but currently they're thought to be a very early offshoot of the Euarchontoglires, the branch of placentals that includes modern rodents, lagomorphs, treeshrews, colugos, and primates.
Living in what is now western Europe during the mid-Eocene, around 47 million years ago, Heterohyus nanus was a small apatemyid about 30cm long (~12") – although just over half of that length was made up of its tail.
Like other apatemyids it had a proportionally big boxy head, with large forward-pointing rodent-like incisors in its lower jaw and hooked "can-opener-shaped" incisors in its upper jaw.
Example of an apatemyid skull from the closely related American genus Sinclairella.
From Samuels, Joshua X. "The first records of Sinclairella (Apatemyidae) from the Pacific Northwest, USA." PaleoBios 38.1 (2021). https://doi.org/10.5070/P9381053299
The rest of its body was rather slender, and fossils with soft tissue preservation from the Messel Pit in Germany show that it had a bushy tuft of longer fur at the end of its long tail.
But the most distinctive feature of apatemyids like Heterohyus were their fingers, with highly elongated second and third digits resembling those of modern striped possums and aye-ayes. This suggests they had a similar sort of woodpecker-like ecological role, climbing around in trees using their teeth to tear into bark and expose wood-boring insect holes, then probing around with their long fingers to extract their prey.
———
NixIllustration.com | Tumblr | Patreon
References:
Kalthoff, D. C., W. Von Koenigswald, and C. Kurz. "A new specimen of Heterohyus nanus (Apatemyidae, Mammalia) from the Eocene of Messel (Germany) with unusual soft part preservation." Courier Forschungsinstitut Senckenberg 252 (2004): 1-12. https://www.researchgate.net/publication/263714512_A_new_specimen_of_Heterohyus_nanus_Apatemyidae_Mammalia_from_the_Eocene_of_Messel_Germany_with_unusual_soft-part_preservation
Koenigswald, W. V., and H-P. Schierning. "The ecological niche of an extinct group of mammals, the early Tertiary apatemyids." Nature 326.6113 (1987): 595-597. https://www.researchgate.net/publication/232761846_The_ecological_niche_of_early_Tertiary_apatemyids_-_extinct_group_of_mammals
Samuels, Joshua X. "The first records of Sinclairella (Apatemyidae) from the Pacific Northwest, USA." PaleoBios 38.1 (2021). https://doi.org/10.5070/P9381053299
Silcox, Mary T., et al. "Cranial anatomy of Paleocene and Eocene Labidolemur kayi (Mammalia: Apatotheria), and the relationships of the Apatemyidae to other mammals." Zoological Journal of the Linnean Society 160.4 (2010): 773-825. https://doi.org/10.1111/j.1096-3642.2009.00614.x
[Summary. A piglet's most precious possession
Is the teat that he fattens his flesh on.
He fights for his teat with tenacity
Against any sibling's audacity.
The piglet, to arm for this mission,
Is born with a warlike dentition
Of eight tiny tusks, sharp as sabres,
Which help in impressing the neighbors;
But to render these weapons less harrowing,
Most farmers remove them at farrowing.
We studied pig sisters and brothers
When some had their teeth, but not others.
We found that when siblings aren't many,
The weapons help little if any,
But when there are many per litter,
The teeth help their owners grow fitter.
But how did selection begin
To make weapons to use against kin?]
Fraser & Thompson (1991)
H/T to an anonymous submitter for sending this in!
Eurotamandua joresi lived during the mid-Eocene, about 47 million years ago, in the lush subtropical forests that covered what is now central Germany.
When it was first described in the early 1980s it was classified as an anteater due to its close resemblance to some modern species… but there were big problems with this interpretation. Anteaters have a sparse fossil record, but they're known to have originated during the early Eocene in the isolated island continent of South America – so Eurotamandua's ancestors making it all the way to Europe within just a few million years would be pretty remarkable!
Also, on closer inspection it didn't have the distinctive skeletal features of a xenarthran mammal, suggesting it wasn't an anteater after all.
Instead more recent studies have identified it as a close relative of pangolins, part of an early branch of the group that didn't have the characteristic large scales.
About 90cm long (~3'), Eurotamandua would have a lifestyle much like the anteaters it convergently resembled, using its large claws to rip open ant nests and a long sticky tongue to feed.
The end is coming! Only three days left to back the next issue of PALEOCENE, the comic book series about our tiny, primate ancestors 66 million years ago, who survived the extinction that ended the reign of the dinosaurs.
It’s going away soon! Four days left to back the latest issue of PALEOCENE, a comic book series about our ancestors who survived the asteroid impact that ended the Age of Dinosaurs. https://www.kickstarter.com/projects/keesey/paleocene-4-comic-book
Only 5 days left to back the latest issue of PALEOCENE, a comic book series about our tiny primate ancestors who survived the asteroid impact that ended the Age of Dinosaurs.
Check it out: https://www.kickstarter.com/projects/keesey/paleocene-4-comic-book
Synapsid Notebook @synapsidnotebook - Tumblr Blog | Tumgag
