Paleolithic Migrations and Ice-Age Worlds
Summary
Big Era 2 is the long Paleolithic epoch during which anatomically modern humans peopled the globe. This chapter focuses on the where and how of human dispersal: hunter-gatherer band organization, the Out-of-Africa migration, the peopling of Eurasia, Australia, and the Americas (including the Bering Land Bridge, coastal migration routes, Sundaland–Sahul, Madjedbebe, the White Sands footprints, and Misliya Cave), the ice-age climate frame, the Last Glacial Maximum, and megafauna extinctions. Students will learn to read the recent evidence base — ancient DNA, dated footprint sites, submerged-landscape research — alongside foraging-lifeway concepts like ethnographic analogy and gender in foraging bands.
Concepts Covered
This chapter covers the following 20 concepts from the learning graph:
- Paleolithic Period
- Hunter-Gatherer Societies
- Band-Level Organization
- Out Of Africa Migration
- Peopling Of Eurasia
- Peopling Of Australia
- Peopling Of Americas
- Bering Land Bridge
- Ice Age Climate
- Last Glacial Maximum
- Megafauna Extinctions
- Stone Age Toolkits
- Coastal Migration Routes
- Ethnographic Analogy
- Foraging Lifeways
- Gender In Foraging Bands
- Misliya Cave
- Madjedbebe Site
- White Sands Footprints
- Sundaland And Sahul
Prerequisites
This chapter builds on concepts from:
- Chapter 1: Foundations of Historical Thinking
- Chapter 2: Cosmic and Biological Origins
- Chapter 3: Hominin Evolution and the Genus Homo
The longest road trip in history.
Welcome back. By the end of Chapter 3, Homo sapiens existed in Africa. By the end of this chapter, our species occupies every habitable continent except Antarctica, has crossed open ocean to reach Australia, walked across what is now the seafloor of the North Sea, and left footprints in New Mexico that should not, by the canonical textbook, exist yet. The Paleolithic is the longest and most consequential road trip in our species' history.
What "Paleolithic" Actually Names
The Paleolithic Period ("Old Stone Age") is the long stretch of human prehistory characterized by chipped stone tools, foraging subsistence, and the absence of agriculture, pottery, or permanent settlements. It begins, depending on whose definitions you accept, with the earliest stone tools (~3.3 Mya, Lomekwian) or with the genus Homo (~2.6 Mya, Oldowan), and it ends with the transition to the Neolithic that we will take up in Chapter 6 — at different dates in different regions, ranging from ~12,000 BP in Southwest Asia to as late as ~5,000 BP or later in some parts of the world.
The Paleolithic itself is conventionally divided into three subperiods. The Lower Paleolithic (~3.3 Mya – 300 kya) covers the long age of australopith and Homo erectus toolmaking. The Middle Paleolithic (~300 – 50 kya) corresponds to the world of Neanderthals and the early Homo sapiens expansion, with prepared-core (Mousterian) industries. The Upper Paleolithic (~50 – 12 kya) is the age of long blade industries, sophisticated bone and antler tools, and the explosion of figurative cave art and personal ornamentation that we will pick up in Chapter 5.
This chapter is mostly about the Middle and Upper Paleolithic experience of Homo sapiens — what foraging societies were like, how they were organized, and how, between roughly 70,000 and 12,000 years ago, they reached Australia, Beringia, the Americas, and most of habitable Eurasia.
Hunter-Gatherer Societies and Band-Level Organization
Throughout the Paleolithic, the universal human subsistence strategy was hunting and gathering — foraging wild plant foods, hunting and trapping animals, fishing, collecting shellfish and birds' eggs, and exploiting whatever else the local environment offered. Foragers did not store food on the scale that later agricultural societies did; they moved with the resources rather than fixing the resources to a place.
The political and social unit that organized foraging life is what anthropologists call the band — a small, mobile, kin-based group, typically 15 to 50 individuals, with no formal leadership, no specialized roles other than age- and sex-based divisions of labor, and very flat status hierarchies. Decisions were made by discussion and consensus. Disputes were resolved by mediation, by the disputants moving apart, or, in extreme cases, by violence — but the small scale of band life made anonymous large-scale conflict impossible. Band-level organization is the social form humans lived in for the overwhelming majority of our species' history, and many of our cognitive intuitions about fairness, gossip, reciprocity, and group membership are likely calibrated for it.
| Social Trait | Typical Forager Band Pattern |
|---|---|
| Group size | 15–50 (a "camp"); ~150 in regional networks |
| Leadership | Informal, situational, persuasion-based |
| Inequality | Low; food sharing is widely enforced |
| Mobility | High; moves measured in days/weeks per camp |
| Property | Mostly portable; communal access to land/water |
| Specialization | Age- and sex-based; no full-time specialists |
A small but important caveat. The picture above is reconstructed from a combination of archaeology and ethnographic analogy — drawing inferences about Paleolithic foragers from twentieth-century foraging societies (such as the Ju/'hoansi of southern Africa, the Hadza of Tanzania, or the Inuit of the Arctic). Ethnographic analogy is a powerful tool, but it has clear limits: modern foragers live in marginalized environments shaped by surrounding agricultural and industrial states, and they are not "living fossils." The Hadza are a contemporary people with as much history as anyone else, not a snapshot of the Pleistocene. Use the analogies; don't fetishize them.
Foraging Lifeways and Gender in Bands
The classic textbook framing — "Man the Hunter, Woman the Gatherer" — has been substantially revised over the past two decades. Detailed dietary analyses of contemporary foraging societies show that gathered foods (roots, tubers, nuts, fruits, small game, shellfish) typically supply the majority of dietary calories in non-Arctic environments, while large-mammal hunting, when successful, supplies high-quality but irregular protein. Skeletal evidence from Paleolithic populations shows widespread participation in physical labor across both sexes, and several recent ancient-DNA and isotopic studies have identified individuals biologically female buried with hunting equipment in contexts that suggest they used it.
A reasonable current framing is that the division of labor in foraging bands was real but flexible, organized more around childcare obligations and individual specialization than around rigid sex roles. Women who were not currently nursing infants could and did hunt; men contributed to gathering and to childcare. The Paleolithic was not a feminist utopia, but it was also not the rigidly gendered tableau of older textbook illustrations. Gender in foraging bands is best read from the evidence, not back-projected from later, agriculturally-organized societies.
A historiographical pivot worth noticing.
The "Man the Hunter" framing came out of a single 1966 conference of that name. It then dominated textbooks for decades — including textbooks written by people who knew the dietary statistics did not support it. The shift away from it is a textbook example of historiographical change: the evidence did not move much, but the framing did, as new generations of researchers asked different questions of the same data. When you encounter a confidently stated "everyone knew" claim about prehistoric people, ask which conference, which decade, which assumptions locked that claim in. That is bias and misinformation detection — applied to academia rather than to social media, but the same skill.
Ice-Age Climate and the Pleistocene Stage
The expansion of Homo sapiens across the globe happened on a stage we can no longer see: the ice-age world of the Pleistocene epoch (~2.58 Mya – ~11,700 years ago). The Pleistocene is characterized by repeated glacial-interglacial cycles — long periods of cold (glacials) interrupted by shorter warm periods (interglacials) — driven primarily by orbital changes (the Milankovitch cycles) modulated by atmospheric and oceanic feedbacks. During glacial maxima, vast continental ice sheets covered much of North America and northern Eurasia; during interglacials, ice retreated and forests expanded.
The cold side of the Pleistocene cycle had two consequences that are easy to forget. First, sea levels were much lower — locking up enough water in continental ice sheets to drop ocean levels by ~120 meters at the Last Glacial Maximum (LGM, ~26,500–19,000 years ago). This exposed land bridges and continental shelves that are now underwater: Beringia (between Siberia and Alaska), Doggerland (between Britain and continental Europe), and Sundaland (a vastly expanded Southeast Asian shelf). Second, much of the world's ice-free landmass was arid, treeless, and windy, with mammoth-steppe grasslands stretching across northern Eurasia and supporting megafaunal communities (mammoths, woolly rhinoceros, steppe bison, cave lions) that have no modern equivalent.
| Feature | Glacial (LGM) | Interglacial (Holocene) |
|---|---|---|
| Average global temperature | ~6 °C cooler | Modern baseline |
| Sea level | ~120 m lower | Modern baseline |
| Continental ice cover | Massive in N. Hemisphere | Greenland and Antarctica only |
| Sahara | Variable; partly more arid | Mostly hyper-arid |
| North Sea | Land (Doggerland) | Sea |
| Beringia | Land bridge | Submerged |
The Last Glacial Maximum is particularly important for this chapter. Many of the dispersal events we discuss either happened before the LGM (Australia, possibly the Americas) or rebounded after it. The LGM acted as both a barrier (ice sheets blocking interior North America) and an accelerator (lower seas exposing coastlines for movement).
Out of Africa: The First Successful Dispersal
Modern humans had been in Africa since at least 315,000 years ago (Chapter 3). They had also made earlier excursions out of Africa — including specimens at Misliya Cave in Israel dated to roughly 177,000–194,000 years ago, the oldest Homo sapiens fossils known outside Africa. But these earlier dispersals appear to have been dead ends that did not contribute substantial ancestry to modern non-African populations.
The genetic and archaeological evidence converges on a successful Out-of-Africa migration beginning roughly 70,000–60,000 years ago through a southern dispersal route — across the Bab el-Mandeb strait at the southern end of the Red Sea, then along the southern coast of Arabia and South Asia. This migrating population is the source of the bulk of non-African modern human ancestry. Within roughly the next 30,000 years, descendant populations reached western Europe, Central Asia, East Asia, and Australia. Ancient-DNA evidence shows that this expanding population picked up modest amounts of Neanderthal admixture in southwest Asia (~50–60 kya) and Denisovan admixture further east — evidence of ongoing biological encounter with the other hominin species we will pick up in Chapter 5.
The earlier "African Eden" model often presented this dispersal as a single wave from a single source population. The current best inference, after a decade of paleogenomic work, is more structured — a main southern-route expansion supplemented by smaller earlier dispersals (some of which left genetic traces in Australasian populations) and by complex internal African migrations. As with hominin evolution, the model has moved from a clean ladder to a textured tree.
Diagram: Peopling of the World, Animated Map
Peopling of the World — interactive animated migration map
Type: map with timeline overlay
sim-id: peopling-of-world-map
Library: Leaflet
Status: Specified
Learning objective (Bloom: Understanding/Analyzing): The student can trace the major dispersal routes of Homo sapiens from Africa across the globe, identify the role of land bridges (Beringia, Sundaland) and coastal corridors, and recognize the dating uncertainties at each stage.
Visual structure. A world map at LGM sea level (sea level shifted ~120 m lower), showing exposed land bridges in pale brown. A time-slider at the bottom (200 kya at the left, present at the right) animates migration arrows as time advances. Major route arrows: Out-of-Africa southern route at ~70 kya; Sundaland-Sahul crossing at ~65 kya (Madjedbebe); Eurasian expansion through Levant; Beringia entry by ~20–25 kya; coastal Pacific route into the Americas; possible pre-LGM entries (White Sands ~22.8 kya).
Interactivity. (1) Time-slider drag updates arrow visibility and labels. (2) Clicking a route shows source population estimates (where known from ancient DNA) and key sites along the route. (3) Toggle: switch between modern coastline and LGM coastline — the "aha" moment when students realize Australia was a single landmass with New Guinea and the Bering Strait was a thousand-kilometer-wide grassland.
Default layout. Responsive Leaflet map; minimum height 600 px. Default coastline at LGM; toggle to modern coastline.
Color palette. Migration arrows by date band: pre-LGM #6A1B9A (purple), LGM #B71C1C (red), post-LGM #1565C0 (blue), Holocene #2E7D32 (green). Land bridges #BCAAA4 (sandy tan). Ice sheets #ECEFF1 (off-white).
Implementation: Leaflet with custom layers for paleocoastlines and animated polyline plugin; data in data.json. Deploy at docs/sims/peopling-of-world-map/.
Peopling Eurasia
Within Eurasia, modern humans spread quickly. Sites in the Levant (Manot Cave ~55 kya), in eastern Europe (Bacho Kiro, Bulgaria, ~45 kya), in western Europe (Grotte Mandrin, France, ~54 kya), and in East Asia (Tianyuan Cave, China, ~40 kya) document the rapid expansion of Homo sapiens across the Eurasian landmass during a period when Neanderthals still occupied much of Europe and Denisovans occupied parts of central and East Asia. Within roughly 10,000–15,000 years, Neanderthals had largely disappeared as a distinguishable population, surviving only as a small genetic legacy in modern humans. The reasons for Neanderthal disappearance are debated — competition with modern humans, climate-driven population pressure on already-small Neanderthal groups, absorption through interbreeding, or a combination — but the timing is clear: Neanderthals as a coherent population end by roughly 40,000 years ago.
The stone age toolkits of these expanding populations evolved rapidly during the Upper Paleolithic (~50–12 kya). Long, parallel-sided blades replaced flake-based industries; specialized tools for hunting, hide-working, and bone-working appeared; and regional traditions — Aurignacian, Gravettian, Solutrean, Magdalenian in Europe — succeeded one another with technological elaboration we will examine more closely in Chapter 5. The point to register here is that technological diversity exploded as soon as modern humans were the dominant species across Eurasia.
Sundaland, Sahul, and Madjedbebe
The most striking single fact about Paleolithic dispersal may be the peopling of Australia. At the LGM, Southeast Asia was an enormously expanded continental shelf called Sundaland (modern Indonesia, Malaysia, and surrounding seas), exposed because of low sea levels. Australia was joined to New Guinea and Tasmania to form a single continent called Sahul. But Sundaland and Sahul were never connected — they were always separated by deep oceanic straits, with islands in between (the Wallacean archipelago) requiring multiple open-water crossings, the longest probably 70–100 km of open ocean.
The earliest unambiguous Homo sapiens occupation of Sahul comes from Madjedbebe rock shelter in northern Australia, where 2017 redating placed human occupation at approximately 65,000 years ago — pushing the peopling of Australia back by roughly 15,000 years from previous estimates and arriving uncomfortably close to the Out-of-Africa departure date. The site preserves grindstones, ochre crayons, and a sophisticated stone-tool kit. The implications are enormous: people reached Australia by deliberate maritime crossing in the Middle Paleolithic, well before the Upper Paleolithic technological flourishing in Europe. Whatever was happening in Europe later was not a prerequisite for the cognitive sophistication required to plan, build, and crew an oceangoing watercraft.
Some of the Madjedbebe dating has been challenged, with a few researchers arguing for a younger age based on questions about luminescence dating of sand grains. The current consensus accepts the ~65 kya date with appropriate caveats about ongoing methodological debate. This is a case where the honest summary is "the strong claim is ~65 kya; the conservative claim is at least ~55 kya; the field is still working it out."
How to read a contested date.
When a textbook (including this one) reports a date as ~65 kya for Madjedbebe or ~22.8 kya for White Sands, the date is doing two jobs: it summarizes the current strongest claim in the literature, and it often glosses over an active methodological debate. A useful student practice: when a date in this chapter feels surprising, type the site name into a scholarly database and look at the most recent reply paper. The disagreement, when there is one, is part of what makes the science live. This is one of your superpowers — recognizing that "we know" almost always has a "currently" attached.
Peopling the Americas: Beringia, the Coast, and White Sands
The peopling of the Americas has been one of the most aggressively rewritten chapters of Paleolithic prehistory in the last fifteen years. The classical model — known as "Clovis-first" after the Clovis projectile-point industry — held that humans entered the Americas through an ice-free corridor between the Laurentide and Cordilleran ice sheets at roughly 13,000 years ago, then spread quickly southward. That model is now untenable on multiple grounds.
The current best model has three components. First, humans entered eastern Beringia (now Alaska and Yukon) from northeast Asia, probably during a window after 23,000 years ago, possibly earlier. Second, the interior ice-free corridor was not biologically viable until well after the LGM (probably after ~13.4 kya), so the earliest entrants likely moved along the Pacific coast — a coastal migration route with island-hopping and shoreline travel that exploited marine resources. Third, and most decisively, several pre-Clovis sites now have credible dates: Monte Verde in Chile at ~14,500 years ago; Cooper's Ferry in Idaho at ~16,000 years ago; and most spectacularly the White Sands footprints in New Mexico, dated to approximately 22,800 years ago in a 2021 publication and reaffirmed by independent dating methods in 2023.
The White Sands footprints are 60+ human footprints preserved in lakeshore sediments. They show adults, adolescents, and children walking near a now-vanished lakeshore. If the dates hold — and they have survived two rounds of independent verification — humans were in the American Southwest at the height of the LGM, roughly 10,000 years before the canonical "first peopling" date. This requires either an entry into the Americas before the LGM (with movement during a very narrow demographic window) or a substantial revision of when the entry occurred. Either way, "Clovis-first" is dead; how exactly to replace it is a live research question.
| Site | Date (Approx.) | Significance |
|---|---|---|
| Bluefish Caves (Yukon) | ~24 kya | Earliest possible human presence in Beringia |
| White Sands footprints (NM) | ~22.8 kya | Pre-LGM occupation of interior North America |
| Cooper's Ferry (ID) | ~16 kya | Pre-Clovis Pacific Northwest occupation |
| Monte Verde (Chile) | ~14.5 kya | Pre-Clovis southern South America |
| Clovis sites (widespread) | ~13.0 kya | Distinctive fluted-point industry; once thought earliest |
Diagram: Bering Land Bridge MicroSim
Bering Land Bridge — interactive sea-level MicroSim
Type: MicroSim
sim-id: bering-land-bridge
Library: p5.js
Status: Specified
Learning objective (Bloom: Applying/Analyzing): The student can manipulate global sea level over the last 30,000 years and observe how the Bering Land Bridge appeared, expanded, and disappeared. They will be able to identify which dispersal-route hypotheses are viable at which sea-level state.
Visual structure. A 900 px wide responsive canvas (with updateCanvasSize() first in setup()). Top region: a stylized map of northeast Asia, Beringia, and northwest North America. Below: a slider for time (30 kya to present) and a slider for sea level (-130 m to 0 m, calibrated to the time slider but adjustable). As the user moves the time slider, sea level animates to its known historical value at that time, and the Beringia land bridge expands/contracts visibly. The Cordilleran and Laurentide ice sheets also expand and contract as a function of time. A small panel shows the current ice-free corridor status (blocked/partially open/open), the current coastal route status, and the current archaeological "tickmarks" (Bluefish, White Sands, Monte Verde, Clovis) lighting up as their dates are passed.
Controls. createSlider() for time; createSlider() for sea level (linked to time by default; freeable with a "manual override" createCheckbox()). createButton('Reset'). createButton('Show coastal route') toggles the coastal-corridor highlight. createButton('Show interior route') toggles the ice-free-corridor highlight.
Behavior. Sea-level history sourced from a Late-Pleistocene reconstruction (Lambeck et al. or similar) and stored in data.json. As time advances, archaeological sites with established dates light up in chronological order. A small "viability check" panel updates dynamically: at LGM, the interior corridor displays "BLOCKED — biologically inviable"; at 13 kya, displays "OPEN."
Implementation: p5.js, canvas.parent(document.querySelector('main')), all controls native p5.js. windowResized() triggers responsive recompute. Deploy at docs/sims/bering-land-bridge/.
Megafauna Extinctions
Across most of the world's continents, the late Pleistocene saw the disappearance of much of the large-bodied animal fauna — the megafauna. Australian megafauna (giant kangaroos, Diprotodon, the marsupial lion Thylacoleo) declined sharply within a few thousand years of human arrival ~65–50 kya. North American megafauna (mammoths, mastodons, giant ground sloths, saber-toothed cats, short-faced bears, glyptodonts) collapsed in a window centered around 13,000–11,000 years ago, broadly overlapping with the Clovis horizon and with the Younger Dryas climate reversal. South American megafauna followed a similar pattern. Eurasian megafauna (woolly mammoth, woolly rhinoceros, cave bear, Irish elk) declined more gradually but were largely gone by the early Holocene, with the woolly mammoth surviving in isolated refugia (Wrangel Island) until ~4,000 years ago.
The causes of these megafauna extinctions are intensely debated, with two main hypotheses and a sensible synthesis. The overkill hypothesis attributes extinctions primarily to human hunting pressure on naïve, slow-reproducing large animals. The climate hypothesis attributes them to rapid environmental change at the end of the Pleistocene, especially in North America where climate shifts during deglaciation were severe. The current synthesis is multi-causal: humans arrived in landscapes where megafauna had no co-evolved fear of bipedal predators, applied novel hunting pressure, and tipped already-stressed populations over edges they would otherwise have survived. In Australia, the human arrival date and the megafauna decline match too well for climate alone to be the explanation. In North America, the timing involves both human arrival and climate reversal. In Africa, where humans and megafauna co-evolved, megafauna largely persist — consistent with the idea that co-evolution (rather than naïveté) is the protective variable.
The implications for systems thinking are useful. A new predator entering a community in which prey have not been selected for vigilance against it is not a small perturbation; it is a regime change. Today, when an invasive species reaches an ecosystem with no co-evolved defenses, ecologists predict severe disruptions for the same structural reason. That is the same lesson the Pleistocene record teaches us, scaled to a continent.
Don't take the dichotomies too seriously.
A common pitfall in writing about megafauna extinctions is treating "humans did it" and "climate did it" as mutually exclusive. The honest answer in most regions is "both, with weights that vary by continent." When you see a public-facing science article that frames it as a dramatic either-or — humans bad vs climate guilty — that is framing, not science. The same caution applies to many modern controversies: real causes are usually compound. Single-villain explanations make for satisfying narratives and unreliable history.
What Went Where, and What It Means
By roughly 15,000 years ago, Homo sapiens had reached every habitable continent except Antarctica. They had crossed open ocean to Sahul, walked across grasslands that are now seafloor, hunted megafauna in the Mammoth Steppe, and left footprints at White Sands that should not, by yesterday's textbook, exist. They had done all this with band-level organization, with stone-age toolkits they refined regionally, and within the framework of an ice-age climate system that simultaneously enabled some of the dispersal (low sea levels, exposed bridges) and constrained other parts of it (interior corridors blocked by ice).
- The Paleolithic is the long age of foraging, chipped stone tools, and band organization.
- Hunter-gatherer bands of 15–50 individuals were the universal social unit; gender roles were real but flexible.
- Out of Africa ~70 kya through a southern route; Misliya Cave ~177–194 kya represents an earlier, dead-end excursion.
- Australia reached by 65 kya at Madjedbebe via deliberate ocean crossing.
- The Americas entered through Beringia and along the Pacific coast; White Sands footprints push interior North American occupation to ~22.8 kya, killing the Clovis-first model.
- Megafauna extinctions followed human arrival on most continents, with multi-causal contributions from hunting and climate change.
- Sea-level swings during the LGM exposed and submerged land bridges (Beringia, Sundaland-Sahul, Doggerland) that were essential to dispersal.
The next chapter picks up the symbolic and cultural side of the same Paleolithic world — the cave paintings, personal ornaments, and other-hominin encounters that share the stage with this dispersal story.
You have inhabited the planet.
From a few thousand modern humans in late-Pleistocene Africa to occupants of every habitable continent, in roughly 2,500 generations. Considering that nobody had a map, a written language, or anything most of us would recognize as a state, the Paleolithic dispersal is one of the most quietly impressive accomplishments in the whole archaeological record. You have now placed humans on the map. In the next chapter, we will hand them charcoal, pigment, and a wall.