Archaeology and traversing America’s pre-Columbian fault line

The world has long struggled to understand Indigenous America. The very name “American Indian” is a misnomer reflecting the confusion of a disoriented European explorer. Half a century ago, anthropologist Eric Wolf described a globe divided between the so-called civilized nations and noncivilized “people without history” (1). Uncertainties over the fault line separating the pre-Columbian (pre-1492) and Columbian periods continue to tangle our sense of the deep American past. The seminal PNAS study by Kelly et al. (2) demonstrates how a new proxy derived from massive radiocarbon datasets can help bridge the long-standing pre-Columbian-Columbian divide.

The frontier thesis of Frederick Jackson Turner held that colonizing American wilderness established a manifest destiny distinguishing American and European cultures (3). In so doing, Turner relegated “powerless” Indigenous Americans to the world of natural history shared by trilobites and extinct dinosaurs. In a sweeping counternarrative, Pekka Hämäläinen’s Indigenous Continent rejects this view of United States history as an irreversible colonial march toward Indigenous destruction (4). After a brief nod to the archaeology of pre-Columbian times, Hämäläinen describes a sovereign Indigenous world not as passive victims of colonial violence, but rather a dominating force that controlled the continent for centuries after the first European arrival. Ned Blackhawk’s (5) award-winning The Rediscovery of America: Native Peoples and the Unmaking of U.S. History continues that theme, no longer viewing the American past solely through the lens of European settlers, but with the understanding that Indigenous Americans were actively involved in societal stability/instability, wealth inequality, democracy, warfare, violence, innovations in technology, and social organization. Unlike Hämäläinen, Blackhawk sidesteps the archaeological past as irrelevant, warranting only the occasional anecdotal footnote. The unintended consequence of beginning US history with European arrival is that pre-Columbian Americans remain a people without history.

The historiographic fault line separating pre-Columbian and Columbian histories ignores the nearly 1 million recorded archaeological sites in the United States that span the last 14,000 y. Today’s increasingly historicized archaeology is capable of dating past events and processes with an accuracy approaching the human lifespan. When empowered with appropriate proxies, the emerging datasets offer the chance to explore long-term trends in deep American history to those willing to cross conventional disciplinary lines.

Kelly et al. (2) transcend the pre-Columbian/Columbian divide by applying radiocarbon dating at a massive scale to define a previously unrecognized Indigenous demographic shift that spanned North America and rocked the pre-Columbian world. This new synthesis both confirms the long-known demographic drop after European arrival and generates compelling evidence of earlier population flux. Kelly et al. document a significantly larger Indigenous population across the United States before 1150 CE, followed by a previously detected (but unassessed) depopulation. Pre-Columbian impacts encompassed agricultural and foraging populations and involved interactions among climate change (especially drought), endogenous disease, emigration, and warfare. Although clues about these regional demographics had previously surfaced, this large-scale study clarifies the magnitude, scale, and scope of pre-Columbian downturn.

Kelly and colleagues previously compiled a radiocarbon database with more than 100,000 archaeological, geological, and paleontological dates for the lower 48 states (6); for the PNAS study, Kelly et al. augment this dataset with thousands more dates collected since 2022. After extensive chronological hygiene, the resulting database of >61,000 ¹⁴C dates spans the last 2,000 y across the continental United States, partitioned into 18 hydrological units reflecting major ecological and cultural divisions. These investigators applied the “permtest” function in the “rcarbon” package in R (7) to calibrate the date estimates and define summed probability distributions (SPDs). Permutating the 200-y running means of the regional SPDs through 1,000 Monte Carlo simulations, they identified statistically significant demographic peaks and troughs relative to a simulated null model for the continental United States.

Fig. 1 shows how the SPD proxy works in the Great Basin (the Hydrological Unit 16 in ref. 8). In general, abrupt climatic change tends to create more significant impacts to farming populations, with hunter-gatherers more vulnerable to “gradual” climate events, and Fig. 1 clearly reflects this. After millennia of exponential growth, human population dropped significantly during the Late Holocene Dry Period, then rebounded across the Intermountain West with horticultural Fremont and Virgin Branch occupations peaking at 1089 CE. The blue trough reflects the steep demographic plunge when the drought terminating 1150 CE wiped out the corn farmers. But because foraging populations typically retained more flexibility in social organization and land use strategies than the horticulturist populations, these hunter-gatherers prospered until 19th century European arrival when the Indigenous population plunged again. Resilient Paiute and Shoshone descendants still live in the Great Basin.

Fig. 1.

Summed probability distribution (SPD) for Great Basin (Hydrological Unit 16 in Kelly et al.’s study). The 4727 ¹⁴C dates are plotted as a 200-y running mean (dark line) and compared to the 95% CI (gray band) from the null exponential model. The statistically significant positive deviations are highlighted in pink and significant negative deviations appear in blue (ref. 8, fig. 13.2). Image Credit: Reprinted with permission from ref. 8.

Like all scientific proxies, potential biases attend radiocarbon SPDs. Kelly et al. mitigated these through judicious database cleaning protocols, thinning and binning procedures, extraordinarily large sample sizes, and statistical smoothing. Still, some may be uncomfortable with the baseline assumption that large samples of radiocarbon dates provide an adequate proxy of population size. Because the number of available ¹⁴C dates varies among foraging, horticultural, and agricultural populations, SPD proxies better track energy consumption rather than absolute population size (9). This is why Kelly et al. deliberately refrain from estimating exact population sizes. However, they do caution that if taken at face value, SPD radiocarbon proxies suggest a continental-wide population loss on the order of 30% between 1150 CE and 1500 CE. Because the settlements of nomadic foragers and horticulturalists are likely overdated relative to the largest sedentary settlements (such as Chaco Canyon and Cahokia), actual population sizes could be greater during the pre-Columbian peak, rendering the 30% loss before ~1150 CE a conservative estimate.

The demographic synthesis by Kelly et al. opens the door for more precise reconstructions of population flux through additional proxies including site size, demographic structure, settlement patterns, mortuary practices, disease, food availability, abrupt and gradual climatic change, age structure of past populations, variable birth rates, and stable isotope measures of diet (10, 11). Fine-grained chronologies like Fig. 1 also help archaeologists bridge the pre-Columbian/Columbian fault line by increasing collaboration with Indigenous scholars bringing their own oral histories into the mix (12–15).

Combining oral history with tribal history and archaeology sheds new light on democratic origins in North America. Historians and political scientists have long traced the birth of American democracy to the emergence of the United States. Turner’s frontier thesis proposed a direct link between rugged individualism and the national identity of democracy and egalitarianism that spread across the world (3). Blackhawk’s “unmaking” of US history emphasizes the irony in “the world’s most exemplary democracy” being defined by a white male-dominated institution founded on the homelands of dispossessed Indigenous peoples (5).

Indigenous scholars and archaeologists explore instead that the deep roots of democratic principles extend well before European colonialism. Thompson et al. (16) argue that Ancestral Muskogean and Cherokee council house/earth-lodges reflect collective governance and democracy involving both men and women. High-precision radiocarbon dating at Cold Springs (Georgia) and elsewhere establishes that such council houses are more than 1,500 y old. Indigenous democracy survived through the fractured colonial landscape and persists today in the Muscogee (Creek) Nation (Oklahoma)—where the capitol building was deliberately constructed in the form of a council house/earth-lodge.

The de Soto and Coronado expeditions lucked out because they showed up a few centuries after America’s Indigenous experiments in political hierarchy. In their heyday, the Chacoans were among the world’s most violent societies and the Cahokia Mississippians numbered in the tens of thousands. Chaco fell after the Great Drought of 1276 to 1299 CE, when abrupt climate change ended centuries of episodic raiding, intragroup male/male violence, and group executions to coerce compliance (17–19). The multitudes of Mesa Verde and Chacoan survivors decamped to the northern Rio Grande and purposefully downsized into something like 130 smaller pueblo settlements. Ten generations later, when Coronado showed up looking for the legendary “Seven Cities of Cibola,” he found only places like Hawikuh, a dusty pueblo of 700 people. Modern Pueblo people explain that “resilience runs through our blood. We are here today because our ancestors … developed a reputation as a peaceful tribe” (20). But after more than a century of Spanish colonialism, the three dozen fully independent Puebloan tribes revisited their Chacoan roots and banded together to banish the Spanish from New Mexico. Today’s Pueblo Indian Cultural Center proudly labels the Pueblo Revolt of 1680 “the most successful uprising against colonial intruders by an assembly of Indigenous people” (20).

The seminal PNAS study by Kelly et al. (2) demonstrates how a new proxy derived from massive radiocarbon datasets can help bridge the long-standing pre-Columbian-Columbian divide.

One thousand miles away, the largest town encountered by de Soto’s expedition was Coosa (Georgia), headquarters of a tribal confederation governing 2,500 to 4,500 Mississippians stretching from Tennessee to Alabama. Coosa was a minor obstacle to de Soto, and the Coosa core was depopulated by the end of the 16th century. But had de Soto’s army arrived two centuries earlier, it would have encountered Cahokia (Illinois) and the largest pre-Columbian settlement north of Mexico (21). De Soto’s force of roughly 600 soldiers, explorers, and enslaved Africans would have had little chance against the highly organized Cahokia population of 15,000 to 20,000 people with a vast hinterland of dependent communities and farms. Though separated by 1,000 miles, Cahokia and Chaco Canyon were both pre-Columbian experiments in centralized power and hierarchical structure. What processes tied together these and other simultaneous continent-wide reshufflings?

Kelly et al.’s bird’s eye view demonstrates that something big happened in pre-Columbian America about 1150 CE. America’s pre-Columbian people without history have long suffered from Euro-American storytelling that duped the world into thinking that pre-Columbian America collapsed due to Old World diseases and superior European civilization, leaving only hapless Indigenous victims in their wake. Kelly et al. put the lie to this misconception by demonstrating that—like elsewhere in the world—pre-Columbian demography fluctuated dramatically in response to shifting headwinds and tailwinds. When abrupt climate change and social discontent knocked out the largest population centers, resilient survivors opened new windows of opportunity to downsize, restructure, and move forward. Decolonizing the legacies of these diverse histories requires that we transcend the pre-Columbian fault line with fresh historiographies appreciating deep Indigenous oral histories and drawing upon archaeology’s dates-as-data approaches to redress that balance.