Introduction: Collapse, Resilience, and the Centralized State in Historical Climatology

The specter of global warming has, over the past couple of decades, amplified climate scientists’ and historians’ interest in one another’s research. While some historians have looked anew at the long-term climate impact of historical epochs such as the agricultural and industrial revolutions while others have examined how past climate change was a shaping factor of earlier human events (Chakrabarty 2009; White 2011, 2017; Degroot 2018). At the same time, climate scientists increasingly recognize that any effective response to global warming must take into account sociological factors. Climate researchers are turning to the historical record to better understand how reconstructed paleoclimatic conditions affected people in real life, and seek to “learn lessons from history” while designing global or local responses to current anthropogenic climate change predictions (Costanza et al. 2007; Gelorini and Verschuren 2012).

Two narrative devices—“collapse” and “resilience”—have come to prominence at this intersection of climatology and history (Butzer and Endfield 2012; Walker and Cooper 2011). Recently, humanistic scholars have questioned the utility of these concepts, pointing out that they are both overly simplistic and carry an implicit normative judgment in favor the status quo (Haldon et al. 2020; Cote and Nightingale 2012: 476; Endfield 2014; Izdebski 2018). Following decades of critical scholarship “from the bottom,” historians and others are poised to ask just whose “resilience” we are speaking of, and whether the “collapse” of elite power structures proven ill-equipped to respond to climate change is truly to be mourned? Consider a classic example of climate-induced historical collapse: the Old Kingdom of ancient Egypt, c. 2180 BCE. After the Sixth Dynasty perished through a mixture of toxic elite politics and failing annual Nile floods, “[w]ealth was dispersed to new centers, with economic growth, artistic and cultural change, and a shift to a different style of social complexity. Some of the elite were deeply disturbed by the course of events…” (Butzer 2012: 3633). Historians cognizant of social inequity and wary of privileging elite narratives of history might readily ask, whose “collapse” was this, really?

Left mostly unchallenged in these critiques is another way that collapse/resilience narratives are skewed: they focus on centralized states, leaving aside the majority of past human experience (White et al. 2018; Lieberman and Gordon 2018; Scott 2009: 3). This over-focus on centralized states, which usually entails a parallel focus on sedentary farming or industrial economies, impoverishes the “lessons” that climate scientists can learn from past examples of collapse or resilience (Cosmo et al. 2018). Non-centralized political systems found across much of the premodern non-western world represent some of humanity’s most flexible forms of social organization, and have proven resilient in the face of large-scale challenges, including climate change (McIntosh 1999: 1–31; Mattalia et al. 2018). Typical hallmarks of decentralized and mobile societies are distributed risk, inter-group alliance building, and flexible boundaries—in other words, factors that can contribute to resiliency. Historical study contributes improves decision-making not by providing a “how to” manual or even a roadmap from the past, but by generating a reservoir of informative experiences beyond what any single person could accrue in a lifetime—a point perhaps best articulated by military historians (Luvaas 1982; Griess 1988: 37). By overlooking past examples of non-centralized societies in our search for historical examples, we are missing a critical piece of the human-climate story, and thereby limiting the usefulness of historical study for policy makers.

The habit of visualizing “collapse” and resilience” through the lens of the state impedes a more expansive understanding of these twinned concepts that centers ordinary people rather than their governments. To be sure, this tendency reflects biases inherent in research produced within the framework of today’s nation-state system. Yet, one of the biggest hurdles to incorporating decentralized societies into climate history is more mundane: the documentary records available to climate historians are almost exclusively produced within sedentary centralized states (Brӧnnimann et al. 2018: 30; Haldon et al. 2018: 3212; Carey 2012: 236). Recent innovative work combining archaeology and documentary evidence with paleoclimate data has pushed the study of climate history to include rural areas on the margins of states (Roberts 2018). Nevertheless, times and places with little archaeological research or lying well outside the purview of centralized states remain almost entirely absent from climate historiography. Even when climate proxies such as sedimented pollen indexing innovative agricultural activity can provide some indication of broad economic responses to climate change in undocumented regions, historians still lack an “insider perspective” that is necessary to understand the social implications of such shifts (Butzer and Endfield 2012: 3629; Rosen and Rivera-Collazo 2012).

This paper uses the methodology of historical linguistics—a largely untapped source of climate history—to explore the longue durée history of social, political, and economic responses to climate change among the politically decentralized Ateker-speaking populations in the grasslands of east Africa (for examples of similar methodological convergence, see Ehret 2011; de Luna 2016; Bostoen et al. 2015; Friedrich 1970). The earliest Ateker speech community emerged in the aftermath of widespread southward migration between 1500 and 500 BCE spurred by the desiccation of the eastern Sahara desert after the end of the African Humid Period (AHP). This culture remained linguistically contiguous for more than a millennium, during which time the Ateker people adapted to a new variable environment in part by borrowing a slew of economic and cultural practices from neighboring groups. Much later, aridity during the semi-global Medieval Climate Anomaly (MCA) c. 900 to 1250 CE spurred some Ateker-speakers to migrate further south, while others endured in their homeland by adopting new cattle herding techniques. After the return of rains c. 1250 CE, Ateker-speakers drew on these arid-period innovations to rapidly expand across a wide area of today’s Uganda-Kenya-Ethiopia-South Sudan borderlands. Throughout all this history, Ateker-speakers never established centralized political institutions. Examining this history shows how linguistics can be combined with paleoclimatic data to yield a fuller picture of human-climate interaction in understudied historical settings, while also shedding light on non-centralized responses to climate change.

500 BCE—900 CE: Resilience Through Interaction

The first significant climate event to shape Ateker history actually occurred before the Proto Ateker language was spoken (Fitzsimons 2020). Around 5,000 years ago, seasonal rainfall in the Nilotic Sudan began shifting south, inaugurating the end of the African Humid Period (AHP) and desiccation of the Sahara (Shanahan et al. 2015; Berke et al. 2012; Krӧpelin et al. 2008). Over the following two millennia, southward-retreating rains were followed by diverse Sudanese populations, many of whom eventually settled along today’s Uganda-South Sudan border, forming a culturally heterogenous zone (David 1982).Footnote 1 There, by about 500 BCE, Proto Ateker emerged as a distinct language spoken in the mountainous mixture of grasslands and acacia forest on this zone’s eastern edge. The early Ateker lived in close proximity to a range of diverse protolanguage communities, including especially those of the Southwest Surmic, Lwo, and Rub language families (Ehret 1982). Proto Ateker culture and language was first formed, therefore, in a context of migration, cultural contact, and re-settlement in response to long-term climate change. This language community would remain basically intact until c. 900 CE (Fig. 1), demonstrating a marked resilience in the face of climate change.

Fig. 1
figure 1

Linguistic tree of Ateker language family, including postulated homelands and dates

Having left behind a flat savanna with intermittent swamps and rivers, Ateker-speakers settled in an area of ecological diversity marked by hot low-lying plains tapering off into scrub land to the drier north and cooler forested mountains reaching over 6,000 feet ASL in elevation to the south. When Proto-Ateker speakers first occupied this land, they were already familiar with cultivating finger millet (Eleusine coracana) and keeping cattle, goats, and chickens. Archaeological excavations demonstrate the antiquity of livestock keeping in the region, and genetic studies point to the domestication of finger millet in East Africa by this period. We can further pinpoint Ateker participation through linguistics (Kay et al. 2019; Marshall and Hildebrand 2002; Dida et al. 2008) because reconstructed words for “finger millet” *-kima, “goat” *-kine, “cow” *-kiteng and “chicken” *-(ko)-kor, were all inherited from Proto Eastern Nilotic (the * symbol marks a reconstructed protoword; evidence underlying all linguistic reconstructions can be found in Fitzsimons 2020). From their arrival, Proto Ateker-speakers were probably the most skilled herders in their neighborhood, so they found little reason to borrow foreign words to describe sophisticated herding techniques. Proto Ateker-speakers inherited a robust herding lexicon, including items such as *-woro “cow dung,” *-dong “to castrate by pounding,” *-gelem “to castrate by cutting,” *-mong “ox,” *-dak “to graze (of livestock),” and *-kyok- “to herd livestock.” Newly innovated words, including *-manangit “very young calf,” *-kori “he-goat,” *-dongot “cow-bell,” *-lepit “milking can,” *-kere-t “gourd for churning milk,” *-doot “salt lick,” *-bela “herding stick,” *-gum- “to bleed cattle,” *-coto “cow urine,” and *-tub(w)a “watering trough” were all derived from native Eastern Nilotic sources, while only *-masanik “bull” was borrowed, probably from Southwest Surmic. Globally, livestock have always served as an important hedge against drought and crop failure, and a robust herding culture undoubtedly contributed to Ateker resilience. Throughout this period, livestock of all kinds continued to be prized by Ateker-speakers, who coined a generic term *-bar-en to mean “livestock (not species-specific),” derived from an earlier word meaning simply “wealth.”

In the realm of cereal agriculture, Proto Ateker-speakers had more to learn from their new neighbors. Though they could still grow the finger millet inherited from their Eastern Nilotic ancestors, experience with periodic drought appears to have spurred cereal crop diversification. Drought-resistant sorghum (Sorghum bicolor) had long been domesticated in the wider region and Proto Ateker speakers paid renewed attention to the crop, borrowing *-momw- from an unknown source language to name it specifically (Winchell et al. 2017). This lexical innovation accompanied a wider revision in the Proto Ateker lexicon for producing cereal-based foodstuffs. About half of these new terms had Eastern Nilotic etymologies, such as *-kut “digging stick,” *-kir(i)ya- “grinding stone,” and *-tap “bread.” Others were borrowed from an array of neighbors, including *-pyet “to winnow” from Proto Lwo, and *edula “granary” from the Moru-Madi language family. Proto Ateker speakers adapted to uneven climatic conditions by growing a more drought-resistant secondary staple and developing new techniques for processing and storing grains, both with the input of neighboring groups.

Food collection–hunting, fishing, gathering–underwent an even more far-ranging transformation during the Proto Ateker period. Ateker speakers learned from already-established communities how to exploit natural resources in their new mountainous homeland. From the Rub, Proto Ateker speakers borrowed the verb *-pok- “to ensare in a trap,” as well as the local name *-jeje for the honeyguide bird (Indicator indicator) that led followers to rich beehives. Hunting was also part of the early Ateker economy. The root *riƔ, borrowed from the nearby Proto Nuer-Dinka language with the meaning “to encircle and compress” formed the Proto-Ateker verb *-ri(g)(k)a “to hunt in a group,” as well as the noun for the *-ri(g)(k)ak “hunting party” that used this tactic. Finally, Proto Ateker-speakers were ecumenical in borrowing fishing technologies, adopting *-kol- “fish” from Rub, *-biti “small fishing spear” from Lwo, and *-golo “fish-hook” from Southwest Surmic.

New subsistence practices supporting life in a variable environment were buttressed by new technology. Most significantly, the settlement of Proto Ateker speakers in their new homeland marked the start of their participation in East Africa’s Pastoral Iron Age (Ehret 1982). They did not smelt iron themselves, but instead acquired iron implements from their neighbors. The iron spear haft (*-morok), iron hoe (*-melek), and iron fish-hook (*-golo), were all borrowed from Southwest Surmic, while the iron axe (*-jep) used to clear forests was borrowed from Lwo. This period also saw improvements in transportation. The donkey, called *-sigiria, was borrowed from Southwest Surmic, and was affixed with a pack-saddle, *saaja-t, to stabilize loads. Other innovations included a gourd bottle, *-tuo, for carrying water on journeys, and new word for a head-carrying pad, *ikit, innovated from a Rub borrowing.

The key to successful farming and pasturing two thousand years ago in East Africa was rainfall, and Proto Ateker-speakers would have paid much attention to precipitation. Here, greater elevation changes in the Proto Ateker community’s new home played a crucial role (Jackson 1956). In a region where temperatures in low-lying plains routinely exceed 100 degrees Fahrenheit, moist air coming from the Indian Ocean became super-heated and rose rapidly after encountering physical barriers such as the region’s plateaus and mountain complexes. As moist rising air began to cool after reaching elevations between 15,000 and 20,000 feet ASL, the tumultuous blend of falling icy precipitation and rising hot air created static charges that frequently caused powerful thunderstorms and heavy precipitation around mountains. It was, perhaps, after witnessing this entirely new weather pattern that the Proto Ateker language underwent a semantic shift in which the Proto Eastern Nilotic root *-kudyu “rain” (pronounced *-kuju in Proto Ateker) came to denote the “sky” as a conceptually distinct entity, while a new word for the noun “rain” was formed by attaching a verb prefix /aki-/ to the inherited root *-ru “to water plants/animals.” Whereas the Proto Eastern Nilotic language had described rain as a noun without agency—something that just happened—Proto Ateker-speakers spoke of rain, *akiru, as a transitive verb performed by a distinct actor, *-kuju “the sky,” with the implied intention of watering plants and animals. Although a fully developed theory of akuj as the “High God” most likely did not enter Ateker philosophy until a later period, the Proto Ateker language community had already trained their thoughts upwards. To name phenomena such as *-top “Venus,” *-kacer “star,” and *-gir “thunder,” they borrowed words from their Rub neighbors. Terms for unfamiliar high-elevation water sources, such as *-bur “mountainous water pool,” and *ecoa “rocky ground spring,” were also innovated in Proto Ateker, although these words’ origins are unknown.

Over more than a millennium, Proto Ateker-speakers slowly refined their economic and social practices to achieve their own ends. The primary Proto Ateker sociopolitical group was the exogamous clan, called *ateker, of which there were probably no more than two dozen. Livestock were preeminent in Proto Ateker culture, and cattle belonging to each clan were marked with special brands, called *-macar. Hospitality revolved around meat-feasting, with the word *-pej “visitor” deriving from an earlier root meaning “to roast meat.” Animal sacrifice, in the meantime, was seen as essential for both mourning rituals, called *-puny, and for harnessing spiritual forces to particular ends through a rite called *-sub-an. These and countless other Proto Ateker practices and values were well-adapted to their local environment, while economic diversification and trade created what was in all likelihood a prosperous and growing community.

900 to 1400—“collapse” and Transformation

Rainfall in the Proto Ateker homeland between the end of the AHP and c. 900 CE, though marked by decadal variations and intermittent drought, was overall sufficient to support the Proto Ateker way of life. Subsistence diversification and intercultural exchange generated enough flexibility in the local socio-ecological system to enable the resilience and longevity of the Proto Ateker language community. However, during a three-hundred-and-fifty-year period beginning c. 900 CE and corresponding with the semi-global MCA, excessively dry conditions caused a permanent rupture to this equilibrium (Lüning et al. 2018; Lüning et al. 2017). Finger millet cultivation became nearly impossible and sorghum cultivation challenging, while water sources for livestock were fewer and farther away.

Water scarcity creates a challenge for the reconstruction of local climate histories, because the best tropical palaeoclimatological sources are found at the bottom of standing lakes, which are not common in the Proto Ateker homeland (Gasse 2000). Bodies of standing water contain piled lake sediments that hold a sequenced history of climatic detritus which can be analyzed to determine the nature of past environments. Types of pollen or the photosynthetic material embedded in stromatoliths found in datable sedimentary layers can index varying climatic conditions at the time of their stratigraphic formation, as can the chemical compositions of lake sediment soils. Throughout the Northern Ateker world, there is only one standing water body—Lake Turkana—that paleoclimatologists have closely studied, and it stands 150 miles away from the Proto Ateker homeland (Morrissey and Scholz 2014; Garcin et al. 2012). This lake can be considered a proxy for the regional environment (Bloszies and Forman 2015). Halfmann et al. measured ratios of fine-grained carbonite isotopes in Lake Turkana sediments as a proxy for water levels, with the assumption that lower ratios would indicate higher overall water volumes. Using this evidence, they conclude that Lake Turkana experienced overall low water levels in the period c. 900–1100 CE (Halfman et al. 1994: 94). Looking throughout the wider region affected by the same monsoon-derived seasonal rains, Verschuren et al. combine evidence from Lake Naivasha in Kenya and Lake Tanganyika in Tanzania with Halfman’s study from Lake Turkana to suggest a regional period of aridity on either side of c. 1000 CE that stretched from Tanzania to Northern Kenya (Verschuren et al. 2000). A final unique source for climatic conditions is the “Nilometer” on Rhoda island in Cairo, which has been used to document annual Nile flood levels since the seventh century CE. Because the lower Nile is partially fed by rainfall in northern Uganda, Nile flood levels can be considered an imprecise proxy for climate conditions in the Ateker region. The Rhoda Nilometer matches paleoclimatological studies, indicating periodic severe droughts amid overall low water levels from CE 900–1250 CE (Hassan 2007; Kondrashov et al. 2005; Putter et al. 1998).

A convergence of evidence indicates regional drought in eastern Africa during the MCA at the end of the first millennium CE. This period corresponds with language-based estimates for the divergence of Proto Ateker into Proto Northern Ateker (PNA) and Proto Teso. The story that emerges is one where Proto Ateker speakers faced an existential question at the onset of this arid period: should they migrate south towards wetter climes to maintain an economy based on finger millet and localized livestock-rearing, or undergo a fundamental transformation towards subsistence practices better suited to reduced rainfall? In the end both choices were made, leading to the bifurcation of the Proto Ateker language into two distinct sub-groups. Those who remained behind formed the PNA. They adapted to dry conditions through the use of herding technique called transhumant pastoralism in which households split during the annual dry season so that young adults can take herds in search of far-out grazing lands while elders watch over permanent homes and young children. In contrast, the Proto Teso, remembered in oral traditions as the Ngikatapa or “finger millet bread people,” moved south and ultimately settled in wetter eastern Uganda (Lamphear 1976). There, they established finger millet farms and joined other migrating families in loose neighborhood associations. Both groups embarked upon paths that fundamentally transformed their social and political lifeways.

Lexical reconstructions of selected flora help track the local impact of the MCA climate shift on the PNA. The most economically significant of these innovations was *erau, for “pearl millet,” borrowed from Lwo speakers. Pearl millet (Pennisetum glaucum) is a global cereal crop that was first domesticated in West Africa by the third millennium BCE, though it was not significant in Proto Ateker agriculture before 900 CE (Manning et al. 2011). Pearl millet has the advantage of being farmable in areas receiving an average of 400–650 mm per annum of rainfall, as opposed to finger millet’s minimum requirement of 600 mm (Dida and Devos 2006). Pearl millet supplemented the cultivation of sorghum, which requires 300–700 mm of rain per annum (Singh and Lohithawa 2006).

Increased aridity is also captured in the PNA arboreal lexicon. When a local climate dries, thirstier tree and shrub species die out except in areas near consistent water sources such as rivers, and they give way to new “pioneer” species with lesser rainfall requirements. We can see evidence of this process playing out in the linguistic record. All newly innovated PNA words for flora refer to species with minimum annual requirements of 200–450 mm and some with upper limits of 800 mm.Footnote 2 These include Commiphora africana or *ekadeli (250–800 mm), Hyphaene compressa or *-kVngol (300–900 mm), Salvadora persica or *esiokon (300–800 mm), and the fodder grass Bothriochloa insculpta or *elet (450–1500 mm).

Words describing the dispersed cattle herding techniques that enabled PNA-speakers to adapt to arid conditions can be dated to this same period. These include the annual assembly inaugurating cattle movement, called *-ud-akin and derived from Proto Ateker *-ud “to prod, to push together,” and the act of blessing to “release” cattle, called *wos from a root meaning “to let go.” PNA-speakers also innovated the term *-bor for dry season camps from the root “to depart, to break away,” and directionally specific verbs *-ram “to drive cattle toward” and *-twar “to drive cattle away” in order to describe the annual to-and-fro. Finally, they adopted new objects to assist with long-distance herding, including *-ku-wos(i) “gourd for carrying cow urine” (used for cleaning), *ku-tam “leather sack for carrying butter,” and *-gec “stirring stick (used for mixing milk and blood).”

Greater spatial separation challenged inherited political ideals emphasizing “gathering together” while also causing anxiety among elders, who gave up control over life-giving cattle herds for half of each year (Lamphear 1983: 114). The most significant political innovation during this period was an age-class system, called *asapan, which effectively addressed both concerns (Fitzsimons 2020: 157–210). With clans and lineages thinly spread across an arid landscape, the formation of overarching social classes based on age provided a substitute ideology of group-making that could bind together fragmented settlements. The particular structure of *asapan, which granted political power to the generation of the “fathers” over that of the “sons,” helped quell elders’ anxieties about the growing economic responsibility of the youth while also providing young men with a predictable pathway to authority in due time. Women, who had traditionally exercised power through clan structures, were initially disenfranchised by this new system, but later adopted parallel initiation institutions taking the form of “singing groups” (Gourlay 1970; Fitzsimons 2020: 240).

Turning south, the Proto Teso who migrated south to continue cultivating finger millet met climate change with an entirely different set of responses, also altering the older patriclan system of the Proto Ateker. As families and friends migrated together in small pioneer groups, the approximately two dozen over-arching Ateker patriclans lost their practical coherence (Webster et al. 1973). These migrants continuously invested in the invention and re-invention, naming and re-naming, of new smaller and more localized “sub-clans,” leading to the proliferation of well over one thousand Teso sub-clans by the eighteenth century. When they reached new lands, migrants encountered new peoples, many of whom spoke versions of Lwo and Rub languages. The general pattern of settlement, then, was one of small groups of families and friends living in diverse neighborhoods comprised of both numerous Teso-speaking sub-clans and entirely foreign cultures. The Proto Teso language community eventually came to fully occupy the area of eastern Uganda between the lakes Bisina and Kyoga, the heart of today’s Teso sub-region.

To coordinate territorial activities between unrelated small-scale settlements, Proto Teso-speakers innovated a political assembly called *etem, derived from the Proto Ateker word meaning “hearth.” These assemblies proved effective for integrating non-Teso communities, including especially members of the Lwo-speaking Inomu and Ikomolo clans, which became thoroughly “Teso-ized” during this period (Cohen 1972: 144–145; Cohen 1988: 67–68). As they integrated these Lwo communities, Teso-speakers also learned about their new landscape from them, borrowing words for two important fodder grasses dominant in the region—Guinea Grass (Panicum maximum), called *edinyo, and Elephant Grass (Pennisetum purpureum), called *egada.

Within the etem institution, Proto Teso-speakers took care to maintain a balance of power between local exogamous sub-clans, preventing any single sub-clan from becoming too dominant. Assembly speakers, called *airabis from the root word *rab “to speak,” were elected according to meritocratic principles, and the position of speaker was not inherited from father to son. Structural equality between sub-clans was also pursued through strict protocols related to marriage exchanges. To prevent one sub-clan from becoming indebted to another through marriage, bridewealth exchanges were converted into a years-long process called *-yit, from the Proto Ateker root meaning “to drip slowly.” Marriage transactions involved a series of reciprocal gift-giving calculated so that families united through marriage would end up with a clean balance sheet between them (Nagashima 1981).

The practice of migration within a sedentary farming economy also spurred new social norms at the individual level, with an increased expectation that young families should build their own separate homes. A new word for “marriage,” *-many, was derived from an earlier verb meaning simply “to dwell together,” while a married man was called *-duk-okina, literally “one who has built a home.” That new homesteads should have been separate from one’s extended family is implicit in an innovated word for “sibling,” *inac, derived from the root “to avoid, to pass by.” Notions of prosperity also followed greater investment in cereal agriculture. Whereas the Proto Ateker had only conceptualized wealth in terms of livestock, cereal-growing Proto Teso-speakers innovated a corollary term, *amion, meaning “wealth in crops” specifically (Stephens 2018).

By the end of the thirteenth century, rainfall in East Africa returned to pre-MWP levels. Although the aridity that had stimulated transformations among Ateker-speakers disappeared, the fundamental social changes it had wrought were permanent. Political institutions such as asapan age-classes and etem assemblies, initially created to ensure the survival of Ateker communities during hard times, provided a foundation for the rapid expansion of Ateker-speakers across a large swath of the region once rains returned. The flexible and decentralized strategies that enabled resilience during a climate crisis also proved effective for constituting political communities in times of relative plenty. By 1800, linguistic descendants of the PNA and Proto Teso peoples dominated more than 100,000 square miles of grasslands in Uganda, South Sudan, Kenya, and Ethiopia, while the asapan system was adopted by neighboring groups throughout the region.


Reconstructing the longue durée socio-ecological history of a decentralized society such as the Ateker offers new insights into the themes of collapse and resilience which have long been at the heart of historical climate research. To begin with, this focus offers a new set of metrics by which such concepts can be measured. While state-centric historical climatology tends to be framed by the start- and end-dates of particular dynasties or government systems, these neat dividing lines are unavailable for undocumented and decentralized societies, and they can be misleading as climate trends do not adhere to political histories. Protolanguage longevity can serve as a useful alternative proxy, because lack of linguistic divergences may indicate an absence of major disruptions to other long-term social factors. In a climatic regime characterized by decadal variability, such as that experienced by the early Ateker, linguistic longevity implies a certain degree of socio-ecological resilience. In the Ateker case, resilience was produced by a flexible subsistence system comprised of internal innovations and foreign borrowings.

The bifurcation of Proto Ateker as a result of climate-induced migrations beginning c. 900 CE provides a useful example of the complex relationships scholars have recently noted between collapse, resilience, and transformation. Traditional narratives pose collapse and resilience as opposing outcomes, but analysis of stateless histories reveal this as too simple of a dichotomy. There is no obvious standard to judge whether the more “resilient” Ateker speakers were those who migrated south to maintain their basic subsistence practices, or those who adopted transhumant pastoralism to remain in their homeland. At the level of the society, if not the individual, both strategies seem to have been successful, since descendant communities of both groups increased in population and territory through the rest of the millennium. But they did so by either undertaking a wholesale transformation of their economic and political systems, or leaving their homeland entirely. The ultimate value of bringing examples from decentralized societies into the field climate history may be that, by removing the deceptively straight-forward phenomenon of the state from consideration, these very categories themselves are shown to be insufficient.

So what are the “lessons,” if any, that climate researchers might take away from this Ateker case study? Two come to mind. The first is the value of building bridges to foreign communities–welcoming new ideas and new people. The unusual longevity of Proto Ateker language and culture in a marginal environment surely would not have been possible without borrowing subsistence practices and material technologies from a diverse range of neighbors. Beyond these practical borrowings, however, Proto Ateker speakers also adopted foreign ways to speak about topics ranging from the cosmos and ritual practice to trade and the geographical features of their homeland. After c. 900 CE the *asapan age-class system enabled the Proto Northern Ateker to forge cohesive political communities from multiple far-flung lineages, while the *etem assemblies did the same for Proto Teso-speakers.

The second is that a community’s survival during periods of climate change may require fundamental transformations to basic lifeways and shared ideals. Ateker speakers endured the arid period of 900–1250 CE—and thrived thereafter—because they were willing to take bold collective steps in response to environmental pressure. For the Proto Teso, this meant leaving the homeland they had known for over a millennium and finding their way among new people living in a new place. For the PNA, this meant giving up a lifestyle of sedentary farming and revising family practices to cope with months of geographical separation.

For a world facing an anthropogenic climate change crisis today, these “lessons” are not directly applicable, of course. No one would suggest we should all become transhumant pastoralists. However, the general themes of openness to foreign people and their ideas, and a willingness to undergo fundamental economic and socio-political transformation do seem relevant today. In the end, the greatest value that we can glean from examples of historical responses to climate change is to build a richer repertoire of narratives about our shared transhistorical human condition of living in societies that are inextricable from their natural environment. Focusing only on examples from centralized states impoverishes that repertoire, and threatens to stifle the collective creative imagination that must be at the center of any effective response to global warming.