Before the State

The Long Dawn

The Institute's register of states opens with Eridu, a city-state of southern Mesopotamia dated from 5400 BCE. By then people had been living in villages for thousands of years. At the end of the last Ice Age the Natufians of the Levant, hunter-gatherers, lived in settlements larger and more permanent than any before them, some with round stone houses, as at Ain Mallaha, and with the first known food stores. In Hilazon Tachtit cave a woman who seems to have been important was laid in a ceremonial pit with the remains of at least three aurochs and 86 tortoises, thought to have been brought for a funeral feast. Obsidian from Anatolia and shellfish from the Nile valley reached Ain Mallaha. The Natufian ended about 9,500 BCE, and its communities may be the ancestors of the builders of the region's first Neolithic settlements.

At the other end of Asia the people of the Jōmon culture made the world's oldest known pottery, about 14,500 BCE. For some 14,000 years they lived mainly by hunting, gathering and fishing, yet many settled in villages of pit-houses, stored nuts and acorns for the winter in underground pits, left great shell middens along the coasts and tended groves of chestnut and lacquer trees, a way of life somewhere between hunting-gathering and agriculture. The Middle Jōmon produced large burial mounds and monuments. In Hokkaido two Jōmon populations later merged to form the proto-Ainu.

Farming villages brought their own forms of order. At Mehrgarh, on the Kacchi Plain of Balochistan, a small farming village was founded in about 7000 BCE; its people grew wheat and barley, kept sheep, goats and cattle and stored their grain in granaries, and eleven molar crowns from nine adults had been drilled during life. In Mesopotamia the Ubaid period left the earliest known settlements on the alluvial plain: large unwalled villages of mud-brick houses and the first temples in Mesopotamia. Villages held specialised potters, weavers and metalworkers. Graves suggest a society becoming steadily less equal. Yet there were no real cities until the Uruk period, warfare seems to have been extremely rare, and Ubaid pottery and customs spread from south-eastern Turkey to the Persian Gulf, in Gil Stein and Rana Özbal's account, largely through the peaceful spread of ideas rather than conquest. Eridu itself is an Ubaid site, where the pottery was first found: the register's first state begins in the middle of the period that built Mesopotamia's first temples.

China shows the same sequence. Along the middle Yellow River the Yangshao culture, c. 5000–3000 BCE, farmed millet and lived in villages with semi-subterranean houses set round a central square. Communal storage pits and large public buildings suggest that some production and distribution were managed together. Only in the late Yangshao did settlement hierarchies spread widely, the first rammed-earth wall in China rise at Xishan, and richly furnished tombs at Xipo suggest an early form of chiefdom. In the Yangtze Delta the Liangzhu culture, c. 3300–2300 BCE, went further: its city near Hangzhou enclosed 290 hectares within clay walls. Jade, silk, ivory and lacquer have been found only in elite burials while poorer graves held mostly pottery, and this division is the main reason the culture is described as an early state. Liangzhu stands at the threshold that the other cultures of this section never crossed.

None of these peoples left a written word. What they governed, they governed without record: the storing of food against winter, the laying out of a village round its square, the drilling of a living person's tooth, the feast at a grave. Settled life, storage, craft specialisation, exchange over hundreds of kilometres, monumental building and ranked burial all appear in this record before any state does, in the Levant and Japan by thousands of years.

Order Without a King

How peoples without a state ordered themselves is known partly from outsiders who wrote about them and partly from their own traditions, and the oldest texts record a sequence that ran from council to king rather than the reverse. Early Vedic society in the Punjab, c. 1500–1000 BCE, was organised into tribes rather than kingdoms and lived mainly by herding; clans as a whole held rights over land and herds. The tribal chief, the rajan, stood at the top with the nobility, but tribal councils, the sabha and samiti, limited him. With the later kingdoms, among them the Kuru, hereditary kingship emerged and the councils declined.

In Iron Age Europe the Greeks and Romans described societies of nobles, their clients and their war bands. In Caesar's day the larger Gallic peoples, such as the Aedui, were governed by councils of nobles and elected magistrates; Caesar names the druids and the knights as the two honoured classes, with the common people little better than slaves. Women of high rank could hold power among the Celts: Cartimandua ruled the Brigantes of northern Britain. The Thracians, some 200 tribes of south-eastern Europe who shared much of their culture, were rarely united in one state. When states did form among them they were unions: the Odrysian Kingdom joined more than 40 Thracian tribes and 22 kingdoms.

On the steppe north of the Black Sea, Herodotus describes the Scythians as a people without towns or walls who carried their homes with them, all of them mounted archers who lived from their herds rather than the plough. At the time of Darius's invasion Herodotus names three kings, each with his own part of the country, and below them stood district governors and a warrior nobility.

In the river valleys of eastern North America the Hopewell communities of about 100 BCE to 500 CE were small, temporary settlements of one to three households, yet they built geometric earthworks covering tens to hundreds of acres, with only sparse settlement around them. Their leaders did not have the centralised power to command armies; some scholars see the rise of “big-men”, leaders whose influence rested on their skill at persuasion in trade and religion and on ties of mutual obligation. Those who came after built differently. The societies of the Mississippian culture, from about 800 to 1600, were chiefdoms: a chief from a ranked lineage ruled a town and the farmsteads around it, and the strongest, the paramount chiefs, drew tribute and warriors from several towns. Such powers rarely lasted long, and Cahokia was largely abandoned by about 1350.

The living peoples of the Pacific and the Americas kept forms of order that outsiders recorded in detail. The Māori of the Classic period lived in autonomous settlements of several whānau, extended families, forming a hapū, and hapū with shared ancestry formed iwi, the largest social unit; society was divided into rangatira, the chiefs and ruling families, tūtūā, commoners, and mōkai, slaves. Systems such as tapu and rāhui used religious sanctions to stop people from taking certain species at particular seasons or from certain places, and genealogies, whakapapa, were memorised and passed down. The Mapuche of south-central Chile were organised around lineages of local communities called lov, joined into the larger rehue; each extended family was led by a lonko, or chief, and in time of war the Mapuche united in larger groupings and elected a toqui, “axe-bearer”, to lead them, while the machi, usually a woman, healed and performed the ceremonies. The Taíno of the Greater Antilles lived in villages under a cacique and grouped their chiefdoms, the cacicazgos, into confederations. A cacique's power rested on the villages he controlled and on alliances of family, marriage and ceremony, and the most elaborate courts for the ball game, batey, lie on the boundaries between chiefdoms, where games are believed to have settled disputes between communities.

In the north the scale was smaller still. Ainu society was built around villages called kotan, usually of fewer than ten households, set along rivers where salmon ran; several kotan sometimes formed a local group led by a headman who handled its outside and inside affairs. Wrongdoers were judged by members of the community, there was no imprisonment or death penalty, and a beating was usually the final punishment. The Sámi lived and worked in siidat, groups of several families who shared the management of their land and, later, their herds. Across these cases the frame of order is the same: kinship, from the whānau to the lov and the kotan; a leader raised for a purpose, the toqui for war or the headman for dealings with outsiders; religious sanction in place of written law; and genealogy, memorised, in place of archives.

Cultures That Never Became States

Some of the widest and longest-lived cultures in the Institute's collection never formed a single state at all. The Lapita culture, c. 1500–500 BCE, carried the first settlers from the Bismarck Archipelago to Vanuatu, New Caledonia, Fiji, Tonga and Samoa. Its communities were hamlets or small villages on beaches and sand spits, and little is known directly of how they were governed: Patrick Kirch has argued, from comparison with later Oceanic societies, that they were organised in “house societies”, groups of kin attached to a founding house and its ancestors. What held the scattered communities together was exchange and a shared style. Obsidian from Talasea on New Britain reached Vanuatu, New Caledonia and Fiji, and even Bukit Tengkorak in Sabah, on Borneo. The motifs of Lapita pottery were repeated with striking consistency from the Bismarcks to Samoa. Lapita did not end so much as turn into its successors: some 450 Oceanic languages descend from the language its people most probably spoke.

In Europe the Bell Beaker culture, c. 2800–1800 BCE, is an archaeologists' name for a set of objects and customs led by a drinking vessel shaped like an upside-down bell. Its spread along long-used sea and river routes seems to have been linked directly to the search for copper and other rare raw materials: at Ross Island in Ireland, the earliest known Irish copper mines, from about 2500 BCE, local ores were smelted into the first copper axes of Britain and Ireland. In its mature phase the culture combined metalworking, long-distance exchange and archery with growing social ranks and regional elites, but no Beaker state is known. The trade routes the Beaker people opened stayed firmly in use.

The Yamnaya culture of the Pontic–Caspian steppe, c. 3300–2600 BCE, is likewise an archaeologists' label. Its people were nomads or semi-nomads who moved their herds in two-wheeled carts and four-wheeled wagons drawn by oxen, and society was organised in chiefdoms. David Anthony saw the culture as the visible expression of a social adjustment to high mobility, the political arrangements needed to manage larger herds from mobile homes on the steppe. Without a state, the Yamnaya left a larger mark than most states have: Yamnaya-related ancestry makes up about 38.8–50.4% of the ancestry of modern Central and Northern Europeans, and under the widely accepted Kurgan hypothesis of Marija Gimbutas they spoke a stage of Proto-Indo-European.

The Scythians had kings, but there was never one Scythian state. The earliest known burial with the “Scythian triad” of weapons, horse harness and art in the animal style lies far to the east of the Black Sea, at Arzhan in Tuva, a great mound of about 800 BCE, and a wider Scythian world of related cultures reached across Kazakhstan to the Altai and Mongolia. What travelled was a way of life and a style. When Darius I crossed the Danube around 513 BCE to attack the Scythians, they refused battle, filled in the wells and destroyed the grass, and he withdrew; the campaign showed the limits of the Achaemenid Empire's power against mounted nomads.

The Sogdians held together through the rise and fall of every empire around them. Sogdiana, between the Amu Darya and the Syr Darya, was seldom united under one ruler: it was a land of oasis city-states of which Samarkand was the largest. What made the Sogdians one people was their Eastern Iranian language and above all trade. From the 4th century CE at the latest their merchants were the chief carriers of trade between Central Asia and China; the Ancient Letters deal with business and debts, and their language served as a common tongue of the Silk Road. In 568 the Sogdian Maniakh led a Turkic embassy to the emperor Justin II in Constantinople. By the 11th century the Sogdians had become Muslim and merged into the Persian- and Turkic-speaking peoples of Central Asia. Their alphabet outlived them: adopted by the Uyghurs in the 8th century, it became the traditional Mongolian script and is still written in Inner Mongolia today.

What held these cultures together was not a government. It was exchange, which carried Talasea obsidian to Borneo, Irish copper to the Netherlands and a Sogdian embassy to Constantinople; kinship, in Lapita's house societies and the Yamnaya's chiefdoms; and a shared craft and style, the dentate stamp, the bell-shaped beaker, the animal style and the painted hall, repeated across thousands of kilometres with a consistency that no king enforced.

Peoples Who Outlived States

Five of the Institute's forty-five cultures are living peoples: the Ainu, the Sámi, the Māori, the Mapuche and the Taíno. Each lived for centuries beside states and then under them, and each record is explicit about what those states did.

The Ainu of Hokkaido, Sakhalin and the Kurils traded between Honshu and Northeast Asia until the Tokugawa shogunate gave the Matsumae clan exclusive rights to trade with them. Under direct shogunate control in the early 19th century Ainu women were separated from their husbands and raped or forcibly married to Japanese men, and Ainu men were sent to merchants for five- and ten-year terms of labour; with smallpox, these policies made the population fall steeply. After the Meiji Restoration and the founding of the Hokkaidō Development Commission in 1869, Ainu land was given to Japanese settlers. The Hokkaido Former Aborigines Protection Act of 1899 took Ainu land and was designed to erase Ainu culture and identity: the rights to hunt, gather and speak their language were denied, and children were sent to schools where Ainu was forbidden. Hokkaido Ainu now probably has no native speakers. Revival came through Ainu effort: Shigeru Kayano opened an Ainu language school in 1987 and fought the Nibutani Dam. The Sapporo District Court's ruling of 1997 recognised the Ainu as an Indigenous people and the 1899 act was repealed; Japan passed recognition into law in 2019, and in 2023 Hokkaido Ainu sued to regain the right to fish salmon in rivers.

The Sámi were taxed before they were ruled. From the 15th century Denmark–Norway, Sweden and Russia all claimed the right to tax them, so that Mountain Sámi following their herds might pay up to three states. Sámi faced witch trials and burnings in the 17th century, and in Norway the missionary Thomas von Westen led a campaign around 1720 that burned drums and destroyed sacred sites: of an estimated thousands of drums, about 70 survive. The border treaty of 1826 stopped herders who had wintered in Finland and summered on the Norwegian coast from crossing. From the 1860s Norway turned to Norwegianisation: from around 1900 Sámi could not be used in public school or in church, and anyone buying or leasing state land in Finnmark had to prove knowledge of Norwegian. The Alta hydro-electric controversy of 1979 made Sámi rights a national issue. Sámi Parliaments were founded in Finland in 1973, Norway in 1989 and Sweden in 1993, while Russia recognises none; Norway's Finnmark Act of 2005 gave the Sámi Parliament a share in managing most of Finnmark's land.

The Māori met the British Crown by treaty. On 6 February 1840 the Treaty of Waitangi was signed at Waitangi; it gave Māori the rights of British subjects and guaranteed their property and tribal authority in return for British sovereignty, but the Māori version the chiefs signed did not cede sovereignty to the Crown, and its meaning is disputed to this day. Disputed land sales and the rise of the Kīngitanga, the Māori King Movement founded in the 1850s to stop the loss of land, led to the New Zealand Wars of the 1860s. The colonial government invaded the Waikato in 1863 and confiscated land as punishment for what it called rebellion; the Native Land Court turned communal land into individual titles that could be sold, and between 1840 and 1890 Māori sold 95 per cent of their land. In 1975 the Crown set up the Waitangi Tribunal to investigate breaches of the Treaty, and since the 1990s governments have negotiated settlements with many iwi, though some Māori call them far below the value of what was taken; te reo Māori is an official language, and at the 2023 census 887,493 people identified as Māori, 17.8 per cent of the population.

The Mapuche held a state off for nearly 350 years. From 1550 the Spanish built cities and forts in their lands and so began the Arauco War. In 1553 Lautaro defeated and killed Governor Pedro de Valdivia at Tucapel, and after the Battle of Curalaba in 1598 the uprising that followed destroyed or emptied the Spanish cities south of the Biobío River, which, except for Chiloé, stayed free of Spanish rule. Over time the frontier became a place of trade and treaties, the parliaments such as Quillín in 1641. In the 19th century Chile and Argentina conquered them: Chile's Occupation of Araucanía ran from 1861 to 1883, and Argentina waged similar campaigns east of the Andes in the 1870s. Thousands of Mapuche died in combat, from pillage, starvation and smallpox, the traditional economy collapsed, and laws dividing community lands, in 1931 and in 1979, broke up their holdings further. The Indigenous Law of 1993 recognised the Mapuche people, Mapudungun and their culture, and disputes over land and forestry plantations in Araucanía have led to violent confrontations and the prosecution of activists under anti-terrorism laws.

The Taíno were the first people of the Americas met by Columbus, and Spanish documents declared them extinct as early as 1550. From 1493 Columbus demanded tribute in gold or spun cotton from every adult on Hispaniola. Under the encomienda system the Spanish forced the Taíno to work in fields and mines and seized their land. Hatuey, who had fled from Hispaniola to Cuba to unite its peoples, was burned at the stake on 2 February 1512, while on Hispaniola Enriquillo led more than 3,000 Taíno in a successful rebellion in the 1520s and won land and a charter from the crown. Within 30 years between 80 and 90 per cent of the Taíno died. Many historians blame epidemics of smallpox, influenza, measles and typhus above all; others, such as Andrés Reséndez, argue that enslavement and forced labour in the mines were a major killer, and David Stannard calls the encomienda a genocidal system. Yet census records of 1514 show that 40 per cent of Spanish men on Hispaniola had Taíno wives, and in eastern Cuba observers from the 19th century onwards recorded Indigenous communities, some of which have kept a Taíno identity into the 21st century. DNA studies show that 61 per cent of Puerto Ricans carry mitochondrial DNA of Indigenous origin, and in the 2010 United States census 9,399 people in Puerto Rico identified as Taíno. Scholars still disagree about how much survived; descendants answer that mixing is how their culture lasted.

What survived in each case was the frame described above rather than any institution of rule. The Ainu kept the Yukar and the kamuy, and sue today for the salmon of their rivers; the Sámi kept their languages, the joik and the gákti, and have won a share in managing the land of Finnmark; the Māori kept the marae, the whānau and whakapapa, which remain central to Māori life; the Mapuche are reviving the machi's knowledge of medicinal herbs, and their women still weave; the Taíno kept their descent and their words, hammock, canoe, tobacco and hurricane among them, and have taken back their name. The states that set out to erase these peoples have in several cases been made to recognise them by their own courts and parliaments: a court in Sapporo in 1997, the Diet in 2019, Chile's Indigenous Law in 1993, Sámi Parliaments in three of the four states that divide Sápmi, and a tribunal in New Zealand that has sat since 1975 to judge the Crown's breaches of a treaty it signed in 1840.

The Lesson for Institutions

The record of these forty-five cultures reverses a common assumption. Most of what people need from one another, food stored against the winter, a village laid out round its square, disputes judged, the dead honoured, knowledge handed down, was supplied for thousands of years by kin groups, councils and ritual before any state existed, and among the peoples of the last section it went on being supplied by them under the state and against it. The Natufians stored food and the Ubaid villages built temples before Eridu; the Ainu judged wrongdoers in the community without prison or death penalty; the Māori restrained the taking of fish and birds by tapu and rāhui; the Taíno settled disputes between chiefdoms on the ball court. A state that arrived among such a people did not fill an empty space. It displaced an order that was already there.

The states of the nineteenth and twentieth centuries understood this well enough to aim at the frame itself. The Native Land Court turned Māori communal land into individual titles that could be sold; Chile's laws of 1931 and 1979 divided Mapuche community lands; the Hokkaido act of 1899 denied the Ainu the right to hunt, gather and speak their language; Norway required knowledge of Norwegian of anyone buying state land in Finnmark; the border treaty of 1826 stopped the Sámi herds from crossing. Each measure struck at kin-held land, at language, or at the movement on which a way of life depended, and each is what the descendants have since worked to undo, from the Finnmark Act of 2005 to the Ainu suit of 2023 for the salmon. For anyone responsible for an institution, the lesson is that the things a state can abolish by statute are not the things that hold a people together, and that a people whose frame survives will in time use the state's own instruments, its courts, its censuses and its treaties, to claim what was taken.

The record also shows what a state can do that a culture cannot. The Lapp Codicil of 1751, an annex to the Strömstad Treaty between Sweden and Norway, regulated how Sámi used the land across a border they had not drawn; the Parliament of Quillín in 1641 and the Treaty of Waitangi in 1840 were agreements between a state and peoples who were not one; and each later became ground on which rights were claimed, as the Waitangi Tribunal judges breaches of the treaty of 1840. A culture holds people together across space and time; a state can bind itself in writing, and be held to it. The peoples in this essay who lasted longest did so by keeping their own frame and, where they could, by making the states around them sign.

What the Data Shows

The Institute profiles forty-five cultures: ten of the Americas, ten of Central and East Asia, nine of Europe, eight of South and Southeast Asia, three of the Ancient Near East, two of Oceania, two of Sub-Saharan Africa and one of Egypt and North Africa. Five are living peoples. Their spans are conventions: most begin with a “c.” and several are archaeologists' brackets around a style of pottery or burial rather than the life of a people. The register's states are dated by the same kind of convention, and 5,280 of its 5,468 states carry both a start and an end. A culture's span and a state's lifespan also measure different things: a state ends when its government does, while a culture's dates close when its pottery, its burials or its language are no longer found. With those cautions, the comparison is still telling.

The median culture in the collection spans 1,000 years, and the mean 1,656; the longest, the Jōmon, spans 13,700 years and the shortest, the Igbo-Ukwu culture, 200. Among the register's dated states the median lifespan is 191.5 years and the mean 255. A third of recorded states, 1,746 of them, or 33 per cent, lasted under a century; 793 lasted over 500 years and 76 over 1,000. The longest-lived state in the register is Eridu, 2,800 years, from 5400 to 2600 BCE, and six cultures span longer than that: the Ancestral Puebloans, the Maya civilisation, the Jōmon culture, the Mehrgarh culture, the Natufians and the Sámi.

The living cultures can be set against the register in another way, by counting the recorded states, in every region of the world, that began and ended within their span. Within the span of the Sámi, since 1000 BCE, 5,080 recorded states have begun and ended; within that of the Mapuche, since 500 BCE, 4,757; within that of the Taíno, since 800 CE, 3,785; within that of the Ainu, since 1200 CE, 2,942; and within that of the Māori, since 1300 CE, 2,590. A Māori hapū tracing its whakapapa to the waka of the fourteenth century has, on the register's count, outlived more than two and a half thousand states, and a Sámi siida almost the whole register. The figures do not say that cultures are better than states, or more just, or that their members lived longer or more safely, and some of the states in the register, Cahokia and the Tuʻi Tonga Empire among them, were built by the descendants of these very peoples. They say only that the unit historians count most carefully, the state with its dates of founding and fall, has usually been the shorter-lived of the two, and that the long dawn before the state has in some places not yet ended.