Study

Ubaid Culture: The World Before Sumer’s Cities

An extended study of a topic, with evidence and references.

Pottery, water, houses, networks and Eridu: how can we read the transformations preceding Sumer’s cities without assigning Ubaid communities an unproven language, state or divine names?

Ubaid bowl from Eridu cemetery, grave 136; mid-sixth–fifth millennium BCE. Metropolitan Museum of Art, 49.133.2. A specific archaeological object, not a representation of all Ubaid potteryThe Metropolitan Museum of Art — Gift of Egypt Exploration Fund, by exchange, 1949 · CC0Object record

Record and revision details

These are descriptive fields, not conclusions.

Edition: 1

Dates

Dates have distinct roles. The timeline uses creation, composition or historical coverage, rather than publication or discovery dates.

Historical coverage: Approximately From 6500 BCE to 3800 BCE

Text

Before Uruk became one of the earliest great cities, and before tablets preserved the names of Inanna, Enki and Enlil, southern Mesopotamia underwent long transformations in its relationships with land, water, work and place.

Permanent villages, painted pottery, expanding houses and connections reaching the Gulf and distant regions were already part of that world. At some sites, buildings exceeded an ordinary household’s scale and possible functions. Historical cities had yet to take shape, but several conditions that would support their development were emerging.

Today we call this world Ubaid culture. It predates the texts that let us read its inhabitants’ and deities’ names, so its study must distinguish material evidence from interpretations of that evidence.

What is Ubaid culture?

Ubaid culture is an archaeological designation for material traditions and chronological phases that began in southern Mesopotamia, with some elements subsequently spreading more widely. It is not an ancient people’s documented name, nor does it identify a known language. The designation comes from Tell al-Ubaid near Ur, rather than a document left by these communities.^[1]

Its beginning and end were not identical everywhere. A broad southern framework including phase 0 and terminal phase 5 extends approximately from 6500 to 3800 BCE. Some reference works and museum collections use 5500–4000 BCE or other ranges for portions of these traditions; northern Mesopotamia has a shorter sequence.^[1]^[3]

Different dates need not mean that one is correct and all others mistaken. We must first ask: which region, phases and definition of Ubaid? This study’s dates provide a broad approximate frame, not a claim that every site was Ubaid throughout the same interval.

A long sequence, not a single moment

Southern traditions are commonly divided into Ubaid 0, particularly associated with Tell el-Oueili; Ubaid 1, or the Eridu phase; Ubaid 2, associated with Hajji Muhammad; and phases 3 and 4. Some sequences add a terminal fifth phase. These divisions order changes in pottery and architecture; they do not represent separate peoples appearing and disappearing with each number.^[1]

“Ubaid communities” is therefore modern shorthand for societies sharing aspects of material culture. It cannot automatically establish a single ethnic identity or language, or the same economy, rituals and social organization wherever similar pottery occurs.^[2]

Southern Iraq before cities

The southern heartland lay in the alluvial plain of present-day southern Iraq. Rivers, marshes, farmland and aquatic resources shaped village distribution and livelihoods. The Lagash project’s overview describes communities drawing on cereals, fish, birds and shellfish across varied floodplain, marsh and Gulf environments.^[4]

This landscape was not a fixed map. Watercourses, sedimentation and shorelines changed, and reconstructing them requires geomorphological evidence and local dating. A single outline cannot show where the Gulf lay throughout the Ubaid period, nor can historical Lagash’s irrigation systems be projected onto much earlier villages.^[4]^[10]

Settlement, production and relations between communities were developing, with differences in size and organization. These villages were not simply waiting rooms for cities: they were complete worlds inhabited for many generations.

Water: managing the landscape without assuming an early state

Southern Mesopotamia’s rainfall generally cannot sustain extensive, reliable cultivation on its own. Access to water and its management consequently mattered to farming. General accounts describe Ubaid irrigation agriculture alongside the exploitation of rivers and marine resources.^[3]

Yet “irrigation” need not initially mean a vast network run by a centralized state. Water management could include using natural channels and river breaches, subsequently modifying them or digging local canals. Each form must be dated before attributing to it an administrative apparatus for which we lack evidence.

A study of the Eridu region published in September 2026 combines remote sensing, fieldwork and dating of samples from the water landscape. The oldest date among the studied samples falls approximately within 4540–4443 BCE, while other examples of more regular networks belong to later periods. This dates a sample in a particular context, not definitively every canal’s construction, or a mature irrigation system throughout the south from 6500 BCE.^[5]

The research makes relationships between water and settlement testable, rather than accepting a ready-made story of a state born to dig canals.

Pottery: a traceable signature, not a complete identity

Among the best-known Ubaid objects are pale vessels, sometimes greenish or yellowish, bearing dark bands, lines and geometric designs, occasionally with more complex motifs. Changes in form and decoration help order phases and connect sites, but similarity does not turn every potter into a member of one society.^[1]^[2]

This study’s cover photograph shows a specific Ubaid bowl in the Metropolitan Museum of Art, excavated from grave 136 in Eridu’s cemetery. The museum dates it to the middle of the sixth–fifth millennium BCE. Its documented context matters; it cannot alone stand for all Ubaid pottery.^[6]

An industrial style can spread through learning, imitation, exchange and the movement of people. Vessels alone cannot determine whether their arrival elsewhere reflects migration, trade, gifts or local production influenced by another community.

A broad interaction sphere

Ubaid elements extended into parts of northern Mesopotamia, Syria, southeastern Anatolia, southwestern Iran and the Gulf coast. Their presence differs regionally: imported vessels, influenced local traditions or wider combinations of shared practices may be involved.^[2]^[3]

Some researchers therefore describe an interaction sphere connecting diverse communities, rather than a Ubaid empire or one people occupying every site with this pottery. Gil Stein’s analysis emphasizes local identities and economic and social variation within the horizon.^[2]

Objects can travel farther than their makers. Techniques can spread without a language or political authority moving with them. This distinction prevents a pottery distribution map from becoming a map of an unproven state.

Distant materials and networks of contact

The southern plain supplied clay, land and water, while certain crafts required materials from elsewhere. Connections formed part of a wider world of exchange, kinship, movement and intermediaries. They need not resemble modern markets or the trade of later kingdoms.

A geochemical study published in 2016 traced obsidian from Syrian, Iraqi and Gulf sites across periods including the Ubaid and other phases. Such analyses identify volcanic sources and routes by which materials became available, but cannot alone identify traders, payment systems or the political organization governing movement.^[7]

Ubaid pottery in the Gulf is an important indication of contact, provided the object’s presence is distinguished from proposed explanations of its journey.^[3] We need not invent traders serving a unified Ubaid state to recognize that villages were connected.

The house and the architectural language of institutions

Several Ubaid sites contain tripartite plans: a central hall with rooms on either side. Examples and functions vary, and plan alone cannot classify a building. An architectural language serving a house could also serve a structure extending beyond a household.^[2]

This suggests an appealing possibility: some later institutions may have borrowed, enlarged and specialized the household model. Jason Ur proposes understanding Mesopotamian institutions and cities partly through transformations of household organization and relationships, rather than the sudden arrival of bureaucracy without antecedents.^[8]

This is an interpretive tool, not proof that every temple began as a house or every large house became a temple. Functional changes require evidence of use, contents, spatial arrangements and local history, beyond architectural resemblance.

Were houses equal?

Some sites show differences in house size and possessions without necessarily presenting the royal classes known from later periods. Tell Abada in the Hamrin basin preserved a fifth-millennium BCE Ubaid village with houses, pottery workshops and varied facilities through three occupation levels.^[9]

One structure, House A, was larger than others and associated with distinctive objects and features. Researchers discuss whether such differences reflect higher-status households or functions extending beyond them. House size alone makes neither a king nor an inherited system of authority.^[2]^[11]

Status could develop through access to distant resources, organizing work, memories and burials, or collective activities. These possibilities need evidence; they allow more than a choice between perfect equality and a fully stratified state.

Work beyond the household

Some northern sites during the later Ubaid provide large buildings, production and storage activities, and sealing evidence, enabling investigation of organization beyond a single house. A recent study discusses Tell Feres and other examples within a northern framework of roughly 5200–4500 BCE.^[11]

But a large Hamrin house, a northern public building and southern Eridu’s sequence cannot simply be combined into a system supposedly present in every Ubaid village. These examples suggest organizational trajectories differing in time and place, rather than a uniform social model.^[2]

An institution may emerge when a collective role continues beyond one household: production, storage, assembly or ritual. Such a role still does not conclusively define how a community was governed.

Eridu: continuity of place and limits of interpretation

Eridu, or Abu Shahrain, is a key southern site for studying architectural continuity and change. Excavations in the 1940s uncovered a famous sequence of eighteen architectural levels in the “Temple Sounding”, contributing to Ubaid periodization. The modern expedition is re-examining the complex and surrounding landscape.^[10]

It is tempting to turn that sequence into a simple story: a small house, then a larger house, then a temple. Yet a structure’s position beneath a later temple does not automatically establish its function. Each level’s installations, finds and use require consideration, including evidence lost or removed by later construction.

The report of the 2022 campaign, published in 2025, particularly investigates the complex, terraces and later architecture. We do not use it to claim that every earlier level was conclusively a house or a temple; that would require more specific analysis than position in the sounding alone.^[10]

Eridu ziggurat remains, photographed by David Stanley on 11 April 2016. The visible architecture postdates the Ubaid; it is not a photograph of its earliest temples or evidence of their deity’s identityDavid Stanley — CC BY 2.0 · CC BYImage source

The photograph shows Eridu’s ziggurat remains from periods later than the Ubaid. It illustrates the place’s continuing importance, not one of its earliest Ubaid buildings.^[13]

Can we read Enki’s name in Ubaid Eridu?

Historical Mesopotamian traditions associated Eridu with Enki, the E-Abzu temple and the deep fresh waters of cosmological belief. Knowing that identity from historical texts and architecture does not automatically carry it into every earlier layer at the site.^[10]

Sacred significance may have continued, the sacred figure or its interpretation may have changed, or a name may have had an oral history before its earliest surviving written occurrence. These are different possibilities, which Ubaid pottery cannot distinguish.

Two statements must remain separate: Eridu was associated with Enki in historical times, and the sacred being’s identity in its earliest phases is unknown to us. Spatial continuity is significant evidence, but it is not a written divine name.

Seals and tokens: distinguishing things before writing

Seals were used before the Ubaid, but also appear within its traditions, with varied images and practices. A stamp seal is a carved surface that leaves an impression when pressed into soft material. Depending on context, it may help distinguish a person, object or procedure.^[3]^[12]

Ubaid stamp seal and modern impression; Syria or Anatolia, sixth–fifth millennium BCE. Metropolitan Museum of Art, 1984.175.13. The rectangular block on the right is a modern impression, not an ancient tabletThe Metropolitan Museum of Art — Bequest of Lester Wolfe, 1983 · CC0Object record

The illustrated Ubaid seal is attributed by the museum to Syria or Anatolia and dated to the sixth–fifth millennium BCE. Beside it is a modern impression clarifying the design, not an ancient tablet. Animal imagery does not reveal its maker’s beliefs or the precise function of every impression.^[12]

Clay tokens and seals indicate ways of distinguishing and organizing things. Marks alone do not make them writing, nor establish a direct chain from every Ubaid token to every later cuneiform sign.^[3]

Was the Ubaid a state? Was it Sumerian?

We lack Ubaid royal names, taxation records or a written apparatus allowing the description of a state as in historical Mesopotamia. Buildings and resources suggest possible inequality and organization, but political classifications depend on definitions of the state and local evidence. The absence of writing does not prove the absence of politics.^[2]^[11]

Nor do we securely know the inhabitants’ language. Population and cultural continuity may connect the Ubaid south with the world in which Sumerian was later written, but pottery and architecture cannot determine language. Ubaid is not simply another name for Sumerian, and our lack of linguistic knowledge does not prove that an earlier form of Sumerian was never spoken.

The sounder approach is to investigate continuity and change, leaving linguistic identity open where no texts answer the question.

Towards Uruk and writing: an uneven transition

During the late fifth and fourth millennia BCE, parts of the south developed features associated with the Uruk period: larger settlements, wider specialization and institutional organization, and changes in recording practices and architecture. Earlier traditions contributed, but the transition was neither a single day nor an identical trajectory everywhere.^[1]^[2]^[3]

At late-fourth-millennium Uruk, administrative accounts became central to the earliest known written record. Tablets were not solely commodity accounts, however: they also included lists of professions and signs. Writing cannot be reduced to an inevitable next step after an Ubaid seal, or to one purpose.^[14]

The emergence of cities and writing needs explanations combining scale, resources, relationships and technologies. No single cause—irrigation, temples or trade—explains the whole history.

The world before the city

Ubaid culture offers a long history preceding written names: villages managing water, unequal houses, specialized crafts, traveling objects, buildings potentially exceeding households, and places rebuilt through generations.

Collective activities may intersect with ritual; ritual may require organization; administration may confer status. These are possibilities and multiple historical paths, not a compulsory ladder by which every house becomes a temple and every village a state.

The Ubaid was more than a footnote before Sumer. It was a world with traditions, communities and differences, whose languages and divine names we cannot read with certainty. In some regions, however, we can examine institutions growing beyond households, the accumulating significance of place, and changing conditions that later supported cities and writing.

Before tablets let us read Enki, Inanna or Enlil, these communities had spent many centuries building a social and material world larger than a single home. That is what archaeology offers: transformations before texts, with a clear distance between the building we see and the name we cannot read.