ANT 417S

 Updated 27 December 2003
 The Origins of Villages  

Course Schedule Evaluation Readings FAQ ARH Prog Anthro Events

 Introduction Today's lecture concentrates on two of your readings, since I dealt with Wilson in some detail last week. I'll begin with Flannery's (1972) article on the origins of villages as a settlement type, along with mention of some sources outside your reading package that are relevant to Flannery's argument (including Flannery's update of 2002). Later we'll consider the quite different approach in Hodder's Domestication of Europe.
  K. V. Flannery (1972). The origins of the village as a settlement type in Mesoamerica and the Near East

Flannery's paper is now a bit dated, as considerable research since 1972 has altered some of the evidence for early villages. In addition, as we'll see, there are some other potential flaws to the argument. Among the updates are Flannery's own article (Flannery 2002). However, the original paper also makes some quite interesting points and deserves close attention.

One of the points that remains valid is that, as mentioned last week, reasonably permanent settlements, including even villages, actually preceded agriculture in some parts of the world, including the Near East.

Flannery suggests that we should not simply assume an association between agriculture, sedentism, and villages. Each can occur without the others, including even villages that are seasonal instead of fully sedentary.

Flannery's main point is that there are major differences between the trajectories of Mesoamerica and the Near East. At about the time that sedentism began in the Near East, the economy included an important component of hunting fairly large ungulates that had herding habits, particularly gazelles, while other food resources were fairly abundant and concentrated. In Mesoamerica, however, the only reasonably large ungulate hunted was white-tailed deer, which does not form large herds, and food was scarce in the dry season. The former tended to encourage the formation of larger hunting parties. The latter tended to encourage people to form smaller, more dispersed groups that hunted and gathered in a more "patchy" way.

Consequently, according to Flannery, Near Eastern hunter-gatherers of the Epipalaeolithic tended to settle in somewhat larger corporate groups, with 3-8 men, possibly twice that many women and three or four times as many children, which he calls circular compounds. In Mesoamerica, he thinks, the main corporate and residential group was already something like a nuclear family (one man, one woman, and a few children).

Flannery attempts to marshall archaeological evidence that Near Eastern Epipalaeolithic sites were compounds, contrasting with later sites in both areas that were villages.

For the number of people who lived in the sites, and in their constituent huts or houses, Flannery depends mainly on Naroll's paper concerning the relationship between roofed floor area and population size. Narroll presented a graph with village population on the x-axis and roofed area on the y-axis. This suggests that, on average, we would expect about 10 square metres of roofed area per person. Later we'll return to this evidence.

Flannery also depends on Whiting and Ayres's article concerning the relationship between house shape and sedentism. This suggests a fairly strong relationship between circular dwellings and nomadic or semi-nomadic societies, and between rectangular ones and fully sedentary societies.

Flannery notes that rectangular structures tend to replace circular ones over time. In the Near East, Pre-Pottery Neolithic B rectangular houses replace PPNA oval huts or house-pits. In the American Southwest, rectangular pueblos replace round pit-houses.

Flannery makes the useful observation that rectangular architecture facilitates the accretion of rooms.

"Circular Hut Compounds"

Here Flannery bases his comparisons on modern examples from various parts of Africa. In fact, they vary quite a bit, but Flannery claims that the following characteristics are quite widespread:

1. Compounds consist of a group of circular huts, each designed to house one, or at most two, people.

2. A typical residential hut in a compound is about 3 m in diameter, or about 7 m2 in area, and houses one adult.

3. Some of the huts are not used for sleeping, but as kitchens, stables, storage buildings, or reception and entertainment of guests.

4. Huts are often arranged in a circle, surrounding a courtyard where many household activities, and possibly storage, occur.

5. Storage of food is mainly joinly owned by the whole compound, or at least kept centrally even if some granaries are technically owned by individuals.

6. As in hunter-gatherer bands, the compound's residents form a corporate group that shares in labour as well as consumption of food.

'Ain Mallaha (right): A Natufian site in Israel, dating around 9 000 BC, has a large number of stone-founded huts, with diameters ranging from 2.5 to 9 m. Flannery uses special pleading to account for the fact that the huts range up to 64 m2 in area. Flannery construes the distribution of the huts as forming an arc, but that was probably due to incomplete excavation at the time.





Nahal Oren (right): With Natufian and Pre-Pottery Neolithic A, probably dating around 8 000 BC, the huts were arranged on terraces and range from about 4 to 15 square metres in size.









Tell Mureybet: The Natufian-like levels I-VIII have traces of a few huts 5 to 14 square metres in area, plus one estimated at 112 square metres in area.

Beidha: This Pre-Pottery Neolithic B site, dating from about 8 000 to 7 000 BC, displays the transition from only circular huts (early levels) to only rectangular houses (upper levels). The huts have areas ranging from 5 to 28 m2, while the later houses have areas that Flannery describes as about 30 m2, but in fact he is including the thickness of the very thick walls. The preserved floor areas are actually only about 8 m2, and were probably basements. The roofed floor area above was probably about 20-25 m2, and thus rather similar to that of the earlier huts.

The hut compounds do not seem to occur in Mesoamerica (although note that there are later plaza groups).

"Villages of Rectangular Houses"

Here Flannery has examples from the archaeology of the Near East and Mesoamerica. Flannery claims the villages tend to have the following characteristics:

1. They consist of rectangular houses designed to house families, and are mostly 25-35 square metres in area. Rather than "family," most archaeologists today would use the term, "household."

2. Each household controls its own storage facility or granary. Flannery claims that the household stores were "not subject to the same kind of obligatory sharing as those of the compound."

3. Rectilinear architecture makes it easy to add or subtract rooms, an important observation related to the need for ease of renovation in response to changing needs of the household where there is sedentism over long periods. In Mesoamerica, the houses are also arranged around rectangular plazas, the plaza groups forming the standard units of the village.

4. Sodalities, which cross-cut kinship relationships, help to unify the villages. For example, men of different households but the same sodality, such as a dance group, could collectively build houses, clear land, and then divide the landholdings among their households.

5. Although some of the activities of sodalities, such as hosting fiestas, can act as an economic "levelling mechanism," living in households, Flannery claims, increases the opportunity for some households to accumulate unusual amounts of wealth.

Some points of critique

As often happens in archaeology, Flannery misuses Naroll's evidence for the relationship between roofed area and population size. He forgets that this relationship is not based on the amount of space dedicated to each person in a sample; instead it is based on the aggregate roofed area of whole villages as compared with their total populations. Therefore he is not right, or perhaps right for the wrong reason, when he says on p. 31 that the fact that some huts are not used for sleeping would lead us to overestimate the population of a compound. Certainly this would happen if, as Flannery suggests, we treated the number of huts as significant. But Naroll's estimates are based on total roofed area and, despite their other faults, are not biased by the fact that some roofed area is for storage, kitchens or entertaining. In addition, Naroll's association has large statistical errors, which reflect the tremendous variability among, and even within, societies, in the number of people per unit area. You simply cannot conclude that a hut housed only one person just because its area is only 10 square metres, or that it could not have been used for living or sleeping if it is smaller than this.

Some of his contrast between compounds and villages is due to a confusion of scale. Most of the compounds he is mentioning are really more analogous to houses than to settlements, the huts being essentially rooms. His comments about sharing within compounds, for example, is equally true of households, as we will see later in the course. Sharing of food within compounds is in fact directly analogous to sharing within households. In addition, some villages and towns are actually made up of many compounds or, in Mesoamerica, plaza groups that may be structurally similar to compounds.

Although Flannery claims on p. 42 that "it is not the 'circular' or 'rectangular' shape of the house [that] is critical, but whether it is intended for a single individual or family," in fact he seems rather heavily influenced by house shape, and tends to assume, automatically, that the circular and oval buildings belong to compounds, even where there is little or no evidence that they are arranged around an open space.

In fact, compounds can have rectangular huts in them, and Prof. Michael Levin from this department has shown that in modern compounds the decision to make the huts rectangular can be based simply on the availability of sheet metal roofing material, which is considered more prestigious because it is imported. Saidel's paper discusses other factors, such as anticipated mobility, that contribute to rectangular or round shape, and next week you'll see Hunter-Anderson's views on this.

Furthermore, some of the examples Flannery mentions (such as the reverse trend toward circular buildings after the Samarran in Iraq, mentioned on p. 43) have nothing to do with compounds. The Halafian circular houses to which he must be referring are all stand-alone houses, or are associated with large, rectangular complexes, but are not grouped in compounds. Even the Epipalaeolithic ones Flannery gives as examples are probably not huts belonging to compounds. The round huts at Nahal Oren, 'Ain Mallaha and Beidha, once you exclude the tiny ones that were probably for storage, all have distributions of floor area that overlap with those for the rectangular houses in villages (e.g., around 15 square metres in the former, and 18-30 square metres in the latter). At least one of the 'Ain Mallaha huts and one of the Mureybet ones are in fact quite a bit larger than most Neolithic rectangular houses. Flannery uses special pleading in an attempt to dismiss these.

In Whiting and Ayres's article, which Flannery cites, and elsewhere it is clear that both compounds and polygyny are strongly associated, not with hunter-gatherers, but with agricultural societies in which women's and children's labour is highly valued for agricultural and food-processing tasks. This provides further grounds to doubt that the huts Flannery discusses constituted compounds.

In his 2002 article, Flannery revises his view by agreeing that the huts of recent hunter-gatherer bands are probably better analogues to Epipalaeolithic huts than are the compounds that he favoured in his original paper. With this revision, key parts of Flannery's argument, such as his emphasis on shared storage in the Epipalaeolithic and PPNA giving way to storage under the control of nuclear families later on, remain of interest (but note that the scale of storage and role of private property are not entirely clear in this time range).

Adaptive Advantages of Villages

Flannery makes a number of other interesting points:

1. The collective land use by sodalities prevents individual families from dispersing onto isolated homesteads. The fact that each household participated in the use of land in widely separated locations around a nucleated village gave them some security and allowed differentiation of activities (specialization).

2. Villages that stayed nucleated and fairly large must have had some political mechanisms that tended to prevent fission.

3. These political mechanisms were also necessary to intensify production (probably leading to plant domestication). The fact that households controlled their own storage allowed some to accumulate surplus and become wealthier than their neighbours.

4. Nucleated villages are better capable of defence. Furthermore, isolated compounds would not easily defend themselves against nearby villages, and would have to move away, form their own villages, or merge with existing villages.


Saidel and Flannery's Response


    Hodder's Domestication of Europe

Hodder takes a quite different approach. What interests him is less the economic aspects of social changes that took place among the early village dwellers of Europe and more the ideological and symbolic.

Having first explored the symbolism of the Neolithic site of Çatal Hüyük in Turkey, which is rich in iconographic art and human burials, Hodder saw associations that led him to the structuralist oppositions,


Hodder argues that oppositional structures like the ones just mentioned are not just mental. They are publicly reinforced or reinterpreted in the context of people's daily lives and social practices.

"If the symbolic structures are produced in the social realm then there is likely to be a close tie to social structures" (Hodder 1990: 13).

Borrowing from Jacques Cauvin's work on the Neolithic of the Levant, Hodder was intrigued by the possibility that control over the wild was, in the Neolithic, "a metaphor and mechanism for the control of society."

For most of the book, Hodder concentrates on the possible relationship between symbolism of the Neolithic and the beginnings of agricultural society. In the European context, this is also a "domesticated" society in the sense of living in villages.

Lepenski Vir, the first site he discusses, is in fact a pre-agricultural village in the Iron Gates region of the Danube, dating around 5500 BC. It is contemporary with Neolithic villages in the Near East, and shares many characteristics with them, including elaborate art, except that it lacks agriculture and animal husbandry.

Its trapezoidal huts or houses have plastered floors, elaborate hearth areas, and are organized over the site in a patterned way that suggests, to some people, some kind of planning. As we will see later in the course, this does not necessarily mean that there was an overall plan, but simply a set of principles, maybe subconscious ones, that guided people's selection of places to build.

It is noteworthy that there is always a somewhat larger house or two near the centre of the village, although all the houses seem to have roughly the same level of wealth.

Rather than social control based on a concentration of wealth, Hodder suggests that at Lepenski Vir it was based on symbolic control over death and the wild. He discusses the distribution of sculptures, animal parts and human burials relative to the hearth and parts of the houses. He notes an association of death and jaws with the hearth, of human dead with stag antlers, and sculptures of human heads with the back of the house. Hodder suggests the oppositions,


but he admits that burials are also found at the front, light part of the houses.

He also notes that the hearth area, which does seem to have a lot of symbolic import, is clear of bone and stone tools. These occur in the peripheries of houses. Although it is possible that this is just a function of housecleaning habits, he suggests the possibility of a centre:periphery::ritual:mundane opposition.

Later in the book, he notes that SE European Neolithic sites such as Polyanitsa exhibit, not only houses with their inevitable inside:outside distinctions, but the whole village can be construed as having an inside and outside. Inside the village, the environment his artificial, controlled, and "domesticated." The outside, which is natural or wild, is separated by a palisade.

In an attempt to avoid the value-laden terminology associated with our own language, he attempts to coin terms that evoke these distinctions while not having so much of the connotations of our modern terms. Thus he recasts the distinction as domus:agrios. Domus, a Latin term pertaining to houses and connected with our modern term, domestic, he uses to label the inside where the wild is excluded or controlled. Agrios (having to do with fields), by contrast, pertains to the wild, natural world outside the village. Of course we find it difficult entirely to avoid the value-laden nature of the terminology. Hodder still needs to explain what he means by these terms, and we also cannot avoid associating them with such related English terms as "domesticated" and "agriculture."

As the Neolithic spread northwards and westward across Europe, some other types of settlements also exhibited this inside:outside::domus:agrios distinction, including later hillforts with ditches and mounds encircling them.

Another thing that Hodder notes is that the barrows of Western and Northern Europe appear to echo the Neolithic houses of central Europe. Central European longhouses of the Linearbandkeramik (LBK) and associated cultures possibly housed multi-family groups much as Iroquoian and Northwest Coast houses did much later in North America. By the time long barrows were constructed in Western Europe, such settlement were no longer based on such longhouses, but the communal burials in long barrows possibly evoked the tradition of longhouses and symbolized the domus and the sense of community that might have been associated with both multi-family households and burials in which individuals' bones were all mixed together.

Many of you may feel more comfortable with the kinds of materialist analyses represented, for example, by Flannery's 1972 article, but it is important not to exclude the possibility that symbolic aspects of culture had an important influence on cultural trajectories. Even if we cannot be certain that Neolithic people would recognize or accept as valid the structural distinctions that Hodder cites, they are not simply figments of his imagination as long as we can find them consistent with impirical evidence. Structuralist arguments are strengthened when we can show that they seem consistent with a variety of types of evidence, and not only the evidence from one realm, such as houses.

 Readings   Recommended references

Flannery, K.V. (2002).

Ingold, Tim (2000). Building, dwelling, living: How animals and people make themselves at home in the world. Pp. 172-188 in The Perception of the Environment: Essays in Livelihood, Dwelling and Skill. Routledge, London.


Saidel, B.

Whiting, and Ayres

Last Update 27 December 2003

Contents and design copyright E. B. Banning 2001-2004