I would firstly like to thank Khaled Abu Jayyab and Johnny Samuele Baldi, the organisers of the 12th ICAANE workshop, Delineating the end of a world: chronological indicators and material-cultural changes at the end of the Ubaid period in Greater Mesopotamia, for the invitation to contribute to this special volume. I would like to offer my deepest appreciation to Mohammad Sarhan, co-director of the Zeidan excavations, and to the entire field crew, especially Kate Grossman, Khaled Abu Jayyab, Tate Paulette, Abbas Alizadeh, Fahd Sbahi, and Kathryn Hanson. This project was of course only possible due to the support and guidance of project director Gil Stein. Funding for the Tell Zeidan excavation project came from the National Science Foundation, the Sigma Xi Foundation, and the Helen Rich Memorial Fund. Finally, I must thank the anonymous reviewers for making essential suggestions to improve the text and figures. All errors are, of course, my own.
1The boundary between the Ubaid period (ca. 5500-4600 BCE) and the Late Chalcolithic period (ca. 4600-3100 BCE) in Greater Mesopotamia has traditionally been both a significant empirical challenge to interpret (Finkbeiner 2001; Stein 2012) and a productive topic for exploring systemic change and the development of social inequality (Akkermans 1989; Lupton 1996; Wengrow 1998; Frangipane 2012). Changes to material culture throughout the 5th millennium BCE appear to have been gradual overall but asynchronous across space (Baldi 2016; Fisher 2017: 31-34; Abu Jayyab and Gibbon, this volume). However, settlement data demonstrate that Ubaid-period town-centric landscapes eventually grew into, were subsumed by, or abandoned in favour of multi-tiered proto-urban settlement systems in the Late Chalcolithic period (Wilkinson and Tucker 1995; Brustolon and Rova 2007; Ur 2010), when conspicuous consumption of wealth goods (Rothman 2001) and circumscription of commodities (Alizadeh 2006: 17) are attestable. This contribution to the volume Delineating the end of a world: Reassessing the Ubaid/post-Ubaid transition in Greater Mesopotamia examines the earliest Late Chalcolithic levels at the prehistoric site of Tell Zeidan, Syria, in order to better understand how Mesopotamian society underwent major changes in the 5th millennium BCE.
2The Late Chalcolithic 1 (LC1) period in northern Mesopotamia (ca. 4600-4200 BCE) has been difficult to apply as a unit of chronometric demarcation (Nissen 2001: 170), but it is also a period that offers great potential for investigating the developmental modalities of social complexity in the broader region. It is a time of both gradual and punctuated change, evidencing developments in rapid ceramic production (Kennedy 2012), mortuary differentiation (e.g., Ball 1997: 96-99; Marro et al. 2011: 80), increased long-distance trade (e.g., Tobler 1950: 193; Rothman 2001: 380), increased human representation in glyptic art (e.g., Esin 1994; Hole 2010: fig. 15.7-8), and localised trajectories of material culture (Fisher 2017). Over the past decade, study of the LC1 period has benefitted from excavations in Syria (Trentin 2010; Ur 2010; Stein 2011; al-Kuntar and Abu Jayyab 2014; Baldi 2015; Abu Jayyab et al. 2020), Turkey (Parker et al. 2006; Kennedy 2012; Koizumi et al. 2016; Fuensanta et al. 2020); Iraqi Kurdistan (Kopanias and MacGinnis 2016; Nieuwenhuyse et al. 2016; Wengrow et al. 2016; Peyronel et al. 2019; Potts et al. 2019; Stein and Fisher 2019, 2020; Renette et al. 2021), and southwestern Iran (Alizadeh 2014). As publication of these projects helps alleviate the empirical lacunae of the LC1 period, questions still remain as to when, how, and why the Ubaid period—and its modes of sociopolitical organisation—ended. This paper considers these questions from both empirical and theoretical perspectives. It addresses the former by presenting and assessing a site-wide dataset of LC1-period artefacts and architecture from Tell Zeidan. It approaches the latter by proposing a model of sociopolitical change that draws on cognitive science studies to highlight innovations in craft production as a means for promoting material variability and the precipitating conditions for sociopolitical reorganisation.
3Tell Zeidan is a 12.5 ha multi-period archaeological site located in northern Syria, approximately 5 km east of the city of Raqqa, at the southern end of the Balikh River Valley and the confluence of the Balikh and Euphrates rivers (fig. 1). The tri-mounded site is primarily prehistoric, with its earliest levels dating to at least as early as the Halaf period, ca. 5800 BCE (Stein 2009: 136, table 1). The majority of the settlement’s vertical deposition dates to the Ubaid and LC1 periods (Stein 2010a: 111, fig. 13), as occupation of the site appears to have contracted during or just prior to the LC2 period, ca. 4200 BCE (fig. 2).
Fig. 1 – Map showing location of Tell Zeidan and other Chalcolithic sites in Greater Mesopotamia.
Fig. 2 – Occupation of Tell Zeidan through time.
4The University of Chicago’s Archaeological Expedition to Tell Zeidan, directed by Gil Stein and Mohammad Sarhan, conducted scientific excavations, site surface collection, and archaeological analyses over the course of three one-month seasons, from 2008 until 2010. The excavations exposed remains of LC1 domestic and craft activities in six different parts of the site, areas A-F (fig. 3), over the course of six successive LC1 levels (table 1). A doctoral research program at Tell Zeidan assigned building Phases to approximate and tentative site levels based on stratigraphy, ceramics, and radiocarbon dating (Fisher 2017). This chapter revisits, summarises, and reconsiders selected research, theoretical arguments, and conclusions from that dissertation.
Fig. 3 – Topographical map of Tell Zeidan showing excavation operations and areas.
Table 1 – Chronology of Chalcolithic Mesopotamia with Tell Zeidan Site Levels.
- 1 Berkman and Kawachi define “social cohesion” (after Durkheim, 1893) as “the extent of connectedness (...)
5The first goal of this paper is to use distributional and comparative assessments across the excavation areas in order to understand the basic composition of Tell Zeidan society in the LC1 period. The underlying questions are “what is the degree of economic interdependence” and “what is the degree of social cohesion”1 between the different areas of the site. To consider the economy, this section compares evidence for craft industries and subsistence activities across the site areas in the following categories: textile manufacturing, lithic manufacturing, food processing, and administration, with brief consideration of faunal and ceramic analyses (to be published fully elsewhere). An assessment of residential architecture, alongside faunal and ceramic data, helps determine levels of social cohesion and inversely, the potential for sociopolitical transformation, especially with respect to durability of form (Coudart 1994: 9-10, 2015: 321).
- 2 Eleven objects are indeterminate; 213 are classified by functional category, function, type, and su (...)
- 3 Fig. 4 shows the distribution of objects, or “small finds”, across Tell Zeidan, which gives a sense (...)
- 4 The use of site-wide levels ease comparison between the mounds but are delineated according to temp (...)
6With only a maximum number of LC1 objects at n = 2242 spread across six tentative site levels, the possibilities for temporally-controlled statistical analysis are limited.3 As a result, this research takes an “intuitive” approach to distributional assessment, displaying the object distributions according to site area and site level to give an intuitive sense of the correspondence between area and functional activity (fig. 5).4 Finally, it addresses architectural variation through observational assessment, again, applying an intuitive/qualitative approach to identifying variability.
Fig. 4 – Total number of objects from LC1 Tell Zeidan by excavation area and site level.
Fig. 5 – Radar chart of object count in functional category by excavation area and site level.
7The following sections consider distributional patterns of objects, or “small finds”, across the three mounds of the site within the five different artefact classes mentioned above, plus summaries of ceramic and faunal analyses, and discusses the implications of those patterns as they pertain to the site’s economy in the LC1 period. In most cases, sampling biases prevent categorical conclusions, but the general trends produce a fairly evident picture of the LC1 economy at Tell Zeidan as having an overall low level of interdependency between mounds. While the sample sizes and exposures are too small to conduct statistically meaningful intra-mound analyses, the interpretation herein suggests that households or groups of households may have had a greater level of economic integration than the overall site, contrasting with Ubaid levels at Tell Zeidan.
- 5 Operation 3 in area B is 12 m2, while operations 19 and 20 in area F are 4 m2 soundings.
8Textile-manufacturing and leather-working artefacts include awls, loom weights, scrapers, abraders, and spindle whorls. Seventeen total pieces come from LC1 levels at Tell Zeidan (table 2). Ten are from areas E, C, and the Step Trench, on the South Mound, and six are from area A, on the Northwest Mound. None come from area B, on the Northeast Mound, nor any from area F (the Lower Town), though this could reflect their limited exposures.5
Table 2 – Textile production and leather working objects from Tell Zeidan grouped by site level (revised from Fisher 2017: table 10).
Site Level |
Area |
ZD No. |
Object Material |
Object Type |
Object Sub-Type |
Context No. |
Ubaid-LC1 |
A |
8453 |
Baked Clay |
Spindle Whorl |
Conical |
100111 |
Ubaid-LC1 |
A |
3717 |
Bone |
Spindle Whorl |
Worked Bone |
100070 |
Ubaid- LC1 |
A |
7146 |
Baked Clay |
Spool |
Cylindrical |
100095 |
Ubaid-LC1 |
C |
2151 |
Clay |
Loom Weight |
Biconical |
90027 |
Ubaid- LC1 |
C |
2574 |
Stone |
Scraper/ Pecking Stone |
Ovoid |
90042 |
Ubaid-LC1 |
Step Trench |
213 |
Clay |
Spindle Whorl |
Biconical |
10013 |
LC1A |
A |
3620 |
Ceramic |
Scraper |
Worked Sherd |
100062 |
LC1A |
E |
8929 |
Bone |
Awl |
Worked Bone |
150078 |
LC1A |
E |
2136 |
Ceramic |
Spindle Whorl |
Pierced Sherd |
10005 |
LC1B |
A |
9099 |
Bone |
Awl |
Worked Bone |
100033 |
LC1B |
A |
2632 |
Baked Clay |
Loom Weight |
Biconical |
100036 |
LC1B |
E |
9100 |
Bone |
Awl |
Worked Bone |
160062 |
LC1B |
Step Trench |
1921 |
Bone |
Awl |
Worked Bone |
60022 |
LC1B |
Step Trench |
86 |
Baked Clay |
Spindle Whorl |
Biconical |
10004 |
LC1B-C |
E |
7820 |
Stone |
Abrader |
Ground Stone |
150058 |
LC1C |
A |
2137 |
Bone |
Awl |
Worked Bone/ Fragment |
20035 |
LC1C |
A |
879 |
Ceramic |
Spindle Whorl |
Worked Sherd |
20035 |
LC1-2 |
E |
9530 |
Baked Clay |
Spindle Whorl |
Conical |
150018 |
LC1-2 |
A |
718 |
Ceramic |
Spindle Whorl |
Worked Sherd |
20026 |
9Because of the general scarcity of textile-production and leatherworking finds at Tell Zeidan, their main value is in contextual assessment. However, plotting weight and diameter of spindle whorls can indicate the kind of fibre being manipulated (Mårtensson et al. 2006: 11; Sudo 2010: 173). For example, Kosak Shamali LC1 levels feature relatively low-weight (<50 g) whorls, implying wool spinning rather than linen, while a declining trend in whorl size and weight suggests finer wools being spun over time (Sudo 2010: 174, fig. 12.3). Tell Surezha, on the Erbil Plain, demonstrates similar trends for the same time period (Price et al., 2021: fig. 3). At LC1 Tell Zeidan, spindle whorls collected from area A (n = 4) and area E (n = 5) suggest that the inhabitant groups may have employed different wool spinning strategies, spinning finer textiles on the Northwest Mound than on the South Mound (fig. 6). Otherwise, both mounds have similar, household-based textile and leatherworking industries. By contrast, in the Ubaid period, 80% of the awls come from area B on the Northwest Mound, suggesting that there might have been a more specialised leatherworking or textile economy at that time. The concentration of whorls and loom weights in building G at Tell Abada (Jasim 1985: 174, 208) is a possible contemporary parallel.
Fig. 6 – Weights and diameters of spindle whorls from Tell Zeidan by Area and Period.
10Evidence from the LC1 period for stone tool production at Tell Zeidan is also scarce, and so, again, the artefacts in this category are more effective for contextualisation than for intrasite analysis. The ratio of (locally sourced) chert cores to (imported) obsidian cores continues a long-term trend at Tell Zeidan in which obsidian makes up 4-7% of the total chipped stone assemblage. However, the number of specifically LC1 contexts analysed in the formal analysis is relatively few (Khalidi et al. 2016: 747). In the Ubaid period at Tell Zeidan, workshops produced obsidian blades onsite (Stein 2011: 129, fig. 7) and imported both raw and finished obsidian from the widest variety of sources demonstrated to date, from the western Caucasus to eastern Anatolia (Khalidi et al. 2016: 748, table 4). In the LC1 period, that pattern appears to have shifted to evidence only a few sources (Khalidi et al. 2016). Furthermore, obsidian cores of the Ubaid period come from the South and Northeast mounds but not the Northwest Mound (Khalidi et al. 2016: table 3), with the greatest concentration of obsidian finds coming from the Northeast Mound (Khalidi et al. 2016: 747). In the LC1 period this pattern changes slightly and the South Mound seems to revert to a reliance on local chert for production, while the Northeast Mound continues to display evidence of obsidian working. Thus, the LC1 inhabitants of Tell Zeidan continued to import and manipulate obsidian, but from different and fewer sources (Bingöl A, Nemrut Dağ, and Bingöl B), with a different local distribution pattern, and in domestic contexts rather than specialised workshops.
Fig. 7 – Top plan of building S.A1, area A, level LC1b, Tell Zeidan, with schematic of storage jar ZD3199.
11Seven of the twelve artefacts evidencing lithic manufacture come from area E (South Mound), which has the broadest LC1 exposures, while just two examples come from area A, on the Northwest Mound (table 3). Along with an obsidian core (ZD291), obsidian flake debitage, collected as bulk chipped-stone, also comes from area B (Northeast Mound). Hammerstones discovered in area B further identify it as a locus of stone tool production. Sampling and recording biases partially obscure the relative frequencies of the finds, but it is clear that all three areas engaged in stone tool production even as access to imported materials became restricted to the Northeast Mound.
Table 3 – Lithic manufacturing objects from LC1 Tell Zeidan grouped by site level.
Site Level |
Area |
ZD No. |
Object Material |
Object Type |
Object Sub-Type |
Context No. |
Ubaid-LC1 |
A |
7149 |
Stone (igneous) |
Hammer- stone |
Cubic |
100095 |
Ubaid-LC1 |
C |
2574 |
Stone |
Scraper/ Pecking Stone |
Ovoid |
90042 |
LC1a |
E |
9653 |
Stone (chert) |
Core |
Pyramidal/ Lenticular |
150094 |
LC1a |
E |
9348 |
Stone |
Hammer- stone |
Cylindrical |
150085 |
LC1a |
E |
9506 |
Stone (chert) |
Worked Stone |
Ovoid |
150085 |
LC1b |
B |
291 |
Obsidian |
Core |
Crested Blade Core |
30049 |
LC1b |
B |
767 |
Stone (chert) |
Pecking stone |
Spherical |
30049 |
LC1b |
E |
7821 |
Stone (chert) |
Core |
Pebble Cortex |
150058 |
LC1b |
E |
9807 |
Stone (chert) |
Core |
Pebble Cortex |
150058 |
LC1b |
E |
9402 |
Stone (meta- morphic) |
Hammer- stone |
Ovoid |
160079 |
LC1b-c |
E |
7927 |
Stone (igneous) |
Pecking Stone |
Spherical |
150058 |
LC1d |
A |
1550 |
Stone (chert) |
Hammer- stone |
Cubic |
100010 |
12The largest number of storage jars and food processing instruments comes from level LC1b on all mounds, which is the LC1 level with the broadest domestic exposures at the site. Food processing objects concentrate in area E on the South Mound, particularly inside of room S.E3.1, although the general distribution indicates that food preparation occurred in the dwellings of all three mounds (table 4).
Table 4 – Food preparation and processing objects from LC1 Tell Zeidan grouped by site level.
Site Level |
Area |
ZD No. |
Object Material |
Object Type |
Object Sub Type |
Context No. |
Ubaid- LC1 |
D |
7453 |
Baked Clay |
Muller |
Crosshatch Incised |
17001 |
LC1a |
C |
1783 |
Baked Clay |
Muller |
Fragment |
90022 |
LC1b |
A |
2642 |
Baked Clay |
Andiron |
Fragment |
100029 |
LC1b |
A |
2843 |
Baked Clay |
Andiron |
Fragment |
100041 |
LC1b |
A |
2619 |
Baked Clay |
Muller |
Crosshatch Incised |
100029 |
LC1b |
B |
298 |
Stone (basalt) |
Grinding Stone |
Fragment |
30049 |
LC1b |
B |
798 |
Baked clay |
Muller |
Crosshatch Incised |
30049 |
LC1b |
E |
9133 |
Stone (basalt) |
Grinding Stone |
Fragment |
160079 |
LC1b |
E |
9123 |
Ceramic |
Colander |
Perforated |
160079 |
LC1b |
E |
7924 |
Ceramic |
Colander |
Perforated |
150056 |
LC1b |
E |
8449 |
Stone (granite) |
Grinding Stone |
Irregular |
150060 |
LC1b |
E |
7190 |
Baked clay |
Muller |
Crosshatch Incised |
160062 |
LC1b |
E |
9101 |
Stone |
Pestle |
Worked Stone |
160079 |
LC1b |
E |
9408 |
Stone (basalt) |
Pestle |
Worked Stone |
160079 |
LC1c? |
A |
1525 |
Baked Clay |
Muller |
Crosshatch Incised |
100006 |
- 6 For comparison, a pithos from Late Chalcolithic Arslantepe (level VIII) has an estimated capacity o (...)
- 7 Cf. Baldi 2015 and Rothman and Fiandra 2016 regarding combination of private and communal food stor (...)
13Evidence for LC1 grain storage at Tell Zeidan comes primarily from two locations, building S.A1 in level LC1b on the Northwest Mound (fig. 7) and building S.B2 in level Ubaid-LC1 on the Northeast Mound. The S.B2 storage jar remains are fragmentary, but building S.A1 features multiple complete, subterranean storage jars with approximate carrying capacities of ca. 270 litres each.6 S.A1 is not fully excavated, although its visible remains imply a simple house plan rather than the larger, more elaborate tripartite design common to the Ubaid period. Its layout recalls Ubaid Structure 4 at Kenan Tepe (Kennedy 2012: fig. 10), whose main room has an approximate area of 24 m2. This would be roughly half of the size of an Ubaid tripartite house central room, which might have accommodated ten non-infant humans (Roaf 1984: 96). Comparatively, then, the S.A1 house might have held around five non-infant individuals. There is no evidence from LC1 Tell Zeidan for a centralised or otherwise institutional system of food production and redistribution, but the remains of building S.A1 represent a relatively robust food-storage profile for such a small building. With a total capacity to feed a small family through a single winter, the house’s known grain stores were not surplus-oriented, but probably represent a household-centric subsistence strategy.7
- 8 These fist-impressed pieces of levigated clay are also known from Tepe Gawra XIII, found in a “well (...)
14Sixty-eight percent of the LC1 administrative finds (table 5) come from a single context in area C on the South Mound, located inside of the Ubaid-LC1 niched-and-buttressed building (Stein 2010a: 109, fig. 10). All of these are all hand-formed pieces of levigated clay, or “sealing preforms” (Stein 2010a: 110, fig. 11; here fig. 8a).8 The single stamp seal from LC1 levels at Tell Zeidan, whose findspot was a pit that stratigraphically postdates the niched-and-buttressed building, was likely an heirloom due to its Halafian iconography (Stein 2010a: 109, fig. 9). Otherwise, the majority of evidence for administrative activity in the LC1 levels comes from area E and the Step Trench on the South Mound and area A on the Northwest Mound. Two notable site-wide absences from across LC1 Tell Zeidan in this category, however, are door sealings and evidence for sealing large storage jars. There is, therefore, no indication of centralised control over commodities such as grain.
Table 5 – Administrative objects from LC1 Tell Zeidan.
Site Level |
Area |
ZD No. |
Object Material |
Object Type |
Object Sub Type |
Context No. |
Ubaid- LC1 |
A |
3723 |
Clay (unbaked) |
Sealing Clay |
Fragment |
100072 |
Ubaid- LC1 |
A |
7685 |
Clay (unbaked) |
Sealing |
Jar or Bag Sealing |
100089 |
Ubaid- LC1 |
C |
2184 |
Clay (unbaked) |
Sealing Clay |
Fragments |
90027 |
Ubaid- LC1 |
C |
2178 |
Clay (unbaked) |
Sealing clay (bulk) |
Preforms |
90027 |
Ubaid- LC1 |
C |
2158 |
Clay (unbaked) |
Sealing clay (bulk) |
Fragments |
90027 |
Ubaid- LC1 |
C |
1797 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2551 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2554 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2556 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2557 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2558 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2559 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2560 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2561 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2562 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2563 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2564 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2565 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2566 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2567 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2552 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
Ubaid- LC1 |
C |
2553 |
Clay (unbaked) |
Sealing clay |
Preform |
90027 |
LC1a |
A |
3041 |
Clay (unbaked) |
Sealing |
Bag Sealing |
100067 |
LC1a |
C |
1798 |
Stone (steatite) |
Seal |
Stamp Seal |
90023 |
LC1b |
Step Trench |
41 |
Clay (unbaked) |
Sealing |
Bag or Jar Sealing |
10002 |
LC1b |
Step Trench |
1925 |
Clay (unbaked) |
Sealing |
Basket Sealing w/ Stamp Impression |
60022 |
LC1b |
E |
7919 |
Clay (unbaked) |
Sealing clay |
Preform |
150066 |
LC1d |
A |
1530 |
Clay (unbaked) |
Sealing clay |
Fragment |
100008 |
LC1-2 |
A |
709 |
Clay (unbaked) |
Sealing |
Bag Sealing? |
20025 |
Fig. 8 – Object images from LC1 Tell Zeidan: a. Clay sealing “preforms”; b. Figural scene on clay sealing (ZD1925); c-d. Coba bowl; e. Beaded-lip bottom-scraped bowls sherds; f. Internally-incised bowl sherd.
- 9 ‘Shamans’ are religious specialists whose power comes from inspiration, as opposed to “priests”, wh (...)
15Three of the clay sealings from LC1 levels feature stamp seal impressions. The earliest (ZD7685) comes from the Ubaid-LC1 level in area A and is a bag or jar sealing impressed on the obverse with a geometric motif of six circles and lines radiating out from the centre and on the reverse by a rope or string. The other two LC1 sealings depict figures, only one of which has a clearly visible motif. ZD1925 (fig. 8b) comes from level LC1c in the Step Trench area. Its impression shows a human striding to the right next to at least one quadruped, resembling figural imagery from Tepe Gawra levels XIII and XII (Tobler 1950: pls. CLXIII, LXXXVIIIb) and, to a lesser degree, Deǧirmentepe (Esin 1994) and Tepe Giyan (Hole 2010). These motifs often feature a so-called “bird-headed” (or bird-masked) man, whom Hole identifies as a namash, or shaman-like figure,9 who has spiritual faculty (see also Pittman 2001: 412). In the Tell Zeidan example, this figure appears with a quadruped resembling what Hole identifies as saluki dogs (Hole 2010: 234). These few instances of evidence for administration and glyptic art tie Tell Zeidan iconography into the broader Mesopotamian artistic repertoire, with the single preform from area E confirming that local sealing practices continued from the Ubaid but at a household level rather than in the potentially monumental, niched-and-buttressed building of the Ubaid-LC1 level.
- 10 K. Grossman performed the analysis of the LC1 faunal remains (forth-coming); see also Grossman and (...)
16Contextual biases and small sample sizes are also present among the faunal data, which nevertheless suggest that individual areas of the site may have employed slightly different animal consumption strategies. The patterns seem to reflect preferential differentiation, or perhaps competition between groups for particular game-animal niches, rather than specialised animal management (see Fisher 2017: 246-252 for further information).
- 11 Many of the bird bones come from a direct discard context, alongside freshwater shell remains and d (...)
17The Northwest Mound displays the most diverse animal economy, with a focus on deer-like species. The South Mound has the second highest number of species attested but seems to favour ovi-caprids, at 20% of its faunal assemblage. It also has the highest percentage of burnt bones, at 5.6%, which counterbalances its lack of evidence for ovens and could suggest distinctive cooking practices relative to the rest of the site. It has the lowest relative amount of pig consumption (1% of its total fauna compared to 2.3-2.4%) and is devoid of bird bones but has the greatest relative abundance of dama Mesopotamica (Persian fallow deer). The faunal profile of the Northeast Mound stands out as having the least diverse animal economy, but it is the only mound with evidence for consumption of lepus (rabbits and hares). It also has by far the highest relative amount of bird remains, at 2% of its fauna, and gazelle, at 1.7%.11
18Overall, the LC1 animal economy at Tell Zeidan continues trends of the Ubaid, in which the meat of domesticated species became an increasingly important component. Part of this trend includes increased consumption of pig meat, especially on the Northwest and Northeast mounds. Despite the overall preference for sheep/goat, and for ovis (n = 53) over capra (n = 12), the survivorship curve indicates a common age of death at 1-2 years old and a maximum age of 4 years old. Thus, the kill-off pattern implies a meat production strategy rather than intensive secondary-product production. The importance of ovi-caprids for Tell Zeidan is attested, however, in their heavy distribution across all three mounds, though in different relative amounts.
- 12 K. Abu Jayyab performed ceramic analysis for the Late Chalcolithic sequence at Tell Zeidan (forthco (...)
- 13 Similar to the “second” type of wide flower pot Bowl identified at Tell Helawa (Peyronel and Vacca (...)
19The rapidly made bowls common to the LC1 period (Tomita 1998) come in three types at Tell Zeidan: the Coba bowl (fig. 8c-d) has a flint-scraped base, as does the more crudely made beaded-lip bottom-scraped bowl (fig. 8e),13 while the wide flower pot bowl (Fisher 2017: pls. 25m, 27j) was typically mould- and/or coil-made. A fourth common, simply made bowl type common to the LC1 period is the internally-incised bowl (fig. 8f; Fisher 2017: pls. 25c-e, 25s, 25x, 27d-e). Other main open forms are club-rimmed pots (Fisher 2017: pls. 24k-l, 26d), inverted bowls with out-turned lips (Fisher 2017: pl. 27b), and grooved-rim bowls (Fisher 2017: pl. 27c). Typical closed forms found in LC1 levels at Tell Zeidan include flaring rim jars (Fisher 2017: pls. 24, 26), coarse cooking pots (Fisher 2017: pls. 24i-n, 26d, 26k), and globular pots with heavy rounded rims (Fisher 2017: pl. 24g). The Ubaid-style painted tradition also continues, but with decreasing frequency from 5% of the total assemblage in Ubaid-LC1 levels to 1.4% by the end of the LC1 period (Fisher 2017: 253-254).
- 14 Ceramic analysis included surface collection data, on which the presence-absence tallies are based, (...)
20Each mound appears to have its own pottery preferences. The ceramic assemblage of the Northeast Mound in the LC1 period generally includes more types than elsewhere, but produced no internally-incised bowl sherds.14 It is, however, the only mound where wide flower pot bowls are present in significant numbers, in addition to both Coba bowls and beaded-lip bottom-scraped bowls. The most common types found on the Northwest Mound are Coba bowls and internally-incised bowls, along with wide flower pot bowls. Other notable characteristics of the Northwest Mound are the absence of beaded-lip bottom-scraped bowls, the early incidence of Red Ware (which is more typical of the LC2 period at Tell Zeidan), and the low diversity of types identified. On the South Mound, beaded-lip bottom-scraped bowls are present and internally-incised bowls are common, while wide flower pot bowls are absent.
- 15 Coba bowls and other rapidly made bowls come from both consumption (Baldi 2012) and storage context (...)
- 16 See Al Quntar and Abu Jayyab 2014: 95 for discussion of the technical implications of chaff temperi (...)
21Ubaid levels at Tell Zeidan show possible evidence of specialised pottery firing areas, such as the pyrotechnic features of area D (Stein 2010a: 108-109). LC1 evidence for ceramic production is less robust, though a vitrified potsherd (likely a kiln waster) and a pair of diamond-shaped kiln spacers from area C attest to on-site manufacturing. Furthermore, there are six examples of vitrified clay, or “pot slag”, from LC1 levels, distributed across most areas of the site. This limited evidence for firing of vessels in the main excavation areas suggests that ceramic production was localised rather than centralised. Regarding technique, production methods differ between Coba bowls, beaded-lip bottom-scraped bowls, and wide flower pot bowls, but their function is fairly interchangeable.15 Thus, the variable distribution of these types may signify that local producers organised themselves differently across the site. In level LC1b on the South Mound, vegetal tempering occurs at approximately double the rate (34%) of that from the rest of the site (16%) at that time, consistent with the interpretation that there were localised, craft economies at LC1 Tell Zeidan.16
- 17 “Agglutinant” architecture is defined here as structures that do not appear to have a clearly pre-d (...)
22Analysis of domestic architecture factors heavily in understanding the transformative capabilities of a society (Coudart 1994; 2015). The tripartite form appears consistently in domestic architecture of the Ubaid period in both central (Jasim 1985; Roaf 1989) and northern Mesopotamia (Esin 1989; Breniquet 1995; Thuesen 2000; Stein 2011; Parker 2012). The LC1 period, however, evidences a wider variety of house types (see Fisher 2017: 286-291) that have sometimes been described in binary terms as either tripartite or agglutinant (e.g., Balossi-Restelli 2010).17 As a result, the data from Tell Zeidan are important for gaining a more complete picture of an architecturally diverse time period. This section will compare architectural construction methods, forms, and functions across the site areas (tables 6-10).
Table 6 – Structural Remains of Level Ubaid-LC1.
Site Level |
Area |
No. |
Building Type Orientation |
No. of Rms. Fnd. |
Construction Techniques |
Features |
Activities |
Rapidly Made Bowls Found? |
Ubaid-LC1 |
A |
S.A3 |
|
1 |
|
|
|
|
Ubaid-LC1 |
A |
S.A4 |
Rectilinear |
1 |
Pisé and mudbrick; levigated pisé (tauf); crumbly mudbrick |
Domed oven |
Cooking; craft manufacture (textiles, lithics) |
|
Ubaid-LC1 |
B |
S.B2 |
Rectilinear Ordinal |
1 |
Double-wide mb walls; single-wide mb walls |
|
Weaponry |
Yes, on floor |
Ubaid-LC1 |
C |
S.C4 |
Niched Ordinal |
1 |
Triple-wide mb wall |
Double-recessed niche |
Admin; textiles; weaponry |
Yes |
Ubaid-LC1 |
D |
S.D1 |
Concentric Rectangles Cardinal/Ordinal |
1 |
Width-and-a-half mb walls; alternating gray and brown rows |
Sherd- lined pit |
|
Yes |
Ubaid-LC1 |
H |
S.H1 |
Singe Wall Cardinal |
|
Double-wide mudbrick wall |
|
|
Yes, nearby |
Table 7 – Structural Remains of Level LC1a.
Site Level |
Area |
No. |
Building Type Orientation |
No. of Rms. Fnd. |
Construction Techniques |
Features |
Activities |
Rapidly Made Bowls Found? |
LC1a |
B |
S.B3 |
Rectilinear Ordinal |
1 |
Black mb; vertical row; 5-brick wide wall; header- stretcher |
Found- ation deposit |
Ritual? |
Yes |
LC1a |
E |
S.E5 |
Agglutinant Cardinal |
4 |
Densely packed, header mb |
Sherd-lined pit; adult burial |
Lithic, textile manufacture |
Yes, collapse |
LC1a |
Step Trench |
S.E7 |
Single Wall and Floor Found |
|
Compact, granular, light-brown, 45 cm bricks; heavy chaff |
|
|
Yes, discard |
LC1a |
H |
S.H1 |
Rectilinear Cardinal |
|
30 x 60 x 10 reddish and grayish mb; half-brick corner join |
|
Ornament- ation (bead) |
Yes, collapse |
Table 8 – Structural Remains of Level LC1b.
Site Level |
Area |
No. |
Building Type |
No. of Rms. Fnd. |
Construction Techniques |
Features |
Activities |
Rapidly Made Bowls Found? |
LC1b |
A |
S.A1 |
Rectilinear Ordinal |
4 |
Single-row mb walls; yellow mudbricks; reddish- brown mudbricks; sherd- tempered bricks |
Sunken storage jars; domed oven |
Grain storage; cooking; food processing; textile production; personal adornment |
Yes; two rooms |
LC1b |
A |
S.A8 |
Single wall found Ordinal |
|
|
Foundation deposit: ovi-caprid mandible |
|
Yes; nearby |
LC1b |
B |
S.B4 |
Rectilinear Cardinal |
1 |
Pale gray mudbricks; double-coat plastering; double-wide mb wall |
Stone and gravel drainage |
Farming; craft manu- facture (lithics); personal adornment |
Yes; nearby |
LC1b |
E |
S.E1 c-b |
Rectilinear Cardinal |
2 |
Stretcher mudbricks; vertical row of bricks |
|
|
Yes; tertiary |
LC1b |
E |
S.E2 c-b |
Agglutinant Cardinal |
2 |
Single-course wide, laid on top of floor |
Sherd- lined pits (2) |
Textile production; food preparation |
Yes; tertiary |
LC1b |
E |
S.E3 c-b |
Agglutinant Cardinal |
3 |
40 cm bricks; side-lying and flat stretchers |
Semi-ovular ash pit; red- painted walls |
Craft manufacture (lithics); food prep; food dispensing |
Yes, one room |
LC1b |
E |
S.E4 |
Rectilinear Cardinal |
1 |
Double-wide wall; grayish- brown and orange-brown mudbrick |
|
|
Yes |
LC1b |
Step Trench |
S.E8 |
Single wall found |
|
Three rows wide; 45 cm bricks; dark brown, crumbly mb |
|
|
Yes |
Table 9 – Structural Remains of Level LC1b-c and LC1c.
Site Level |
Area |
No. |
Building Type Orientation |
No. of Rms. Fnd. |
Construction Techniques |
Features |
Activities |
Rapidly Made Bowls Found? |
LC1b-c |
E |
S.E1a |
Rectilinear Cardinal |
2 |
|
Sherd-lined pit; bin |
|
Yes; tertiary |
LC1b-c |
E |
S.E2a |
Agglutinant Cardinal |
2 |
|
|
|
Yes; tertiary |
LC1b-c |
E |
S.E3a |
Agglutinant Cardinal |
3 |
Dark brown, crumbly mudbricks |
Non-structural partition wall |
|
Yes, discard |
LC1c |
Step Trench |
S.E9 |
Single wall found |
|
Double-wide mudbrick; 40 cm bricks |
|
|
|
LC1c |
A |
S.A2 |
Limited discovery Cardinal |
|
Single-course mb walls; yellow mudbricks; reddish-brown mudbricks |
|
Personal adornment |
|
LC1c |
B |
S.B5 |
Tripartite/ T-shaped hall Cardinal |
1 |
40 cm bricks; 10 cm bricks |
|
|
|
Table 10 – Structural Remains of Level LC1d and LC1-2.
Site Level |
Area |
No. |
Building Type |
No. of Rms. Fnd. |
Construction Techniques |
Features |
Activities |
Rapidly Made Bowls? |
LC1d |
A |
S.A5 |
Single floor found |
1 |
|
|
|
|
LC1d |
E |
S.E6 |
Single wall found |
|
Triple-row mb wall |
|
|
Yes; nearby |
LC1-2 |
A |
S.A6 |
Single floor found |
1 |
|
|
Craft manu- facture: metallurgy, textiles |
|
23Building methods and materials vary widely during the LC1 period at Tell Zeidan. Among the most notable construction techniques are the different ways that mudbricks are aligned within the wall. For instance, in level Ubaid-LC1 in area D, there is a distinct method of using one full-width brick and one half-width brick to form a slightly wider wall. Another particularity of these walls is the alternation of grey and brown brick courses. Neither of these traits are attestable elsewhere on the site.
24In area A, level Ubaid-LC1, the walls of building S.A4 are mixed material, including cob and dense mudbrick as well as levigated pisé (tauf) and crumbly mudbrick. Area B at this time features a building (S.B2) with both a double-row mudbrick wall and single-row walls. In area C, the exterior wall of building S.C4 is three rows wide but uses half-width bricks to create the double-recessed niche.
Fig. 9 – Composite top plan of Late Ubaid and Ubaid-LC1 levels of building S.D1, area D, Tell Zeidan.
25In level LC1a, again there is variation from building to building, although S.C4 is replaced by pits rather than another iteration of the niched-and-buttressed building. Building S.B3 features a 5-row-wide wall made of black, crumbly mudbricks and a middle course of longer bricks laid on their side (fig. 10). The size of the wall and the style of brick are unique at Zeidan, while the side-lying row of brick is attested elsewhere in a later level. The bricks’ consistency in area B contrasts with the compact, granular, heavily chaff-tempered bricks found in the Step Trench (South Mound) in this level. In area E, building S.E5 uses header-oriented mudbricks for multiple walls, which is also unique at Tell Zeidan (fig. 11). Two of these walls also feature alternating colours of mudbrick clay (grey, brown, and white).
Fig. 10 – Top plan of Building S.B3, Area B, Level LC1a, Tell Zeidan.
Fig. 11 – Composite top plan of LC1a and LC1b buildings, Area E, Tell Zeidan.
26In level LC1b, area A features sherd-tempered bricks in single rows of multiple colours in building S.A1 (fig. 7), while in area B building S.B4 has pale grey mudbricks, a double-coat of plastering (one thick, functional layer and one thin, white, aesthetic layer), and double-row brick walls. The architecture on the South Mound (fig. 11) includes the “agglutinant” style of rooms added onto one another and the transformation of Structure S.E2 from a presumably outdoor court in building phase S.E2c to an enclosed space in building phase S.E2b. Building S.E4, though harder to interpret, seems to have also undergone an agglomerative process, with an orange-coloured mudbrick wall abutting against a parallel brown one. Building S.E8 in the nearby Step Trench area features a wall with three rows of mudbrick.
27Level LC1c in area B includes the T-shaped Hall building (S.B5; Stein 2009: 133, fig. 10), and while the remains of this level are less well preserved, especially in area A, it is still apparent that the construction methods of houses continue to vary across the site. Area E at this time is primarily distinguished by the partition of the main space in S.E3a across the northern sections of rooms S.E3.1 and S.E3.2, creating a more secluded area out of a room (S.E3.2) characterised by red-painted interior wall faces. The new walls wind their way through the space in accordance with the pathway designated by the earlier walls, and it appears as if the northern wall was cut diagonally to accommodate the new partition wall (fig. 11). Thus, while building S.B5 on the Northeast Mound is a planned and newly-built structure in accordance with the tripartite architectural tradition, building S.E3a features a lower-precision reworking of the preceding structural form.
28Along with the varying approaches to construction, architectural layout and house style also varied across LC1 Tell Zeidan. Most of the excavated structures at Tell Zeidan are domestic buildings, even where some degree of focused craft production is attested, with the possible exception of the niched-and-buttressed building (S.C4) on the South Mound. This building, dating the to the Ubaid-LC1 level, may represent an administrative facility, though its limited exposure makes this interpretation uncertain. Otherwise, all of the LC1 buildings demonstrate evidence of domestic occupation, typically as occupational debris such as figurines, beads, cooking vessels, and a general multiplicity of function. Beyond that commonality, though, few of the house plans seem to share design principles. The relatively small exposures limit definitive interpretations about the nature of the dwellings, but several observations emerge. Firstly, most of the houses appear not to be based on the tripartite plan. The single possible exception at Tell Zeidan is the T-shaped or cruciform hall of level LC1c on the Northeast Mound (S.B5, area B; Stein 2009: 133, fig. 10). It potentially follows the tripartite plan of T-shaped houses at the Ubaid site of Tell Abada (Jasim 1989), as opposed to the more northern tripartite style seen at LC1 sites such as Tilbes-Körche (Fuensanta et al. 2020) and Kosak Shamali (Nishiaki and Matsutani 2001).
29Secondly, from the Ubaid-LC1 period onward on the South Mound, buildings can be reused in later levels but often with internal addition of rooms. In area D, just east of area E, the rectilinear plan of Late Ubaid building S.D1 (fig. 9) is augmented in the Ubaid-LC1 level with a rectilinear room built in the centre, directly on top of the floor that abuts the still-in-use Ubaid walls, creating a design of concentric rectangles. In area E on the South Mound, the buildings of levels LC1b-LC1c are all immediately adjacent to one another, and the individual structures are either rectilinear or agglutinant (fig. 11). In some of the LC1a-LC1c buildings (e.g., S.E3), structures seem to follow no plan, with new rooms built inside of older ones. In other cases (e.g., building S.E1), the structures are rectilinear and appear more regularly planned but still demonstrate a density of dwelling space. On the Northwest and Northeast mounds, structures tend to be completely rebuilt in successive levels. For example, LC1b building S.A1 (fig. 7), described above, features an approximately rectilinear construction with ordinal orientation, whereas in level LC1c, no part of S.A1 remains and the poorly-preserved architecture has a new orientation.
- 18 Similar pits are seen at Sakçe Gözü (Çoba Höyük) period IVC in the small sounding, but lined with s (...)
30“Functional features” here are immovable elements of architectural space that have low structural value but high functional value. Types such as ovens occur rarely in LC1 levels at Tell Zeidan, presumably due to the limits of excavation, and so comparison is not warranted. The most recurring type of “functional feature” discovered at Tell Zeidan is the “sherd-lined pit”, although good terminology for these is lacking due to their uncertain purpose. These sherd-lined pits are seemingly drainage features for hydro-management, but could have been used as sumps to collect rainwater, sinks, holding pits for round-based jars (M. Gibson, pers. comm.), or something else.18 The dimensions range from 15 to 46 cm in diameter and 10 to 25 cm in depth, with the majority of the LC1 examples falling into the 40-46 cm diam. range.
- 19 This may be a method for retaining heat inside of fire installations (Balossi-Restelli 2012a: 48).
31The earliest instances of sherd-lined features at Tell Zeidan date to the lower Ubaid levels in areas B and D and both are associated with ovens, while a third Ubaid example comes from inside of a hearth.19 In a few cases, postholes have sherd-and-gravel linings. Building S.D1 in area D (fig. 9) includes a sherd-lined pit (Locus 17) in the floor of an enclosed, presumably domestic space dating to the latest Ubaid level, and this is the most common context in which these pits are attested from that point forward.
32Though they appear in different areas of the site, the greatest concentration of these pits at LC1 Tell Zeidan is in area E (fig. 11) where there are five, one of which is lined with dried clay. All of them are either set into built floors or otherwise inside of a room, though some occur in what were possibly open-air spaces (courtyards or alleyways). The construction is similar in each case, with small, abraded, usually coarse-ware potsherds arranged along the side of shallow, funnel-shaped pits no more than a half-metre across. Their contents are also fairly uniform: a silty, earthen matrix devoid of artefacts. The silty nature of the fill hints at a water-related function, but there is no further evidence for such an interpretation. The most information that can come at present from sherd-lining of features is that the general construction technique persisted over hundreds of years but with a change from oven contexts in the Ubaid period to pit contexts in the LC1 period.
33“Cultural features” here are elements of architectural space that signify social, cultural, or decorative activities not directly related to subsistence or structural functions. The most evident type from LC1 Tell Zeidan is a vessel containing an animal mandible placed beneath the wall of a building. This occurs twice, both times in domestic contexts. One comes from the Northeast Mound in S.B3 (fig. 10, Locus 61), level LC1a, and one from the Northwest Mound in level LC1b. Typical of Mesopotamian foundation deposits (Ellis 1968: 31), it appears that these could have been votive offerings made during construction as part of the building rites. Such votive offerings can occur in a variety of contexts (van Buren 1952: 79), but are often associated with religious activity (see e.g., Mallowan 1946: 124, fig. 2; Ellis 1968: 10; Munchaev 1997: 71; Oates et al. 2001: 90, 389). For the LC1 period, they may signify that religious activity shifted away from centralised spaces such as temples into standard dwellings (see, e.g., Gurdil 2010).
34Another type of “cultural feature” found at Tell Zeidan is the intramural burial. The only LC1 example comes from area E and is an adult grave placed underneath the floor inside of a room in house S.E5 (fig. 11, Locus 95). This burial contains a single fully flexed skeleton of an adult human and contains no grave goods. Its stratigraphy and context are certain, though, and so it is informative as to the changing mortuary practices of the mid-5th millennium BCE. In the Ubaid period, supra-local cemeteries contained burials of adults, sometimes in supposedly kin-based pairs or triplets, and were differentiated in their grave goods and burial conditions only by age and possibly gender (Wright and Pollock 1987; Akkermans 1989). There is a relative uniformity and emphasis on egalitarianism or community identification seen in these cemetery graves (Stein 2010b: 37; 1994). Following the Ubaid period, this changed and burial practices began to feature wealth-based ranking (e.g., see Pollock 1989; Marro et al. 2011; for an example from southern Mesopotamia, see Woolley 1955). The flexed burial in House S.E5 indicates that by the LC1 period, an adult could be buried with no grave goods at all, underneath a floor in (presumably) their own house. This arrangement is entirely lacking in community affiliation, and represents a significant ideological shift from the mortuary conventions of the Ubaid period.
35Finally, excavations in area E produced faint traces of red paint on two white-plastered walls in Room S.E3.2 (fig. 11, Walls 85 and 86). Ubaid (e.g., Tobler 1950: 33) and LC2 parallels (e.g., Hauptmann 1976: pl. 42.3) suggest that painted walls appear in spaces with ritual or ceremonial functions. In the LC1 period, however, wall painting occurs in domestic spaces as seen at Deǧirmentepe (Gurdil 2010: 366) and Arslantepe (Balossi-Restelli 2008: 24). Gurdil interprets the Deǧirmentepe paintings as “direct indications of ritual behaviour”, and views the dwellings generally as “shrine houses” (Gurdil 2010: 370-372). Although the level of exposure and the distribution of painted rooms at Deǧirmentepe both differ from Tell Zeidan, it raises the possibility of household-based ritual activity in area E.
36The most apparent outcome of this intrasite study is that a fair degree of intrasite variability characterises the LC1 period at Tell Zeidan. Looking at relative spatial distributions of craft and subsistence debris, we can see that all three mounds have similar levels of craft production, food preparation, and food consumption (fig. 12). Thus, self-sufficiency at this level of locality is reasonably attested in the data, and the site as a whole displays a low level of economic interdependence based on evidence for mixed-craft production and subsistence activities occurring in most of the domestic contexts excavated. Contrastingly, Ubaid levels at Tell Zeidan yielded evidence for a specialised obsidian and stone-goods workshop (Stein 2011) and circumscribed concentrations of entire craft-production artefact types such as awls. Furthermore, concentrated evidence for economic and administrative specialisation, attested in Ubaid and Ubaid-LC1 strata, respectively, does not appear in subsequent levels. This is consistent with the interpretation that in the LC1 period, structurally similar communities began to organise themselves economically at a localised or household level rather than centrally.
Fig. 12 – Percentage of objects in site area by functional category.
37The site does not appear to have a high degree of social cohesion in the LC1 period, as shown by the variation in dietary preferences, architectural forms, textile spinning practices, and even ceramic types. In some cases, individual mounds or households demonstrate their own cultural preferences. The preliminary ceramics analysis indicates that in terms of pottery usage, the three mounds were functionally similar in the LC1 period, though with particularised approaches to production. Regarding the LC1 animal economy, based on the fauna collected, individual households or areas seem to have had different dietary preferences or access. The pattern in the Ubaid period demonstrates a slightly different kind of faunal variability, as area D features an abnormally high concentration of goat bones while area B had a relatively high abundance of wild fauna, possibly reflecting functional variation at that time. The Ubaid diet in general, however, was fairly consistent across the site (Grossman and Hinman 2013: 210).
38Intrasite architectural assessment indicates that LC1 buildings at Tell Zeidan can be planned, rectilinear spaces, or unplanned and agglutinant; they can be either ordinally or cardinally oriented; and they can employ a variety of construction methods, including ad hoc changes between strata. Building S.C4 features the only potential evidence for monumentality from mid-5th millennium BCE Tell Zeidan along with the highest concentration of administrative artefacts, raising the possibility that it was a public building. Whether or not this supposition is correct, S.C4 predates the LC1 occupation, and so may reflect the more centralised political economy and economic interdependence of the Ubaid occupation at Tell Zeidan rather than the decentralised goods procurement and processing evident in the LC1 levels. It was, however, during the time of the Ubaid-LC1 level that architectural variation started to become pronounced, and so the rise of variability is apparent alongside the final attestation of centralised administration and monumentality.
39The data from the latest Ubaid and earliest Late Chalcolithic levels at Tell Zeidan are too limited to be conclusively demonstrative, but they are suggestive of a societal progression toward intrasite sociocultural variability. Intrasite variation is already evident in Ubaid strata at Tell Zeidan, but its nature appears to change gradually from specialised and integrated to unspecialised and localised. The archaeological availability of LC1 and Ubaid remains across the site mainly results from Tell Zeidan’s reduced occupation size in later periods. This eliminates any potential for understanding the reintegration or urbanisation processes of the LC2 period, either locally or regionally within the Balikh Valley (Wilkinson 1998; Trentin 2010). However, this snapshot from LC1 Tell Zeidan reflects a wider phenomenon of variability and localisation writ large across Greater Mesopotamia (Baldi 2016), and so developing a model to help explain the structural shift benefits from a broader, trans-regional perspective.
40Where some theories of change address a transition between periods, LC1 Northern Mesopotamia forces us to approach it differently because it demonstrates uneven change at an often-gradual pace. The argument herein views the gradual decline of one set of material culture trends and the later rise of a new societal structure as related phenomena. The question itself is difficult to answer with prehistoric data: why would inhabitants abandon a known system, how do aggrandisers dismantle the prevailing ideologies of that system, and then how does that process relate to inhabitants participating in a new, more hierarchical system? Lesure and Blake (2002) use the term “ideological shift”, while Flannery and Marcus (2012) refer to a “manipulation of the social logic”. Stanish and Haley (2004) offer that society members are typically “conditional co-operators”, and if elites can demonstrate to them that they benefit from participating in cooperative, surplus-oriented economic production, then those elites have effectively manipulated the “social logic”. Next-level wealth-accumulation would then become socially permissible, which can facilitate more formal, embedded social stratification.
41In northern Mesopotamia the process appears to have been slow, with traits of the Ubaid system persisting for almost a millennium. The discussion here focuses on pottery, as it is the most apparent instance of 5th millennium BCE material culture exhibiting both stylistic continuity and instrumental innovation, but it could be applied to other material categories as well. Pottery is also a typically conservative medium with high resilience to change (Rice 1984; Gibson and McMahon 1997), and can represent shared, embedded social processes, ideologies of identity, or widespread “traditions” that constrain behaviour and promote “continuities of practice and form” (Carter and Philip 2012). The gradual but definite decline of the dark-on-light painted style attested across the Ubaid interaction sphere, then, would reflect gradual change in a practical or structural sense. Put another way, social change likely happened concomitantly with material change.
42The introduction of new pottery techniques would have clearly impacted the ceramic tradition, but how can we view this in terms of human agency? Recent research in the field of cognitive science addresses “social learning” and the relationship between changes in instrumental behaviour (such as new pottery production techniques) and changes in conventional (or sociocultural) behaviour. The model of sociopolitical change proposed here describes how disruptive technological innovations of the Ubaid period, particularly evident in the ceramics, precipitated a time of widespread cultural variability and lower levels of sociocultural cohesion in the LC1 period. This would have led to a wider variety of economic and sociopolitical strategies available, attestable in the archaeological record (Baldi 2012; Fisher 2017).
- 20 Cf. Roux, who views craft manufacturing motor skills as “embedded” and without room for innovation (...)
43This theoretical approach builds on the behavioural archaeology of Earle (1991), the identification of variability and heterarchy proposed by Stein (1998, 2010b), and the role of technological innovation emphasised by archaeological pioneers such as Childe (1942, 1972; cf. Roux et al. 2017). However, it is rooted in empirical, present-day experimental studies in cognitive science. These identify a strong distinction between two main types of innovation: “conventional” (sociocultural) and “instrumental” (relating to tools), demonstrating that “instrumental innovation” is more readily achievable for human children than innovation of social conventions (Nielsen et al. 2014; Legare and Nielsen 2015; Tomasello 2016).20 Thus, sociocultural continuity in human society is generally quite strong, based on the human predilection for “conventional imitation” that fosters in-group cohesion. However, humans can more easily reimagine and reformulate non-conventional, instrumental behaviours as evidenced through tool innovation at young ages. Those instrumental innovations, if adopted, then provide an opportunity for behavioural variability, which allows groups and individuals to promote new, competing sociocultural conventions, especially when mapped onto the new instrumental behaviours.
- 21 Wen et al. define ‘ritual behaviour’ as “conventional, causally opaque procedures, are uninterpreta (...)
44Because the process of social learning is largely predicated on imitative repetition, instrumental behaviours, as opposed to conventional behaviours, afford children “imitation flexibility” (Legare et al. 2015; Clegg 2016), and “novel behavioural variation [occurs] through innovation” (Legare and Nielsen 2015: 7). Furthermore, human children can be conditioned not to innovate (“functional fixedness”) if they know that an instrumental function is related to a cultural convention (Legare and Nielsen 2015: 6). Members of a society, then, can govern cultural norms by encoding conventional cues in instrumental behaviour. Social cohesion is a product of conventional imitation and it can increase through ritual behaviour,21 which reduces intra-group conflict (Legare and Nielsen 2015: 4). As suggested above, in the LC1 period at some sites, ritual behaviour may have become localised at the household level, while social cohesion was simultaneously in decline, creating the precipitating conditions for imitation flexibility and, more broadly, ideological change.
45Since “robust” imitation of innovative instrumental behaviour is possible across time (Chappell et al. 2013; Beck et al. 2014), once the new behaviour is imitated and transferred enough times (i.e., it has recurrent utility) it becomes part of a society’s instrumental repertoire. Additionally, any non-conventional or purely instrumental behaviour can become conventional by adding conventional language, especially through ritualisation (Wen et al. 2016). This allows society members to potentially govern not only sociocultural stability, but also change, particularly if that change is projected onto technological change.
46The implications of these studies in cognitive science are that sociocultural changes are likely to have been precipitated by variability through instrumental innovations. Through the continuous social learning process of cultural convention, a society member is normally disinclined to operate outside of those conventions. An “aggrandiser”, which is Clark and Blake’s ever-present, and occasionally successful, self-promoting “political entrepreneur” (1994: 17), would be precluded from successfully convincing others to participate in altering that social order to meet their own ends. Adoption of innovative instrumental behaviours, however, can disrupt the chain of conventional imitation. This results in heightened opportunity for behavioural variability, in terms of both socioeconomic strategy and organisational modality, as well as, and often expressed through, material culture. Aggrandisers can then co-opt the new technology and create associations between its instrumental techniques and new social conventions. Centralised administration systems began to grow and crystalise in the LC2 period. This is when administrative hallmarks such as monumental buildings started to co-occur with the outcomes of earlier instrumental innovations such as rapidly produced bowls (Baldi 2015) or figurative glyptic iconography featuring ritually imbued leaders (Rothman 2002).
47Thus, the model offered here can be summarised as:
48stable social structure a transferable instrumental innovations social reflexivity and augmented opportunity sociocultural and sociopolitical (or “ideological”) innovations mapped onto new instrumental behaviours and new social structure.
49Ultimately, this model establishes the precipitating conditions for structural transformation, rather than arguing for one specific technological cause of change, since sociotechnical systems produce “power and meaning as well as goods” (Pfaffenberger 1992: 502). The role of variability in explanations of both societal development (e.g., Stein 1998; Swenson 2007) and the material trajectory of 5th millennium BCE Mesopotamia (e.g., Adams 1971, 1972; Stein 2010b) underlies this approach. It is important here because it is a productive way to frame the various “circuits of interaction” or “cultural provinces” that define the LC1 period (Frangipane 1993; Mazzoni 1999, 2000; Hellwing 2000; Rova 2000; Marro 2012), rather than attempting to rely on limited material commonalities across a varied cultural landscape. Variability is also one of the key outcomes of instrumental innovation in the ubiquitous process of social learning (Legare and Nielsen 2015: 9). It helps underpin the theory that identifying levels of what is called “social reflexivity”, or the extent to which any given individual can exert influence over the nature of their society, could play a role in understanding sociopolitical development. Finally, it informs the method used here of categorising elements of local variability as markers of social reflexivity, which can help determine a society’s susceptibility to structural change.
50There are, of course, other ways to explain how aggrandisers are able in some societies to legitimate the structural changes that they pursue. A good example is Stanish and Haley’s (2004) argument that in order to achieve a state of cooperative surplus productivity, “smart individuals” need to perceive a benefit to themselves in order to participate in the new system. Persuasion toward this end rests largely on neutralisation of the threats of exploitation (such as “free-riding”). Early Central Andean public architecture, they offer, had high visibility and would have been used in conjunction with ritual to amplify the impact of punitive action against deviant members’ reputations, thus convincing adaptive individuals that the system will function effectively and fairly (Stanish and Haley 2004: 59-60). In this model of sociopolitical development, the quantity and magnitude of public buildings correspond to degree of complexity in that the buildings must accommodate the decision-making members of the faction or group (Stanish and Haley 2004: 62-63). The argument, however, rests on the concept of “cost-benefit calculation” among society members in order to avoid the circular logic that attempting to increase monumentality already requires new modes of labour organisation. The “conditional cooperator” theory, like the “instrumental innovation” model proposed above, addresses the connection between materially attested changes in ancient society and sociopolitical reorganisation. However, like many archaeological theories, the former rests on a major assumption, in this case about the human mindset. In contrast, the model presented here invokes the human mindset but relies on practical studies and scientifically-demonstrated conceptions of it.
- 22 Tell Zeidan demonstrates early evidence for the flint-scraped method of rapid bowl production in th (...)
51The proposed model views instrumental innovations of the early 5th millennium BCE, such as rapid-bowl manufacture,22 as conduits for behavioural variability and the multiplicity of socioeconomic strategies generally evident in the LC1 period. An atmosphere of ‘augmented opportunity’ enabled aggrandising actors to pursue formerly unsanctioned strategies such as individualistic and conspicuous wealth accumulation and, possibly, development of technological domains. Such domains would have provided a means of validating, through conventionalisation or ritualisation, ideological innovations such as conspicuous consumption of wealth goods, conscription of commodities, or overtly hierarchical systems of leadership and domination. Put another way, variability engendered a broad landscape of highly varied segments and strategies in the LC1 period that were eventually, in the LC2 period, reconsolidated into a complex northern Mesopotamian sociopolitical framework.
52Overall, we know that inhabitants of each area at Tell Zeidan conducted their own craft and tool manufacture, food procurement, and ritual activity. Without the evidence for centralised administration or specialised workshops present in Ubaid layers at Tell Zeidan, this potentially implies lower levels of site-wide economic interdependence. And yet, the mounds all show, to some extent, variable cultural preferences in categories such as diet, ceramics, and architecture, altogether suggesting lower levels of social cohesion. These patterns may reflect a transformation in spatial usage and material culture from Ubaid levels at Tell Zeidan, which evidence monumental architecture and administration, tripartite buildings, specialised workshops, and functional differentiation within the material assemblages.
53In order to understand the end of the Ubaid period as a time of active transformation rather than stagnation or regression, we can apply Stein’s concept that variation can be the condition that leads to wider contact initially (2010b). Variability is a characterising aspect of the LC1 period, and the theory proposed here states that variability itself is an important precipitating condition for further structural development. In a sense, the otherwise conservative medium of pottery supported a culturally-embedded sociopolitical framework in the Ubaid period, and introduction and adoption of new instrumental techniques allowed producers to disrupt that link and foster greater material-culture variability. This then enabled new modes of social behaviour, economic opportunity, and, ultimately, sociopolitical organisation. These new modalities manifested during the LC2 period, when sites in the Balikh Valley like Tell Zeidan contracted, and sites in the Jezirah such as Tell Brak expanded into proto-urban centres. And so, this kind of theoretical model helps us go beyond empirical limitations when evaluating the nature of the transition from the Ubaid period and the boundaries that have defined it.