Home Page
Archaeological Excavations in Romania, 1983 - 2012.
Preliminary Archaeological Reports

Archaeological Excavation Report

Pecica | County: Arad | Site: Şanţul Mare | Excavation Year: 2008

Excavation Year   2008
Bronze Age
Bronze Age
Site Category
Site Types
Fortified settlement
Map it   Find it on the Romanian map
County / District  Arad
Locality   Pecica
Commune   or. Pecica
Site  Şanţul Mare
Site Sector
Site name   
Persons involved and Institutions
Last nameFirst nameroleInstitution
Barker Alexander Michigan University, USA
Draşovean Florin Muzeul Banatului, Timişoara
O’Shea John Michigan University, USA
Pascu Hurezan George Site director Complexul Muzeal Arad
Sherwood Sarah Michigan University, USA
Szentmiklosi Alexandru Muzeul Banatului, Timişoara
National Arch. Record Site Code 11593.01
Report The great settlement of Pecica “Şanţul Mare” is among the most important archaeological sites in the European Bronze Age. The site occupies a strategic location astride the river Mureş, between the ore producing region of the Western Carpathian Mountains and the metal using societies of the Carpathian Basin and beyond. Similarly, its deeply layered Bronze Age deposits have served as a chronological standard for the entire Bronze Age in Eastern Europe. In November of 2003, the Muzeul Banatului Timişoara, the Muzeul Judeţean Arad, and the Museum of Anthropology, the University of Michigan (USA) entered into a contract to facilitate new archaeological excavation at the site. Following a brief planning visit in 2004, a major, multi-year field investigation was begun in 2005 with funding from the National Science Foundation (USA). The principals of the excavation cooperative are Dr. Peter Hügel (Arad), Dr. Florin Draşovean (Timişoara) and Dr. John O’Shea (University of Michigan). Mr. Pascu Hurezan (Arad), Dr. Alexandru Szentmiklosi (Timişoara), and Dr. Alexander Barker (University of Missouri) rounded out the excavation team and were involved in the day to day management of the excavations. Dr. Sarah Sherwood (University of the South) oversaw site geoarchaeology and micromorphology, Ms. Amy Nicodemus continued to oversee the recovery and analysis of faunal materials, and Dr. Laura Motta (University of Michigan) consulted on the recovery and analysis of plant remains from the site.
The excavation campaign in 2008 saw the continuation of layer by layer excavation at the site. The specific goals of the 2008 season were 1) to continue the layer by layer excavation of a 10 by 10 meter block on the site to expose the core Late and Middle Bronze Age portions of the site sequence; 2) to investigate the large enigmatic D layer platform that was exposed during the 2007 season; and 3) to identify additional Bronze Age house structures. A limited program of off-tell testing was also initiated with the excavation of two test units to the west of the great ditch. In addition to these specific goals, and in accordance with the contract, work in 2007 continued the program of offsite geological testing, and the development of a regional GIS site database. Field operations began on 18 June and were completed on 5 August.
Work at the site in 2008 began with the removal of backfilled sediment and the reestablishment of the 10 x 10 meter grid that was begun in 2006. The excavation zone covers an area of 12 x 12 meters, which includes the 10 x 10 meter excavation block and an unexcavated belt around the excavation area to protect the site profiles. Following the procedure established in 2006, the excavation block was divided into a series of 2 x 2 meter squares for the purposes of data recording and flotation sampling. Given the depth of the excavation, it was necessary to create entry points from the older stratigraphic trench across wooden walkways. Walkways were also used on the site to minimize trampling damage to intact excavation surfaces. Excavation continued on a layer by layer basis, working systematically across the site surface. Excavation was conducted with small hand tools, with the total volume of deposit removed recorded. Within each 2 x 2 meter square, 10 percent of the soil removed was dry sieved through screens with a mesh size of .65 cm. Additionally, two 10 liter samples from each square level were collected for flotation. Flotation samples were processed in the field laboratory using a Flote-tech automatic flotation system. All significant finds were mapped in three dimensions with the Sokkia total station, as were the starting and ending elevations of each unit, and the perimeters, tops, and bases of all features. Throughout the period of excavation, daily three dimensional maps of the excavation were constructed, as were layer photo mosaics. A series of ‘micro-morphology’ samples were also collected over the course of the block excavation. The samples, when analyzed, will provide important information on the creation and composition of the site’s micro-stratigraphy.
A suite of seventy four radiocarbon samples were collected for chronometric dating using AMS techniques. All radiocarbon samples were collected with a clean trowel or other hand tool into folded foil packets, labelled with sample number, material being dated, a reference number identifying the corresponding three-dimensional values recorded by the Sokkia total station, and the sample’s coordinates. Almost all of the samples collected were of wood charcoal, and samples were checked in the field for size and to confirm that intact ring structure was visible in order to distinguish datable charcoal from zones of blackened or reduced soil.
Materials recovered during excavation and screening were bagged separately by square and were returned to the field laboratory each evening. Materials were washed and sorted, with faunal remains being separated from artifacts. Artifacts were counted and weighed in coarse categories (ceramics, stone, metal, slag, shell, daub) with diagnostic artifacts being removed for further description, photography and analysis. Preliminary analysis of the animal bone was undertaken on a systematic sample of the excavations units during the course of the 2008 season (see below). A similar analysis of diagnostic ceramics was also initiated, but was not completed due to the unexpectedly large quantity of recovered ceramics. Due to a miscommunication among project staff, selected ceramics were originally to have been loaned between the cooperating institutions for further study in the off-season; these plans were subsequently changed due to concerns about the ability to complete all documentation required by patrimony laws for typical ceramics in the available time. Diagnostic ceramics were set aside for subsequent analysis. All other materials were packaged and transferred to the Arad Museum for future analysis. Flotation samples were separated into light and heavy fractions (floating and nonfloating) during the flotation process. The light fraction samples were transported to the University of Michigan for detailed analysis of micro-faunal and plant remains.
In addition to the layer by layer excavation of the site units, visible features, primarily deep pits, were excavated as units using natural layers. In order to maintain tighter association and control over internal features of houses, the logical unit of structure was added as a superordinate form of feature, each comprising multiple internal features and superposed occupational surfaces. This approach facilitated mapping of changing house forms and activity areas while keeping materials associated with each together for analytical purposes. Features internal to houses were either excavated at the level in which they were encountered, or mapped and recorded at the level encountered. Structures excavated during the 2008 season tended to have multiple superimposed occupation layers, each of which may have been associated with posts or post pits. For larger pits, the feature was bisected so that a profile of the feature fill would be visible after half of the feature had been excavated. Unless the pit was positively determined to be a modern looter’s pit, the sample of soil from each feature was 100% screened, excluding only that sediment collected for flotation sampling. Each feature was mapped and photographed during excavation, and the top, base, perimeter, and internal levels of each were mapped using the total station.
At the end of the 2008 field season, the excavation surface and all profiles were fully covered with plastic sheeting. Due to the depth of the excavation block, the base of the excavation was fully covered with soil, soil was bermed along the profile walls, and an anchoring layer of soil was placed on top of the plastic sheets lining the profiles to hold them in place. This left the base of the excavation unit, the tops of the profiles, and parts of the profiles covered with backfill, with the remaining portions covered with anchored plastic. All collected materials were inventoried and packed for storage at the CM Arad.
The 2008 field season was generally successful in terms of both the accomplishment of the planned excavation goals and in terms of new discoveries which will guide and inform future excavation. As analysis of all materials is ongoing these results are of a preliminary nature, and subject to change based on the outcome of more detailed studies.
In stratigraphic terms, the excavations removed the lower rubble and fill associated with site layers C3, C4, and C5 in the western half of the site, and extended downward to expose floors associated with site layer D1. In the eastern half of the site block, excavation was focused on the massive D0 construction, which was excavated to the top of the underlying fill layer, designated E1 and E2. In addition to the normal layer excavations, the complexly stratified remains of two architectural features, termed Structures 1 and 2, were excavated in their entirety. Two discrete Middle Bronze Age houses or house complexes, each characterized by multiple layers and rebuilding episodes, were excavated during the 2008 season.
Economic, geoarchaeological and chronometric samples were systematically collected from multiple levels of both houses and house complexes, and are currently undergoing or awaiting analysis. While both houses were interrupted either by subsequent cultural features or the limits of the current excavation area, it is possible that neither is as large as previously defined Periam-Pecica culture houses (e.g., Giric 1987; Horvath 1982; O’Shea 1996), although they are similar in size to house structures previously reported at the site by Crişan (1978). Whether this smaller size represents functional differences, or reflects spatial constraints atop the limited area of the tell summit is not yet clear.
Structure 1 had been identified in Trench 1, excavated in the 2005 season as part of the complex of strata in layer C, and was located on the western edge of the 10 x 10 m excavation block. It was identified as the corner of a major structure, or possibly an element of a thermal feature or kiln, in previous annual reports (see O’Shea 2007). Structure 2 was located along the southern margin of the excavation block, and was associated with site Layer D. Scatters of architectural rubble were associated with the rebuilding or abandonment of both structures, and covered in the SW portion of the excavation block. Both structures were also associated with thermal features which contained discrete and bounded concentrations of partially reconstructable crushed ceramic vessels and loose sherds. The excavators of Structure 2 recorded what may have been a transverse internal partition wall perpendicular to the exterior wall, with distinct stratigraphic sequences recorded on either side of the putative partition.
In light of the discoveries of 2008, it now appears that fragmentary portions of a third, undesignated, structure may have been encountered in the NE corner of the excavation block as part of the 2007 field campaign. Thermal feature similar in form to those documented in structures 1 and 2 were recorded from this portion of the excavation block, as was an extensive capping by a white unfired rubble. While portions of occupational surfaces were recorded and mapped in previous seasons, the characteristic signature of structures and the associations of various structural elements were not recognized until the 2008 season, allowing retrospective identification of this third house or house complex.
In addition to house structures, a massive surface (termed Layer D0), the top of which was exposed in 2007, was excavated in its entirety (figures 5 and 6). The deposit varied in thickness from south to north, with the thickest deposits being 80 cm at the north edge of the excavation block. Stratigraphically, the construction definitely post dates Structure 2, and predates the undesignated third structure. The surface may also pre-date Structure 1, although determining the stratigraphic relationship is made difficult by the lack of continuously distributed deposits linking the two features.
The layer appears to have been created as a single event or rapid series of events, and showed evidence for burning or sustained exposure to high temperature levels. Ceramics were oxidized and in some instances refired or partially vitrified, and bones were calcined, burned or in associated with well-developed and visible vivianite staining. Micromorphological samples collected in each of the last three seasons suggests that there is significant cementing of the deposit due to carbonate precipitates in both the C and D strata units and, in some stratigraphic units like D0, alteration due to sustained heating to relatively high temperatures, and cultural preparation of specific occupational surfaces. The feature also extends well beyond the limits of the current excavation area. A series of auger test cores located around the perimeter of the excavation block demonstrated that the feature continued outward from the excavated area, particularly to the east and the south.
The construction appears to have performed some manner of site levelling or renewal function, although it is not yet clear whether it was designed simply as a surface for industrial or community activity, or whether it might represent the base of a large structure. If it does represent a structure, the construction is unlike the other houses exposed at Pecica, or any structure previously reported for the Periam-Pecica/Maros Culture. While geomorphological analyses are currently underway, it appears that the D0 layer represents burned and redeposited anthropogenic sediments forming a clear lithostratigraphic unit with abrupt and well-defined boundaries, in some areas 70cm or more in thickness. As this represented an unmistakable “natural” layer whose contents may represent chronologically earlier site deposits, analysis of layer D0 and its contents may have significant implications for understanding the strikingly long apparent use life of certain ceramic forms in existing seriations based, in part, on materials from Pecica “Şanţul Mare”.
In addition to site excavation, the 2008 season saw the excavation of two 2x2 meter test units outside of the main fortification ditch. These units were opened to document the extent and character of Bronze Age occupations outside the fortification along the area to the south of the tell. Coring samples collected in previous seasons indicated that there were relatively deep cultural deposits in these areas. These excavation units revealed no identifiable Bronze Age occupations, but did reveal the presence of a series of large and well-defined mediaeval pits in the area.
At total of 74 new samples for radiocarbon dating using the AMS method were collected during the 2008 season, but results are not yet available. Material evidence of Middle Bronze Age occupations included a wide range of local and nonlocal ceramics, lithic and obsidian flakes and artifacts, a diverse assemblage of animal bone, and groundstone tools including a portion of a ceremonial axe casting mold. The recovery of large quantities of casting debris, slag, and raw minerals underscores the significance of metallurgy and bronze production during the second half of the site’s Bronze Age occupation.
Block’ micro-morphology samples and 56 ‘bulk’ samples from the main excavation area. These samples, when analyzed, will provide important information on the creation and composition of the site’s micro-stratigraphy. Continued excavation at the site has revealed increasingly complex patterns of site formation and deposition, and these analyses will be crucial in resolving important issues regarding the character and significance of individual stratigraphic units, and their relation to one another. An additional 32 auger core samples were also collected. Processing of flotation, micromorphology, and metallurgical slag samples is currently underway. A preliminary analysis of the recovered faunal remains, conducted by Amy Nicodemus during the 2008 field season, is included here.
The 2008 field season provides a solid foundation for continued excavation into the deeper portions of the tell sequence. The greater depths of the deposit, and the high density of cultural remains, pose a continuing challenge to our excavation and data recording strategies which must necessarily be adapted to changing excavation circumstances. While understanding the role and character of the extensive D0 deposits will require further analysis and study, stratigraphically sequential houses within both the C and D layers reveal certain signatures for both house construction and the scattering of their remnants following abandonment. These signatures will aid in the future recognition of houses, as well as helping frame questions regarding relationships between discrete structures within the same stratigraphic unit.
Key questions to be addressed during future seasons include understanding the character of the E layers, recognizable by a dense concentration of wood charcoal; the longer-term spatial relationships between houses or house complexes; and the varying occurrences of slag, metal by-products., metals and thermal features in different areas of the excavation block. A particularly important element in these analyses will be the effort to accurately link artifacts, fauna, and other material residues with particular house structures, as a first step towards investigating social and economic differences between tell households.
We continue to benefit from an extremely productive collaborative relationship with the Muzeul Judeţean Arad and the Muzeul Banatului Timişoara, and from the skills and experience of our senior collaborators, Dr Florin Draşovean and Dr. Peter Hügel. The daily presence on site of Dr. Szentmiklosi and Mr. Hurezan has been a great factor in the success of the work. Finally, the friendly and cooperative atmosphere that has developed among the Romanian and American student crews, and with the people of the village of Semlac, is particularly appreciated and valued.

Pecica 2008 Summary Faunal Report
Amy Nicodemus (Museum of Anthropology University of Michigan)
The 2008 Pecica excavations produced a very large and well preserved faunal assemblage. Representative samples from each layer (C4-D0), two structures, and four thermal features were analyzed in the field, totalling 3381 bone and shell fragments (Table 1). Overall, the fauna from all contexts was similar in composition (Table 2). Mammals comprise the bulk of the material, averaging 87.3%. Mollusks are less common (12.4%), while fish (0.3%) and birds (0.1%) are rare. There are no reptiles present in the 2008 sample. Domestic animals far outnumber wild mammals (83.7% versus 16.3%). Sheep and goats are the most frequent livestock (34.3%), followed by pig (27.4%), cattle (22.0%), and horse (14.8%). Dogs are poorly represented, forming only 1.4% of the domestic fauna. Despite the general homogeneity between contexts, there are a few significant differences present. Mollusks are over-represented in the thermal features1 and Structure 1 has relatively few ovicaprids and a very high proportion of horse bones.2 The 2008 animal remains differ sharply from later Bronze Age deposits (O’Shea et al. 2005; 2006; O’Shea 2007). In layers C1-C3, non-mammalian fauna are much more abundant,3 particularly fish, and there are significantly more ovicaprids (except C1) and far fewer horses than in earlier layers.4 In summary, animal husbandry formed the core of the Pecica animal economy throughout the Bronze Age periods analyzed to date. However there was a shift in emphasis within livestock management away from larger animals, particularly horses, towards ovicaprids in later periods.
Concurrently, fishing became increasingly important though time. Large game hunting, focused on red deer, was an important dietary supplement in all periods and additionally provided raw materials for tool and ornament manufacture. Fowling and small animal trapping was only a minor activity.
1 χ²=20.802, df=5, p=.001.
2 χ²=44.772, df=15, p<.001.
3 χ²=40.262, df=7, p<.001.
4 χ²=116.271, df=21, p<.001.
Abstract other lang.
Abstract   The great settlement of Pecica “Şanţul Mare” is among the most important archaeological sites in the European Bronze Age. The site occupies a strategic location astride the river Mureş, between the ore producing region of the Western Carpathian Mountains and the metal using societies of the Carpathian Basin and beyond. Similarly, its deeply layered Bronze Age deposits have served as a chronological standard for the entire Bronze Age in Eastern Europe. The excavation campaign in 2008 saw the continuation of layer by layer excavation at the site. The specific goals of the 2008 season were 1) to continue the layer by layer excavation of a 10 by 10 meter block on the site to expose the core Late and Middle Bronze Age portions of the site sequence; 2) to investigate the large enigmatic D layer platform that was exposed during the 2007 season; and 3) to identify additional Bronze Age house structures.
Bibliographic notes
Source   Cronica cercetărilor arheologice din România
Editor   CIMEC
Language   ENG

Copyright: the authors of the reports and the National Heritage Institute, CIMEC, 2018.
Coordinating: Bogdan Şandric. Documentary - analysts: Iuliana Damian, Oana Borlean, Adriana Vîlcu. Consultant: Irina Oberländer Târnoveanu. ASP, HTML design: Cosmin Miu