Stone Village Project 2005-2006

Contents

Acknowledgements
Aims and methodology
Results
Discussion
Artefacts
Ceramics
Environmental remains
Faunal remains

Acknowledgements

The Stone Village Project is an independently funded subsidiary project of the EES Expedition, and the support of the EES and Expedition Director Barry Kemp is acknowledged with much gratitude. The SCA facilitated the smooth running of the 2005–2006 season; particular thanks are due to inspector Mr Fathy Awad Riyad. The 2005–2006 field season was funded by The Wainwright Near Eastern Archaeology Fund, Oxford University; The Thomas Mulvey Egyptological Fund and The Chadwick Fund, Cambridge University; and The Seven Pillars of Wisdom Trust, London. Several members of the Amarna Expedition provided much appreciated practical support and advice, particularly Pam Rose and Helen Fenwick.

Aims and methodology

Anna Stevens
The season ran from 8/12/2005–2/1/2006. It was intended as a preliminary investigation of the layout of the core site, construction materials and methods, the depth and nature of archaeological deposits, the relationship between the surface and sub-surface features, and the nature and extent of any disturbance to the site. A combination of total-station survey, detailed hand planning and small-scale excavation was employed.

A survey grid was established across the core site, using a Leica TCR307 total station, and general-reference site plan generated, showing the limit of the stone scatter, the major linear arrangements of stone, and the small mounds dotted around most of the immediate perimeter of the site, some of which are certainly areas of ancient dump. Hand planning of surface features then commenced, concentrating on the south-west corner of the core site. Over a period of approximately one week, an area some 16 x 15 metres was planned.

Aerial photograph of the Stone Village taken in 2002, showing the limit of surface planning in the 2005–2006 season (larger rectangle) and the position of the test trench (smaller rectangle). North is to the top of the image.

Midway through the season, a 7 x 3 metre trench (Trench 1) was opened in the north-central quadrant of the core site, where the depth of deposit appeared greatest. It was positioned to encompass samples of the three typical surface deposits: a prominent east-west linear stone scatter across its northern margin, a ridge of gravely sand through the western margin, approximately linear, and slightly sunken loose sand concentrated in the southeast quadrant. Aerial views of the site convey a sense of ‘containment’, raising the possibility that it was surrounded by a retaining wall. The stone concentration formed part of the northern limit of the stone scatter that delineated the core site, and had potential as such a perimeter wall. One question of interest was whether the stone represented the lower courses of a wall that had slumped onto itself, or had fallen from the upper part of an architectural element, retaining its original orientation. The overall linear alignment of the gravely ridge also lent it potential as a buried architectural element. Excavation was conducted over eight days, and at the end of the work the trench was backfilled using the excavation spoil.

Caption: Trench 1 prior to excavation, facing south. The red dots mark the corners of the trench. Note the prominent east-west stone ridge across the northern margin of the trench.

Results

The survey work established the limits of the core site, as far as can be determined from the surface remains. It measures approximately 67 metres north-south and 80 metres east-west, covering an area of around 5360 sq. m, including the mounded areas around the perimeter. The precise limits of the structural remains are difficult to determine at present. The surface of the north-west quadrant, in particular, has less stone than other areas, and less that appears to lie close to its original position, so the edge of the site may step in here. As yet, there is no evidence from survey or excavation (see below) of a perimeter wall.

Stone alignments can be recognised with greatest confidence across the north-east quadrant and southern margin of the site. A particularly prominent north-south alignment, containing a possible entrance way, is visible in the south-west quadrant. The stones here appear to be at or near foundation level. A possible corner can be discerned in the north-east quadrant, and several alignments across this area seem to relate to one another. This is an area of the site that has noticeably less gravely sand both across it, and around its perimeter. If these gravely sand deposits are spoil heaps from illicit digging, as excavation to date suggests (see below), it may be that this area has largely escaped such disturbance.

Caption: Prominent stone alignment in the south-west corner.

Two relatively prominent east-west stone ridges running through the central margin of the site delineate a band of stone-free sand that is possibly an access way; two prominent stone linears running east-west along the southern edge define a similar band. A feature of the north-west quadrant is a small, possibly rectangular, construction just north of the core stone scatter, whilst two large areas of ash-rich ancient midden along the eastern margin run close to possible stone alignments, and there may have been little separation of discard areas and occupational/activity space here.

Excavation

Excavation of Trench 1 underway. Facing north-east

Excavation in Trench 1 commenced through surface windblown sand (11222), revealing more of the stone scatter (11227) visible on the surface. This remained concentrated along the north margin of the trench, but lost much of the linearity that defined it at surface level. The stones were found to lie on sand (11226), some of which bore prominent ashy lenses (11228) or at a slightly lower level a loose deposit of chaff-rich silty sand (11239), and were clearly not in situ. (11239) overlay a thick surface, 11238, comprising laminae of chaffy mud plaster with occasional lenses of ash and other living debris. Its upper horizon was irregular, seemingly from post-abandonment pitting. One possible pit was traced through the overlying fill in the north-west corner, but in general pit edges were not identifiable in the loose fill above.

Stone scatter (11227) in section in the west baulk. It overlies ashy sand (11228), lensing into cleaner sand (11226). Facing west.

Damaged mud-plaster floor 11238 across the north margin. Facing west.

Beneath the surface sand across most of the eastern margin was a loose deposit of chaff-rich sand (11229), which overlay a similar but slightly more chaff-rich deposit, (11243). Both seem to represent destroyed plaster floors, mixed with sand. Removal of these deposits revealed a line of in situ stone, [11240], sunk into the sandy gebel. This clearly represented the foundations of a wall running east-west and ending in the approximate centre of the trench. For most of its length as exposed it appears to be one stone wide, although a collection of stones projecting from the northern side of its west end is mortared in position. This could indicate that the wall was originally wider, or mark the location of a projecting wall or similar architectural element.

East-west wall foundation [11240] with, in the background, the remains of plaster floor 11232 and damaged oven [11233]. Facing west.

Along the west margin of the trench, removal of the immediate surface sand revealed an underlying deposit of slightly ashy sand (11230), which in turn overlay a mud plaster surface, 11232. Built into the northern edge of the surface was part of the south wall of a damaged ceramic oven, [11233]. The ash from this emplacement lensed out to the north-east, presumably after the removal of the oven edge here, and to the north had mixed with windblown sand to form deposit (11228). The floor surface had been truncated along its eastern and southern edges.

Removal of surface sand across the southern margin revealed further sub-surface sand, partially covered by a shallow crust of mortar nodules in the south-east corner. The sand gradually resolved into two separate deposits, (11244) and (11241), contained either side of a rough north-south stone wall, [11245], preserved to two to three courses and running into the southern baulk. At its northern limit, the wall abutted a hard mud plaster floor, 11246/11249. At this point an irregular band of stones [11259], seemingly mortared in place, continued over the plastered surface, curving slightly to the north-west. This may relate to a similar collection of stones [11263] on the surface of the plaster floor 11232 in the central west quadrant. In the gap between the plaster floors was a distinct circular deposit of ash and charcoal, (11247), probably representing the lower fill of another oven that had been entirely removed. The surface of the gebel across most of the south-west quadrant was slightly hardened, probably from indirect exposure to heat. At least two additional features may represent the locations of ovens or storage bins, or less likely partially sunken storage vessels. The southern edge of the floor plaster in the immediate south-east of the trench, 11249, bore an impression from a circular feature of appropriate size for a ceramic oven or bin, whilst a circular pit of similar size cut the plaster floor 11238 in the immediate north-west corner of the trench. Whilst neither feature had distinct concentrations of ash associated with them, ash appeared as a diffuse element in much of the overlying sand.

Southern half of the trench, facing east, showing the roughly constructed stone wall [11245] extending into the southern baulk. The dark circular patch by the scale is (11247), probably in situ residue from a destroyed oven.

Discussion

It is difficult at present to relate the different architectural elements exposed in the trench to one another, and to gain an impression of the spaces represented. The floor surfaces lie on the same upper horizon, supporting the idea that they were part of one space, or related spaces. The top of east-west wall foundation [11240] lies some 15–25 cm below the top of the floor surfaces, although only 5–15 cm below their lower horizons, and it is easy to envisage the floor surfaces building up around the wall. The western edge of the wall appears to abut a patch of floor plaster, 11232, largely truncated so that only its lowest horizon survives.

Trench 1 at the completion of excavations in 2005 – 2006, facing south.

It is likely that the stone wall in the south-east corner, [11245], was erected at the same time as floor surface 11246. The rough line of stone [11259] was clearly laid after these floors, but the lack of intervening deposit between them may indicate that little time elapsed between these construction events. This construction, and [11263], perhaps served as retaining walls for ovens. It is not clear if the east-west wall foundation [11240] represents the northernmost wall of the core site. The thick floor surface 11232 to its immediate north is perhaps unlikely to have been uncontained, but it will be necessary to extend the trench to investigate this.

The lack of obvious roofing material throughout the trench in general raises the possibility that it encompasses open or lightly roofed spaces, which would seem to fit with the presence of ovens.

At least two ovens are represented in the trench, and there may be up to five. If these were used for baking bread, there is some justification for viewing this as part of a bakery quarter, and possibly broader food preparation area. If so, this contrasts with the situation at the Workmen’s Village – which in its location and size provides perhaps the most immediate point of comparison – where ovens were generally found singularly, associated with individual houses. Might this difference in the organisation of space imply greater external – and by implication official – influence on the control of food supply at the Stone Village?

By far the most common building material encountered was stone, few if any of which were faced. The stones of wall foundation [11240] had been laid to provide a level upper surface, but the footings of wall [11256] were set at an angle and stepped out slightly from the overlying courses. No clear trace of a foundation trench was found for either wall. Not infrequently, lumps of gebel-rich plaster some 10 – 20 cm in length were encountered, with concave impressions; these were clearly used as mortar ‘plugs’ between the stones. Occasionally, they remained attached to stones. Less often, smaller pieces of gypsum plaster were found with similar concavities, implying they were also used as architectural bonding, although none were found in situ. One striking feature was the overall lack of brick encountered during excavation, either intact or degraded. Unless the brick from this area was redeposited elsewhere, it does not seem to have been a prominent construction element, at least in this small part of the site. It remains unclear how high the walls were built in stone, and what additional construction materials were used.

A likely explanation of the concentration of stone (11227) initially visible across the north margin is that it is collapse from wall [11240]. This implies that the damage to floor surface 11232 took place before the wall collapsed. It is difficult to understand why the ovens seem to have been particularly badly damaged, but it may simply be that they provided accessible voids through the thick mud plaster floors. It is not clear from the archaeology when this disturbance took place, although it may in part coincide with a period of extensive looting known to have occurred at the Workmen’s Village following the EES excavations there in the 1920s. Several modern items were encountered in the loose surface sand, some wedged around the loose rocks of (11227). This implies that the wall was standing to a substantial height – perhaps over a metre – within the past century, although if this were the case, it is difficult to understand why the site seems to have gone unrecorded until the 1970s.

The excavation provided reassuring confirmation that the stone linears visible on the surface relate to sub-surface alignments, albeit displaced laterally, whilst the undulating deposits of clean and gravel-rich sand probably relate, largely, to areas of post-abandonment pitting. Despite the latter, excavation and survey in 2005–2006 has confirmed the potential of the site to yield data relevant to the project aims.

Adding the final touches to the Trench 1 plans and field notes. Facing south-east.

Artefacts (excluding ceramic vessels)

A fairly extensive range of object and material types is already represented amongst the artefact corpus generated by the 2005 – 2006 season: stone furniture and vessels, textiles, worked wood, faience jewellery, stone tools and small metal implements are all attested. One material type as yet not encountered is glass. A prominent feature is the large number of basalt chips visible both on the surface and in sub-surface deposits, potentially suggestive of a significant stone-working industry. The degree of preservation of organic material is particularly high; comparable to that encountered during excavation at the Workmen’s Village, and noticeably greater than that seen at the recent excavations at Grid 12 in the Main City.

A selection of objects from the 2005 – 2006 excavations is illustrated below.

37186 painted wooden ‘knob’

37306 linen fragment

37384 basalt hand axe

          
37315 part of a stamped gypsum jar seal, the impression unfortunately illegible

Faience necklace and collar pendants. L-R: 37157 (Hathor head), 37374 (Bes-image), 37161 (Bes-image), 37158 (Bes-image) and 37320 (cornflower)

Ceramics

The procedure for dealing the ceramics has been to retain all diagnostics and count and discard non-diagnostics on site. The diagnostics from the 2005–2006 season remain to be processed, although one initial impression from sorting is that ‘meat jars’ and ‘hearths’ comprise a prominent element of the corpus. A rim sherd from a carinated bowl is unusual in that it bears black painted decoration on its interior, probably representing a fish. One sherd had an incised geometric potmark, and another a small part of a hieratic jar label.

Environmental material

The dry desert conditions mean that the site promises much in the way of environmental remains, and indeed the floor deposits in particular preserve much plant material. Samples have been collected from all excavated contexts for future processing.

Faunal remains

Phillipa Payne
The 2005 – 2006 excavations in Trench 1 yielded 232 fragments of animal bone. This material was processed in early 2006, at which time work concentrated on the mammal bone. A preliminary discussion of the results is presented here.

The 232 elements were spread over 12 excavation contexts:

Context	Sampled mammal bone	Context description
11222	–	Loose surface sand
11224	–	Loose sub-surface sand
11225	1	Gravely sub-surface sand forming slight ridge
11226	1–5	Interface layer between clean sub-surface sand and underlying deposit 11239 (see below)
11228	1–16	Loose clean sand with bands of ash-rich sand, the latter once oven fill
11229	1–7	Loose sand with mortar nodules and chaff
11230	1–7	Sub-surface sand with lenses of silt or ash
11232	–	Ashy sand associated with oven
11239	1–29	Loose sand with mortar nodules and chaff
11241	–	Loose sub-surface sand
11242	1–12	Chaffy sand, probably in part from destroyed floors
11243	1–13	Silty sand, in part probably degraded ancient plaster floor

Of the 232 elements, 87 were diagnostic mammal, 67 were un-diagnostic fragments measuring less than 5 cm, three were bird, 32 were fish and 38 were splinters of long-bone shaft, also un-diagnostic:

	mammal 84 fish 32 bird 3
	n=119 sampled

Pieces measuring less than 5 cm and splinters of shaft were weighed and recorded as present or absent but are excluded from this discussion, except to say that their fragmentary state may be indicative of processing, perhaps of grease and marrow extraction.

A sample of 84 mammal elements was obtained. Ribs and vertebrae have been included in the sample as they are a significant proportion of the remains: 16% and 20% respectively. Shaft fragments were also included if they were reused as tools or exhibited marks from knives or axes.

The mammal bone

All the mammals identified were from the caprid, sus or bovid families. Other domesticates that might have been expected (as they feature in other domestic areas) include canis familiaris, felids or equids, but none of these were evident. This suggests that the sample comes from kitchen waste rather than a general rubbish dump. The sample size is too small to produce MNI counts of any significance. 

The break-down of mammal species is as follows:

	sus 33 capra 20 bovid 28
	n=81

Whilst a division of animals according to both species-identification and group size is:

	sus 24 size 4 9 capra 4 cap/ovis 1 size 3 15 cattle 14 size 5 14
	n=81

Both charts exclude one unknown shaft fragment and the size 2 animals which are smaller than sheep/goat or represent very small examples of the capra family.

Overall, the material was in a fragmentary state with just 15 examples of complete bones or epiphyses and only 25 over 50% intact. The number of fragments under 5 cm numbered 67 and weighed just 44.6 gm. This could be suggestive of very intense usage of the carcass, involving extraction of marrow, fat and the less ‘meaty’ parts of the skeleton.

10%  present	43
25%  present	9
50%  present	10
75%  present	5
90%  present	5
100% present	15

There were examples of primary and secondary butchery, suggesting that meat was divided elsewhere and then divided again on site. The first division was done with a heavy axe (as Rosemary Luff suggests for the Workmen’s Village), whilst the on-site practice was to use a knife. Butchered pieces include:

• [11239].21, a piece of cattle pelvis (ischial shaft fragment with nutrient foramen) showing several knife cuts indicative of jointing, whilst the obverse shows damage from a heavier instrument. A small flake of bone remains in situ on the edge of the shaft.

• [11239].25, a cattle long bone showing careful use of an axe or cleaver, producing a step-like pattern cutting away from the unfused epiphysis.

• [11239].4: pig atlas showing knife cuts on the caudal extremities as evidence of dividing the neck between axis and atlas, a possible point of beheading.

• [11239].21: fragment of cattle pelvis with damage from a heavy instrument on one side (top image) and knife cuts from jointing on the other (bottom image)

• [11239].25: cattle long bone, where the careful use of an axe or cleaver has left a step-like pattern – best seen in profile in the right-hand image – cutting away from the unfused epiphysis.

Several of the bones showed gnawing patterns produced by rodents. [11243].3 is a rib from a small sheep/goat-sized animal which has been damaged by gnawing and has three cut marks from a slim blade on the medial surface. Over the entire sample, 8 pieces showed evidence of gnawing by rodents, and two further examples may have been digested by a dog or other large carnivore.

Just 5% of the complete assemblage showed fire damage, suggesting that burning rubbish was not the favoured means of waste disposal. Most of the burned samples were blackened rather than completely calcinated.

Several pieces may have functioned as tools. [11228].13, [11228].9 and [11228].3 all appear to have been worked into a point. A further example, [11242].10, has a direct parallel in wood. They show some, but not hard, use wear suggesting they were employed in working a soft material like leather.

[11242].10: bone fragment worked to a point, perhaps used for working leather or a similarly soft material.

Finally, fragment [11239].1 is an unfused femur shaft of a juvenile pig under the age of three (Schmidt 1972, 49), which had broken and re-healed during the life of the animal. Both epiphyses were present ([11239].2 and [11239].3), suggesting perhaps that the leg had not been butchered but disposed of while the connecting tissue of the unfused ends was still in place. In light of the serious injury that the bone exhibited, it was perhaps considered unfit for consumption. The bone had broken across the diaphysis and the weight of the body pushing down on the limb caused the distal part of the shaft to be crushed by the proximal. The distal was then forced up inside the Medullary cavity, occasioning regrowth of bone material around the fracture. The animal would have been in intense pain but no human intervention, either in the form of splinting or resetting the bone, had been attempted. The amount of regrowth in the region shows that the animal lived some time after the accident, which must have occurred reasonably early in its short life. The break seems not to have jeopardised the early slaughter of the animal though, following what may have been general practise at the site, with 63% of the recovered pig remains falling within the age-range considered juvenile. The break may also indicate that this animal was resident at the Stone Village as it would have been unable to walk a long distance to a market or abattoir. Pig rearing seems to have been a significant occupation for residents at the Workmen’s Village, and this industry should provide an important reference point for future work on the Stone Village faunal remains.

[11239].1: unfused femur shaft of a juvenile pig with re-healed break across the diaphysis

Publications cited

Kemp, B. J. 1978. ‘Preliminary report on the el-‘Amarna survey, 1977’, Journal of Egyptian Archaeology 64, 22–34.

Schmidt, E. 1972. Atlas of Animal Bones for Prehistorians, Archaeologists and Quaternary Geologists, Amsterdam: Elsevier Publishing.