8.0. Discussion and summary.
The foregoing analyses have shown at least four and possibly five episodes of use of the Old Kingdom portion of the excavated area, not all of which were for strictly habitation purposes. This is consistent with the pattern of occupation found by Giddy (1987) at Ayn Asil where building periods were interrupted by abandonment of a portion of the site, the abandoned portion being used for temporary habitation or refuse disposal. In this chapter I will first summarize the depositional history and spatial patterns in the Old Kingdom levels of Kom el-Hisn, compare this pattern to other settlements, and discuss how Kom el-Hisn functioned in the context of the larger Old Kingdom society.
8.1 Occupational history.
Level 0: The majority of Level 0 deposits are part of the Upper Pottery Layer and are thus distributed over most of the upper levels of the site. Apart from the UPL, most of this level's deposits are restricted to a small portion of Middle Kingdom deposits in the southeastern part of the site near unit 1261-3/1074 and 1219/1095. Most of these deposits are dumps or redeposited dumps.
Level 1: This level is the topmost in the stratigraphic section and consists largely of dumps (eight total) along with one wall collapse deposit and a series of column bases in Room 9. Four of the ten deposits from this level come from Room 9, which contains an apparent short-term occupation after the room had begun to collapse. The dumps from this level are all characterized by typical Old Kingdom ceramics and are usually separated from the underlying Level 3 habitation deposits by one or more strata of wall collapse deposits. That these dumps and other deposits are not directly related to the habitation structures of Level 3 is indicated most clearly by the sequence in Room 9. In this case, the floor associated with the structure (DU-7) is covered by two deposits of wall collapse material, the topmost one (DU-5) being heavily decomposed wall material indicating that some time had elapsed since the abandonment of the structure. Two dump deposits (DU-3 and DU-4) were directly on top of this wall material which were in turn covered by the column bases, interpreted as a brief occupation. Similar sequences occur in Rooms 4, 5, and 6 where these types of dump deposits also occur. The obvious implication is that the structures had been allowed to decay for some period of time before dumping of refuse from other parts of the site were deposited within the partially collapsed structures. There is no pattern evident as to where this level's deposits occur.
The synchronicity of the various deposits assigned to Level 1 is somewhat conjectural at this point since chronological control is not fine enough yet to seriate individual deposits. Consequently, that all Level 1 deposits are contemporaneous remains a working hypothesis, yet it seems clear that they all post-date the main Level 3 occupations to one degree or another.
Level 2: The few deposits of this level are all directly associated with various adult and child burials. All of the burials found in this level are intrusive to the Level 3 deposits. This is especially clear in the case of the Room 15 tomb. The tomb structure itself was constructed within the confines of a large Level 3 room structure (of which Room 10 is a part). Thus, all deposits within this tomb can be assumed to be later than Level 3. The dump in the Room 15 tomb (DU-3) was sandwiched between two episodes of wall collapse derived from the tomb structure itself. This dump also contained typical Old Kingdom ceramics. Therefore, the burial itself and the wall collapse derived from the tomb structure (DU-2 and DU-4) were all assigned to Level 2, while the DU-3 dump, having been formed initially after the walls, was assigned to Level 1. Again, the contemporaneity of this dump with other Level 1 deposits is somewhat hypothetical and requires further examination.
None of the child burials contained obvious grave goods or coffin material and seem to have just been placed within the abandoned structures and then covered with nearby sediment (usually wall material). Of the adult burials, only one contained an object directly and unambiguously associated with the occupant, a bronze or copper mirror found in Room 15 directly atop the skeleton. All of the adult burials had evidence of coffin plaster surrounding the bodies. Two of these adults, Rooms 15 and 20, were contained in mud brick tomb structures. The male in Room 20 was disturbed by later excavation/sebakhin activity and any grave goods associated with the upper half of the body would have been removed with that portion of the skeleton.
The obviously intrusive nature of all of the burials with respect to the Level 3 habitation structures points to post-abandonment use of this area as a cemetery. Dumps overlying these burials are usually separated by some wall collapse indicating that some time also elapsed between the burials and episodes of refuse disposal. Since the dumps overlying the burials are of Old Kingdom date, the burials are necessarily Old Kingdom as well. Beyond setting a terminal date of the end of the Old Kingdom for this sequence, chronological control is inadequate to set specific dates for these events.
Level 3: These are the main occupations associated with the excavated Old Kingdom structures. Most of the structures were located within the top 10-20 cm of the surface below the capping UPL. Most of the architecture presents a fairly coherent structure (see foldout map). From the architectural features and artifact analyses in preceding chapters, I would argue that this set of rooms is at least part of a larger functional unit primarily devoted to the preparation and distribution of food. It is not clear whether this is part of a larger habitation unit as yet uncovered or is a service area providing for residents in another part of the site. However, the functional differentiation of the rooms militates in favor of a set of functionally interdependent structures.
Three rooms contain brick pit structures presumably used for storing grain. These rooms – 2, 5, and the room directly north of Room 5 and not used in this analysis -- are all clustered towards the northeastern part of the room block. The artifact contents of the rooms containing these pit structures are not particularly informative. Ceramics are generally sparse and not distinctive either in the floors or the pits themselves. The same is true of lithics, although two deposits from Room 5 (DU-6 and DU-10) contained distinctive globular chert nodules with rubbing wear on several faces. The function of these objects is unclear and not related to the pit structures in any obvious way. Plant remains were also sparse in these rooms as a result of there being no burning evident and hence little chance for preservation. Apart from the structures themselves about the only thing these rooms/pit structures have in common are the presence of fish cranial elements in the pits themselves which were probably dumped there when the main function of the pits was complete.
Room 12 had a distinctive set of artifactual and faunal remains. It contained numerous ground stone fragments, most of which were of a single type of stone (sandstone) and seem to represent both upper (mano) and lower (metate) parts of grinding apparatus. All of these were found in the southern portion of the room along with many of the faunal remains. Evidence that at least some of the grinding stone fragments are in their use locations is indicated by specimen #8 from this room, a thick, loaf-shaped fragment of a metate found resting on a clay setting. This room is also relatively rich in faunal remains containing nine different species, a large number of which are fish. A large portion of a Trionyx skeleton was also found in this room. The ceramics from this room are predominantly bowls, notably Types F (flared bowls), I (bowls with a molded rim), and O (small bowls), but also jars of Types B and C. The absence of either bread mold (Types D and E) indicates that if grain was being processed in this room it was not being subsequently used in the preparation of bread/beer in this location.
Nevertheless, the presence of grinding stones in apparent use locations and a wide variety of faunal remains, including a generally rare Trionyx, suggests this room functioned as a processing location for grain and animal foodstuffs. No evidence of cooking of this material was present which probably accounts for the lack of plant remains from this room.
The adjacent room, 13, is markedly different. The species richness and distribution for this room is similar to that of Room 12 but most of the faunal remains from Room 13 came from a wall collapse deposit that may contain some later dump material; the floor associated with this room contained no faunal remains and the circular brick structure contained only a single ovicaprid bone. The ceramics associated with this room tend to be large, heavy types, either Type D Bread platters or Type E Bread molds. Either of these types are associated with bread baking. As noted in the general description of this room, the ceramics and mud brick structure argue for a bakery as the sole function of this room. Its immediate proximity to Room 12's grain processing facilities, of course, indicates that at least some of the grain processed in Room 12 found its way into the bread oven next door.
The only other room for which a specific function can be reasonably inferred is Room 17. The faunal and ceramic distributions for this room are very distinctive from all other rooms. It is one of only two rooms where ovicaprid remains outnumber pigs, the other being Room 22 (ovicaprids also outnumber pigs in the dump deposits of unit 1192/1035). It is also only one of a handful of rooms where Type L contracting-mouth bowls are found in relatively high abundance. Further, the plant assemblage from the hearth and floor deposits contains a suite of taxa more consistent with the use of cattle dung as a fuel (Moens and Wetterstrom 1988). The co-occurrence of constructed hearth features, a distinctive set of faunal remains, and ceramic types (C and E) that are typically associated with cooking and whose physical properties are suited for such a regime suggest a fairly specialized function for this room, in this case cooking. Nearby, a series of dump deposits in a topographic depression, unit 1192/1035, have similar distributions of ceramics and faunal remains. The obvious implication is that the refuse from Room 17 was transported to 1192/1035 and dumped there. This short-distance transport of refuse indicates, following Hoffman (1974), a non-elite area of activity, though the goods and services produced here may have been meant for elite use.
Room 17 also contained abundant sickle blades, two fragments of a bifacial knife, and numerous blade blanks. This was also one of the few rooms where debitage was found (Cagle 1991). Consequently, at least some stone tool maintenance activities and possibly the production (from pre-existing blades) of sickles were carried out in Room 17.
Redding (1991, 1992, ND) has discussed the significance of different species in the Old Kingdom Egyptian diet. He suggests that, prior to the New Kingdom, pigs were kept and maintained by local populations (that is, the non-elites) primarily as a low-cost protein source, while ovicaprids were kept for their milk and wool with only a portion culled out and consumed primarily by the higher classes. Assuming that Room 17 is a specialized cooking facility with access to high quality fuel, it is reasonable to further posit that the output of this room’s products were intended for elite individuals, perhaps administrators or other specialists.
This is contrasted with other rooms with clear evidence of cooking (Rooms 1 and 8) where pigs dominate and the plant taxa are more indicative of a variety of plant fuels, such as stems and brush. What other functions these rooms may have performed (apart from open-fire cooking) is not clear from the artifact, faunal, or floral material. Both of these rooms contain ceramic types G (Bent-sided bowls), I (Bowls with molded rims), and a large number of Type O small bowls. Room 3 DU-1 adds Type D (Bread platters), while Room 8 DU-4 substitutes Type H (Meidum bowls) and B (Pointed jars). These may all be considered as generalized serving wares.
Other Level 3 structures are difficult to interpret. The curving wall that defines Room 7 does not appear to be related to the rooms on either side (Room 4 to the north or unit 1220/1072 to the south. None of the artifact classes distinguish it in any way, and there are no consistent similarities to other rooms. The charcoal flecking in the DU-2 floor may represent some small amount of food preparation (cooking) but the lack of specific burned areas argues against any form of hearth. Similarly, Rooms 6 and 8 show no distinctive patterns. These two rooms may have originally formed a single unit, but was later divided into two separate rooms.
The two structures excavated in the far western part of the site, Rooms 22 and 23, are problematic. No real features were uncovered during excavation apart from two diffuse lenses of burned material in Room 22 DU-2. These burned areas are somewhat similar to the patches found in Room 1 DU-3 which were interpreted as open-hearth cooking areas. The ceramic assemblages from these two rooms represent a fairly generalized set of types. Most of the bowls are represented with especially high numbers of Type H (Meidum bowls). Type D (Bread platters) are also fairly abundant. This distribution is quite different from that of Room 1 DU-3, but somewhat similar to that of Room 12 DU-2 with which these deposits were clustered. The faunal remains are somewhat similar to those of Room 17 in having a relatively equal number of pig and ovicaprids though in Room 22 this is based on only three identified specimens. This room also contained a distinctive net weight, two grinding stone fragments, and several other ground stone fragments. Floral remains are available from Room 23 DU-2 only and these are low density across all categories.
The overall impression one gains from these two rooms is that they functioned in ways distinct from those in the other portions of the site. Individual artifact distributions resemble one or another of the rooms in the main occupation area, but these similarities do not carry over to other artifact categories. The ceramics argue more for a generalized set of functions rather than the specialized character of the rooms in the other area, perhaps containing a few activities (evidenced by the net weight) not present in the other excavated areas. Clearly, however, there is some major disjunction in the use of space between these two areas.
Level 4: This level represents 38 deposits from ten excavation units. All of these were defined by their relation to the uppermost architecture in a unit, or their depth relative to similar deposits in other excavation units. Most of these deposits are overlain directly by Level 3 structures. Substantial exposures occur in units 1235/1056 and 1166/1066. For the most part, however, Level 4 deposits consist of some form of wall collapse lying below the Level 3 architecture. All ceramic types are present though in generally low numbers, except for Type C (Bread molds) and Type B jars, the latter of which is very abundant in the floor of Room 4 DU-7. The overall distribution of faunal remains is also fairly similar with pigs dominating the mammalian element and fish as a secondary component.
Because of the paucity of deposits no spatial patterns are observable. However, several aspects of Room 18 DU-8 are similar to some Level 3 deposits. This DU contains several distinct patches of burned material which are interpreted as open-air hearths. The plant component of Room 18 DU-8 is very similar to that from other similar black patches in Room 1 DU-3: a large percentage of unidentified plant taxa together with a relatively high proportion of cereal grains and low numbers of other plants. In Room 1 (and Room 8) this type of deposit and distribution of plant taxa was interpreted as a generalized cooking area using a mix of fuels in the form of stems and brush, compared with the more specialized facilities in Room 17.
The architectural plan shown in the foldout map indicates that there may be two distinct areas represented in the main occupation area. These two areas are separated by a large wall running north-south along Rooms 8, 6, and 18 to the west and along Rooms 4, 5, 7, and unit 1220/1072 to the east.
The series of rooms to the west of this wall constitute a coherent larger structure consisting of Rooms 6, 8, 10, 12, and 13. The only entrance to this block of rooms apparent from the visible architecture is along the south wall near the empty space south of Room 8. The main room in this block was not defined as a single entity but is L-shaped with the northern arm excavated as Rooms 6 and 8. This large L-shaped room offers entrance to Room 13 and that containing Room 10 (hereafter, this larger structure containing Room 10 and the Room 15 tomb will be referred to as Room 10). Access to Room 12 is gained only via Room 10 and perhaps through a possible doorway in the northwest corner of Room 12. Thus, this block consists of four main rooms, with the wall between Rooms 6 and 8 being built sometime during Level 3 to split this northern wing of the L-shaped room in two.
The functions of Rooms 6 and 8 are unknown but to some extent were used for cooking over an open fire. Room 13 is thought to be a bakery and Room 12 was probably used for plant and animal processing. The deposits from the excavated portion of Room 10 are not associated with the Level 3 occupation of this structure but represent a mix of wall debris from the construction and subsequent collapse of the Level 2 tomb structure built within it (Room 15).
The eastern block of rooms is less coherent overall, but some structure is seen in the northern half, comprised of Rooms 2 and 3/5 and the structure immediately to the north of Room 5. The jumbled nature of the visible mud bricks makes interpretation of the overall structure difficult. There do not appear to be any connections between these structures and those to the west. Entrance to Room 5 may have been gained through what appears to be a doorway in the center of the south wall and through the southwest corner of Room 2. Rooms 3 and 5 comprise a larger structure in which the individual rooms were divided by a short wall. Room 3 may also have been further subdivided into two very small sections by the divider projecting from the east wall of that room. Entrance to the room to the west of Room 2 (unexcavated in depth) seems to have been gained only via Room 2. The "pavement" to the north of this room may be the north wall of that room that fell over largely intact. The structures to the south of these four rooms (Rooms 4 and 7 and unit 1220/1072) do not have sufficient visible architecture to relate them to the overall structure.
All of the northern four rooms have specially constructed storage pits associated with them. Altogether, this block of four rooms contains at least six pits constructed of mud brick: three in Room 5, one in Room 2, and two visible in the unexcavated room. The distribution of ceramic types in Rooms 2 and 5 are similar with a wide range of types, but Room 3 only contains only types H, K, and O, suggesting that Room 3 may have performed some specialized function within the larger scheme of grain storage.
The structures to the south have no readily discernible function. The Level 3 deposits of Room 4 contain no detectable floor, which may have been obscured by the activities associated with the later infant burial in this room. The curving structure of Room 7 may be associated with Room 4 but, as indicated above, the artifact inventory and the character of the deposits does not suggest any particular function. The deposits in unit 1220/1072 are probably part of a shallow depression in which debris from adjacent structures washed in.
The most inscrutable room in this complex is Room 9. This room, which lies just east of Rooms 4 and 7, contains few occupational debris. The floor consists of a series of rectangular hard-clay structures provisionally interpreted as column bases. Because only a portion of this room was excavated no relation to the other structures in this area is available. However, it is the only room where no evidence of food preparation or consumption is evident. This may be only a portion of a much larger structure reminiscent of the columned porticoes observed fronting the residence blocks at Abydos (for example, Figure 8, p. 15 in Wegner 1998). If this is the case, then the structures in the main excavated areas may be the support structures for as-yet-unexcavated habitation units to the east.
Based on reports of earlier investigators (Edgar 1915-1919; Griffith 1888) the structure of the site as a whole probably consisted of a large oval or circular enclosure wall with a principle temple or administration complex located to the west under the modern village. As at other large towns, such as Abusir and Giza, this administrative district would be surrounded by a complex of buildings housing the administrators and their support staff. The habitation areas of these centrally administered pyramid/temple towns generally consisted of distinct blocks of residential units with repeating room units. The example of the village adjacent to the tomb of Khentkawes given in Chapter 2 has each residential block consisting of 3-4 rooms containing a grain storage room, kitchen area, and a larger general habitation room. This is contrasted somewhat by the larger, more centrally-planned towns of the Middle Kingdom such as at Kahun (David 1986; Petrie 1890, 1891) where room blocks were larger and contained courtyards, columned porticoes, living quarters, and ancillary rooms containing bakery-breweries, granaries, butcheries, etc.
The structure of the excavated areas described herein appears to be closer to the Middle Kingdom model. The east and west sets of rooms may each service a different set of people and the excavations coincidentally straddled two distinct functional areas of each. However, because of the limited extent of the uncovered architecture, these two sets of rooms may be related after all, with access to both obtained by other entrances to the north or south of the excavated areas. If this were the case, and if the columned area of Room 9 represents the habitation area associated with these other rooms, then a case could be made that what we have uncovered are the support areas that service inhabitants of further structures to the east.
The number of rooms represented here is larger than the blocks of 3-4 rooms described for the other Old Kingdom towns around Giza and Abusir. In these cases, the basic habitation unit was relatively small and probably only serviced a few people, perhaps a single family. Here, there seems to be a larger supported unit. That three separate rooms are devoted to grain storage seems to imply more than a few members of a single family. Food preparation is also divided between the intensive features of Room 17 (sharply defined hearths, cattle dung for fuel, and a higher percentage of ovicaprids) and the more generalized cooking areas represented by Rooms 3 and 8 (diffuse hearth areas, a variety of fuels, and a dominance of pig). Minimally, food preparation was directed at two different groups of people. As Redding (ND) suggests, this could reflect a more elite diet which included more ovicaprids contrasted with a commoner diet of predominantly pig.
That there is functional differentiation in such a restricted area implies that this was not a centralized area of industrial activity servicing the entire site such as is found at Hierakonpolis earlier. Rather, it seems more analogous to the structure at Kahun where large room blocks were devoted to an elite family or set of families and their attendant workers. However, at Kahun, each habitation block was more self-sufficient than that which obtains at Kom el-Hisn in that it contained numerous areas devoted to a wider range of activities such as weaving, carpentry areas, etc. While many of these activities may have left no material evidence at Kom el-Hisn, the rooms here tend to be centered around food preparation rather than a complete set of activities.
I interpret this to mean that Kom el-Hisn was in some ways dependent on the surrounding economy for much of its subsistence needs and partly independent of it. Pigs apparently were raised and consumed by some of the resident staff, implying that the residents themselves were not directly involved in intensive agriculture, though the presence of numerous sickle blades in various states of use and manufacture argues for some involvement. This militates in favor of their obtaining much of their grain from local fields rather than as assignments from the central government. However, it also seems reasonable to assume that since the residents were involved in the harvesting of grain that the fields were the property of the residents. Pious foundations often included tracts of land for support of the cult and this could have included fields for growing food to support the human population as well as grazing land or for growing fodder.
The presence of locally obtained fish and at least one net weight further indicates that some of the residents were involved directly with procuring local resources. However, many other resources, such as marine fish and stone, had to have been obtained from a distance. The ovicaprid data also suggest that many of these animals were exported rather than being consumed locally. The same is true of bovid remains. Thus, Kom el-Hisn was also involved in trade of a more national nature, perhaps trading sheep/goat or excess agricultural produce for non-locally available materials.
Rather than being an isolated island of state control, I would argue that Kom el-Hisn formed something of a middle ground between complete integration with the local economy and total independence of it. This model has textual support from the Middle Kingdom temple complex at Kahun where some papyri deal with business involving the surrounding community (see Chapter 2) while others involved strictly temple business.
This interpretation may shed light on the question of who the residents actually were. In some of the Old Kingdom pyramid towns, such as Abusir, the priest inhabitants were in residence for short periods of time rather than full-time inhabitants. The habitation blocks associated with these towns tended to be smaller with only a few rooms devoted to each household unit. Later in the Middle Kingdom residents inhabited the buildings full-time and the support structures were much larger and more complex. As I have suggested, the room structure presented here seems closer to the Middle Kingdom model but probably not as elaborate, having more to do with immediate subsistence needs rather than the full range of activities required for a family unit to be entirely self-sufficient. The differential distribution of ovicaprid and pig remains suggests that at least two populations were actively inhabiting Kom el-Hisn: one group of elites most likely engaged in non-subsistence activities, and another of support staff probably drawn from the surrounding population who provided for some of their own subsistence needs through the maintenance of herds of pigs.
This model implies that the residents were not involved in full-time agricultural production and required an ancillary population of support staff to provide subsistence and other domestic needs. Some of the surrounding land was part of the town's property and provided forage for animals and perhaps produce for the local population. This specialization required some dependence on an extra-local economy for some basic goods and services, which I have argued was a combination of direct procurement of local resources (e.g., fishing), use of local labor, and some long-distance trade/exchange nationally. The exact nature of the duties of the elite residents is not a settled issue. The data presented here lend some additional support to the hypothesis that one of the functions of Kom el-Hisn was as a specialized center for cattle production related in some way to the Estate of the Cattle. The close association with Hathor may also suggest the presence of a cult center at Kom el-Hisn.
This scenario of a major center involved in both the local and national economy lends some support to Trigger's 'territorial state' model and to Kemp's (1983) suggestion that active local economies operated alongside a national system of redistribution of certain goods. While no direct evidence of a strictly appropriative relationship between Kom el-Hisn and the central government has been presented, the fact that it was not wholly self-sufficient in basic goods and services suggests that at least one substantial Old Kingdom community was at least partly left to fend for itself apart from the national state structure.
In terms of developing a general explanatory theory for the development of complex societies, I have argued that Darwinian evolutionary theory can be a productive framework from which to develop appropriate models. If, as Dunnell (1978a, 1978c, 1995) and Dunnell and Wenke (1979) argue, the development of complex society is a result of the scale of selection moving from the individual to groups of functionally interdependent individuals, then determining the empirical units on which selection occurs is of utmost importance. This is the principle methodological problem dealt with by this research.
When dealing with a complex society such as Old Kingdom Egypt, if one accepts that the evolutionary "individual" is an aggregate of functionally differentiated individuals, then the data one uses to describe those individuals are aggregate data as well. In this case, the aggregate data are the phenotypic expression of the individuals: artifacts and other cultural remains. A key issue involves how the aggregate data were created; that is, do they represent a set of functionally relevant units and are they historically related to one another such that their spatial distributions represent their systemic (sensu Schiffer 1972) context?
I addressed the former question by using existing artifact types that are (arguably) largely functional in nature. This provided a basis for arguing that the spatial variation observed is, in fact, functional. I then attempted to establish the historical relations between the artifact types by using sedimentary principles that defined the spatial and temporal relationships between the deposits in which artifacts were physically aggregated. Thus, the aggregate data used to establish functional differentiation are derived empirically from the depositional events that created the assemblage rather than assigned by simple (contemporary) proximity. This strategy is largely in line with much current work involving formation processes. Where it differs is in the attention paid to the depositional history of the deposit as a whole rather than the inferred depositional histories of individual artifacts contained therein.
This study is a first step in establishing a set of units that can be used to explain the development of complex societies within a theoretical framework that can be used in truly cross-cultural fashion. Much further work at Kom el-Hisn and elsewhere is needed to refine the model developed here. Certainly, more rigorous analytic units can be created, units that unambiguously differentiate between homologous and analogous similarity. Once a set of appropriate functional and stylistic classes have been developed, the former may be used to explore functional differentiation in much more detail, while the latter can assist in establishing better chronological controls through seriation analyses and perhaps elucidating the types of social interaction that took place across the site as well. This research has, I believe, provided a sound basis for further exploration of many of these issues of more general archaeological importance.