Differential Consumption of Pig vs. Sheep/Goats at the Old Kingdom Site of Kom el-Hisn.


Paper presented at the 53rd Annual Meeting of the American Research Center in Egypt, Baltimore, Maryland.

Note: Figures and tables appear at the bottom of this document.


       Much of the discussion regarding Kom el-Hisn in recent years has revolved around its possible function  as a specialized center for raising cattle for export to other areas of Egypt. Some e.g., (Redding 1992) have suggested that the complementary nature of the faunal assemblages from Kom el-Hisn and the recently excavated workmen’s areas adjacent to the pyramid complex at Giza indicates that the Giza could well have been the ultimate destination for many of the catle raised at Kom el-Hisn. While the role of Kom el-Hisn within the larger economic structure of Old Kingdom Egypt is certainly an important issue, there are many other questions which may be addressed by the faunal assemblages. Many of these questions relate to the spatial structure of the site and the socioeconomic status of the inhabitants and their subsistence base.

       Over three seasons from 1984 to 1988, excavations were carried out at Kom el-Hisn under the direction of Robert Wenke and Richard Redding. Recently, I performed an extensive spatial analysis of the material recovered from two seasons, 1986 and 1988 and I was able to clarify the structure and function of a good portion of the excavated areas. This paper deals with one aspect of that structure, the differential distribution of faunal remains and its implications for social differences in subsistence; namely, the role of pig versus sheep/goat in the diet of different segments of Kom el-Hisn’s residents.



       The role of the pig (Sus scrofa) in the diet of ancient Egypt has been somewhat confusing, especially from the Old Kingdom, for one simple reason: there is an almost total absence of pig remains from tomb provisions and a similar dearth of pictorial and textual representations from tomb and temple contexts. The main reason cited for this absence of evidence is mythological in nature: the male pig was often a manifestation of the evil god Seth and was thus considered a ritually unclean or impure animal much as it is today among various religions. Nevertheless, there is some scattered evidence for pigs in tomb contexts from the late Predynastic and Early Dynastic. For example, at Minshat Abu Omar, many of the graves of poorer people contain pig remains, though the graves of the rich have cattle instead, and votive figurines have been found from some First Dynasty temples (Houlihan 2001):47). In contrast, sheep/goats and cattle are often represented in tomb paintings and offering lists, cattle are abundant as tomb provisions, and all were used as sacrificial animals.

       Many explanations have been offered for this apparent lack of interest in the pig by Egypt’s elite and the pig’s association with a generally distasteful god (Seth). Certainly, the proclivity of  pigs to eat human fecal matter as well as much of the garbage disposed by people in urban settings contributed to this rather negative view. Certain disease organisms, notably Trichinella and Taenia, have often been associated with pork consumption and have been found in the bodies of ancient Egyptians (Miller 1990); (Cockburn and Cockburn 1983)). However, it is doubtful that the ancient Egyptians made the connection between, specifically, pork and these diseases, and in any case, pigs were raised and eaten by at least some of the population throughout Dynastic times.

       Nonetheless, more detailed excavations of settlement areas in the past two decades has greatly enhanced our knowledge of the role of the pig in Egyptian subsistence. Detailed faunal profiles have been generated from Neolithic Merimda (von den Driesch and Boessneck 1985), Predynastic Hierakonpolis (McArdle 1982) and Maadi (Boessneck, et al. 1989), Old Kingdom Giza (Kokabi 1980); Redding 1992, Middle Kingdom Tell el-Daba (Boessneck 1976), and  New kingdom Amarna (Hecker 1982). These sites show that, far from being rare or absent from the Egyptian diet, pig remains often constituted a large proportion of the mammalian faunal remains. Most remarkable are the extensive pens found at the Workmen’s Village at Amarna (Kemp 1985-86) and textual references of commodity transactions involving pigs at Deir el-Medina (Janssen 1975). It is notable in the latter case that none of the transactions involving pigs deal with deliveries to the site from the central administration, but were apparently strictly local in nature, which also seems to be the case at Amarna (Miller 1990):132). That is, pigs were raised, slaughtered, and used as trading commodities by the local workers and were not part of the larger network maintained by the state.

       Thus we are left with the impression that pigs, although present in varying abundances during all periods of Egyptian history and prehistory, were largely consumed by the lower classes and that the central government did not involve itself in the production and distribution of pigs. There remains the question of whether this phenomenon is visible archaeologically -- that is, can it be demonstrated at the site level -- and how it manifests itself in different parts of the site (reword). It is precisely these sorts of questions that the excavations at Kom el-Hisn were designed to answer.



       The excavations at Kom el-Hisn (Figure 1) took placed in two phases. In 1986 several randomly placed 2-meter units were excavated in addition to a 72 m2 shallow trench. In 1988 the large trench was expanded to reveal more architecture and in this area individual rooms (rather than arbitrary units) were excavated. In addition, two areas were also excavated where architecture was visible on the surface (Rooms 17 and 22/23). All units and rooms were excavated by natural stratigraphy, all ceramics (both body and diagnostic sherds) were weighed and saved, and other artifactual material (bone, lithics, etc.) was retrieved from the screens. At least one sediment sample and sample for flotation to retrieve plant remains were also removed from each sedimentary unit.

       The overall architectural plan (Figure 2) as well as previous analyses (Buck 1990, Moens and Wetterstrom 1988, Redding 1991, Redding 1992, Sterling and Wenke 1997, Wenke and Brewer 1992, Wenke, et al. 1988, Wenke and Redding 1985, Wenke and Redding 1986) suggest that this area was geared towards food preparation and storage, seen most clearly in the numerous grain storage facilities. The ceramics were dominated by coarse bread molds and bread platters (also referred to as offering trays) and various sorts of crude cooking or “beer” jars, often fire-blackened presumably from cooking fires I have also recently argued (Cagle 2001) that several of these rooms indicate specialization in certain aspects of food processing -- such as individual areas for baking, grain processing, butchery -- and thus indicate that this area was probably involved in more “industrial” food preparation for a large group of people rather than a set of domestic habitations.

       The overall goal of my analysis was to determine whether there were specific sets of artifacts that occurred together and how these were distributed across space. A secondary hypothesis was that rooms whose function was geared toward a single or few activities would have limited distributions of artifacts that were geared towards those functions. To accomplish this, I conducted a series of statistical analyses designed to group together room deposits with similar assemblages of artifact types. The main tool in these analyses is cluster analysis, used to group units with similar distributions of artifacts. The resulting clusters are then analyzed further to determine which artifact types are controlling the clustering routine and the functions that these types represent. Types that vary with other functional data, such as particular types of faunal remains or architectural features, represent loci of activity within specific structures. The  result is a map of the site designating discrete functions or sets of functions associated with architectural features.

       Because of a relatively small number of relevant Old Kingdom deposits (N=39) I chose to run each clustering routine on a single artifact type first and then compare the results from each. The first analysis used 12 ceramic types and resulted in four distinct clusters present, and upon closer examination one in particular is very striking, Cluster # 1. This cluster contains all of the deposits from two excavation units, Room 17 and unit 1192/1035. If one charts the types present in all four clusters (Figure 3) one readily sees that the deposits from these two excavation units are dominated by Type C, a form of crude jar, probably the ‘Ordinary traditional offering jar’ of Reisner ((Reisner 1931, Reisner 1932, Reisner 1942, Reisner 1955) Type A-IV; see also (Brunton 1927, Mond and Meyers 1937). In fact, over 95% of all Type C sherds occurred in these two units. Many of these were fire-blackened indicating their use as cooking jars. These vessels were very crude, produced either by coiling or low-rotation wheel, and had thick walls of coarse Nile C clay. These characteristics make them relatively cheap to manufacture and resistant to thermal shock produced in cooking.

       An examination of the stratigraphy and architecture in these two units makes this association clear. The deposits from Room 17 in this group consist of a floor surface (DU-5), a small dump (DU-4), and the contents of a pit structure (DU-3). The pit structure was constructed of clay and stood in the northwest corner of the room; otherwise Room 17 was devoid of any other features. Unit 1192/1035 lies approximately 43 meters southwest of Room 17 and consists of a series of sloping deposits with occasional laminated structures (indicating deposition by running water) and varying concentrations of dumped material. The overall structure and contents of the deposits imply that this unit was in a topographic depression into which material from adjacent rooms slumped into while other material was directly dumped from elsewhere. The similarity in ceramic composition suggests that much of this material was brought from Room 17 or at least a very similar location.

       To test this hypothesis further, I conducted a second cluster analysis using identified faunal remains (NISP) of cattle, sheep-goat, pig, birds, and fish. Once again, the deposits from Room 17 are clustered together, but this time with only one deposit from unit 1192/1035, DU-7. Also included within this cluster are the contents from a brick pit structure in Room 13 (DU-3) and a floor in Room 5 (DU-12). Referring to the graph showing the species composition within these clusters (Figure 4), it is apparent that Cluster #1 is dominated by sheep-goat remains. However, the only reason Room 13 DU-3 and Room 5 DU-12 are clustered together with these other deposits is because they both have a single identifiable bone, both sheep-goats, in relatively small deposits.

       The reason that other deposits from unit 1192/1035 are NOT included with the Room 17 deposits has to do with the presence of other species that are otherwise not well represented in Room 17; that is, they are not so completely dominated by sheep-goat remains. In fact, this cluster (#3) is somewhat misleading with respect to the deposits from 1192/1035. While the bar graph makes it appear as if pigs dominate in all deposits, in fact, only one deposit from 1192/1035 contains more pig than sheep-goat, DU-8 (N=4 and N=1, respectively). Further, if one calculates the the ratio of sheep-goats to pig, only seven deposits have a ratio higher than 1.0. Of these, six are from either Room 17 or unit 1192/1035; the other is Room 22 DU-2 which has only three identifiable sheep-goat and pig remains. Thus, with this one exception, only Room 17 and unit 1192/1035 contain more sheep-goat than pig. At this point, I have demonstrated some sort of functional connection between those deposits in the hearth structure in Room 17 and the dump deposits in unit 1192/1035. To see more specifically what this function or functions may be, it is necessary to look at other data and compare these two sets of deposits with others. A total of 13 deposits were available for examination of plant remains. Of these, data is available for only two of these deposits, both from unit 1192/1035. It must now be assumed that the material from these two deposits can fairly represent that from Room 17.

       Table 1 presents the categorical data for available plant remains, separated into seven categories. Of note are three categories, Chaff, Fodder, and OtherUnid or “Other-Unidentified”. Both of the units from unit 1192/1035 are very high in Chaff, higher than any other units, and both are relatively high in Fodder plants (see (Cagle 2001):267, Table 6.11 for the genera represented by these classes). Conversely, two other deposits from Rooms 1 and 8 (both floors) are both very high in Other-Unidentified remains and relatively low in both Fodder and Chaff. I have also included one deposit from a lower level (4) for comparison, Room 18 DU-X. Note that this unit is most similar to those from Rooms 1 and 8 in having a hig percentage of Other-Unidentified plants and low Chaff and Fodder. These differences become significant when one examines the source of the plant material.

       As Moens and Wetterstrom (1988) note, the primary source of charred plant material at Kom el-Hisn is derived from fuel for fires either directly or incorporated within animal dung. Little charcoal has been recovered, suggesting dung and/or plant stems rather than wood as the principal fuel. Dung fuel is usually prepared by mixing the dung with straw or some other tempering material, forming it into cakes, and allowing it to dry. Such methods are known from modern Egypt and throughout Pharaonic times. The chaff from grain processing could have served as a temper in this context and winnowing debris is also known as a major animal feed (van Zeist and Camparie 1984:4-5). Very few sheep/goat pellets have been found which tend to be common when used as fuel (Bottema 1984:208). Cattle are thus the most likely source for the Trifolium that is especially abundant in the 1192/1035 samples (Cagle 2001: 290) since cattle are often represented being hand-fed cereal, cereal stalks, straw, and cut fodder (Moens and Wetterstrom 1988:170) while pigs and sheep/goats were usually allowed to forage on their own.

       Thus, the samples high in Chaff and Fodder are probably a result of using cattle dung as the primary fuel, while those with large amounts of unidentified remains probably used a combination of various plant stems and brush as fuel. It will be recalled that the source for much of this material in Room 17 is a well-defined, specially constructed hearth structure with a distinct set of ceramics associated with it suggesting a degree of specialization in function, that is, cooking. The sources for much of the plant material in these other rooms -- 1, 8, and 18 -- are distinctly different. In these cases, the burned patches that probably served as hearths have no constructions associated with them, are shallow and diffuse, and presumably represent ad hoc hearths for short-term occupations.

       The conclusion that I have reached is that Room 17 functioned as a specialized cooking facility for some portion of the residents, probably in the more elite classes. The structure of this area is probably analogous to that found at Middle Kingdom Kahun where room blocks of special-purpose rooms were devoted to one or several elite families and their attendant workers (Cagle In press). Individual rooms specialized in different aspects of food preparation and storage suggests that this area serviced more than an immediate family and functioned in a somewhat indistrial capacity. Access to a regular supply of fuel in the form of cattle dung, which in itself required a certain amount of labor and other resources to prepare, further indicates the elite status of the occupants. These deposits are dominated by sheep and goat remains.   In contrast, the small burned patches in Rooms 1 and 8 that are dominated by pigs and are indicative of short-term occupations using a variety of plant matter as fuel and little, if any, dung fuel. These apparently brief occupations may, as argued by (Giddy 1987) , be small encampments (“squatter’s camps”) that took place after this area of the site had been abandoned; the chronology of the deposits is not sufficiently detailed at this time to definitively say whether these were coincident with the others. However, these small occupations were not as specialized nor do they appear to represent the same socioeconomic level as the inhabitants of Room 17.

       Thus, we have two sets of residents inhabiting the same site at (approximately) the same time and consuming a different set of animals, though with some overlap. The more elite residents consumed relatively more sheep/goats than pigs, and the lower class residents consumed predominantly pig. This division in animal use patterns is in agreement with other habitation sites, such as New Kingdom Amarna, and burial evidence from other Old Kingdom sites that show pigs as more or less restricted to the lower classes, with other animals, especially cattle but also sheep and goats, being more abundant in the upper classes. However, it should be stressed that pig remains at Kom el-Hisn are not entirely absent from the “elite” deposits thus showing that there is no absolute prohibition on the consumption of pigs on the upper classes.

Bietak, M.

       1975       Tell el-Dab'a.


Boessneck, J.

       1976       Tell el-Dab'a III: Die Tierknochenfunde 1966-69. Osterreichische Akademie der Wissenschaften, Denkschriften der Gesamtakademie, Band V, Vienna.


Boessneck, J., A. von den Driesch and R. Ziegler

       1989       Die Tierreste von Maadi und Wadi Digla. In Maadi III: The Non-lithic Small Finds and Structural Remains from the Predynastic Settlement, edited by I. Rizkana and J. Seeher, pp. 87-128. Verlag Philipp von Zabern, Mainz am Rhein.


Brunton, G.

       1927       Qau and Badari I. British School of Archaeology in Egypt, London.


Buck, P. E.

       1990       Structure and content of Old Kingdom archaeological deposits in the western Nile Delta: a geoarchaeological example from Kom el-Hisn. Dissertation, University of Washington, Seattle.


Cagle, A. J.

       2001       The Spatial Structure of Kom el-Hisn: An Old Kingdom Town in the Western Nile Delta, Egypt. Dissertation, University of Washington.


Cockburn, A. and E. Cockburn

       1983       Mummies, Disease and Ancient Cultures, Cambridge.


Giddy, L. L.

       1987       Egyptian Oases: Bahariya, Dakhla, Farafra, and Kharga during pharaonic times. Warminster.


Hecker, H. M.

       1982       A zooarchaeological inquiry into pork consumption in Egypt from prehistoric to New Kingdom times. Journal of the American Research Center in Egypt 19:59-71.


Houlihan, P. F.

       2001       Pigs. In The Oxford Encyclopedia of Ancient Egypt, edited by D. B. Redford, pp. 47-48. vol. 3. Oxford University Press, Oxford.


Janssen, J. J.

       1975       Commodity Prices from the Ramessid Period, Leiden.


Kemp, B. J.

       1985-86 Amarna Reports, I-III. London.


Kokabi, V. M.

       1980       Tierknochenfunde aus Giseh/Agypten. Annalen des Naturhistorischen Museums in Wien 83(519-537).


McArdle, J.

       1982       Preliminary report on the Predynastic fauna of the ierakonpolis Project. In The Predynastic of Hierakonpolis -- and interim report., edited by M. A. Hoffman. 116-121 vols. Egyptian Studies Association, Publication No. 1.


Miller, R. L.

       1990       Hogs and hygiene. The Journal of Egyptian Archaeology 76:125-140.


Moens, M.-F. and W. Wetterstrom

       1988       The agricultural economy of an Old Kingdom town in Egypt's west Delta: insights from plant remains. Journal of Near Eastern Studies 3:159-173.


Mond, R. and O. H. Meyers

       1937       Cemeteries of Armant, London.


Redding, R. W.

       1991       The role of the pig in the subsistence system of ancient Egypt: A parable on the potential of faunal data. In Animal use and culture change. MASCA Research Papers in Science and Archaeology, edited by P. J. Crabtree and K. Ryan, pp. 20-30. vol. 8, Supplement.


Redding, R. W.

       1992       Egyptian Old Kingdom patterns of animal use and the value of faunal data in modeling socioeconomic systems. Paleorient 18(2):99-107.


Reisner, G. A.

       1931       Mycerinus, Cambridge, MA.


Reisner, G. A.

       1932       A provincial cemetery of the Pyramid Age: Naga-Ed-Der Part III. Egyptian Archaeology VI. University of California Press, Berkeley.


Reisner, G. A.

       1942       A history of the Giza Necropolis I. Harvard University Press, Cambridge.


Reisner, G. A.

       1955       A history of the Giza Necropolis II. Harvard University Press, Cambridge.


Sterling, S. and R. J. Wenke

       1997       Attribute scale seriation of Old Kingdom Egyptian ceramics. Paper presented at the Annual Meeting of the Society for American Archaeology, Nasville, TN.


von den Driesch, A. and J. Boessneck

       1985       Die Tierknochenfunde aus der neolithischen Siedlung von Merimde-Benisalame am westlichen Nildelta. Mitteilungen des Deutschen Archaologischen Institut, Abteilung Kairo 49.


Wenke, R. J. and D. J. Brewer

       1992       The Archaic-Old Kingdom Delta: the evidence from Mendes and Kom el-Hisn. In Haus und Palast im alten Agypten. Sonderbruck.


Wenke, R. J., P. E. Buck, H. A. Hamroush, M. Kobusiewicz, K. Kroeper and R. W. Redding

       1988       Kom el-Hisn: Excavation of an Old Kingdom Settlement in the Egyptian Nile Delta. Journal of the American Research Center in Egypt XXV:5-34.


Wenke, R. J. and R. W. Redding

       1985       Excavations at Kom el-Hisn, 1984. Newsletter of the American Research Center in Egypt 129:1-11.


Wenke, R. J. and R. W. Redding

       1986       Excavations at Kom el-Hisn, 1986. Newsletter of the American Research Center in Egypt 135:11-17.




Figure 1: Map of the Delta showing the location of Kom el-Hisn (after (Bietak 1975)).




Figure 2: Architectural plan map of Area A of Kom el-Hisn.



Figure 3: Bar graph showing the relative percentages of each ceramic types in the ceramic clustering routine. Cluster 1 contains only Room 17 and unit 1192/1035 deposits.





Figure 4: Bar graph showing the relative percentages of the major taxa of a faunal clustering routine.


Table 1: Percentages of plant taxa by deposit for Level 3 deposits (including two Level 4 deposits for comparison, both Room 18).