Visit Us | Contact Us | Membership | Make a Gift | Calendar | Order Online | What's New

Print this Page

Home > Research > Publications > News & Notes


The Location and Importance of the Pyramid Settlement

By Mark Lehner, Associate Professor of Egyptology
The Oriental Institute
The University of Chicago

(This article originally appeared in The Oriental Institute News and Notes, No. 135, Fall 1992, and is made available electronically with the permission of the editor.)

Eight years ago I rendered the Giza Pyramids Plateau in an isometric drawing adapted from existing contour maps to illustrate how the landscape affected the mobilization of the Fourth Dynasty Egyptians (ca. 2550 B.C.) for the building of the Great Pyramid of Khufu, the first pyramid on the plateau (fig. 1). I thought that this first major construction project must have determined the way that the rest of the architecture and settlement developed across the landscape over the three generations of pyramid building at Giza.

Figure 1: Rendering of the Giza Plateau, with the Great Pyramid of Khufu in the Center.

The location of the workmen's settlement is particularly interesting since the estimates of their numbers ranging from 100,000, according to Herodotus, to 20 or 30 thousand, according to modern Egyptologists, were comparable to the populations of sizable cities anywhere in the Near East in the third millennium. The scientific excavation of such a settlement could tell us something about how the pyramids were built and what affect pyramid building had on the development of Egypt as a nation.

Each pyramid was a functioning temple site at the same time that major construction continued on the next pyramid. A settlement attached to a ritual center might differ significantly from one that involved a large labor force. Giza continued to be a ritual center well after the din of building moved back to Saqqara and on to Abu Sir. It would be interesting to see if this change from labor camp to temple community could be traced in the archaeological record.

Geomorphology at Giza presents important clues about the location of the third millennium settlement. A wadi or valley separates the Mokattam Formation, on which the pyramids rest, from the Maadi Formation to the south (fig. 2). This wadi, known as the "Main Wadi," probably served as the conduit for building materials brought into Giza from elsewhere, such as the granite and fine limestone from quarries across the Nile Valley that were used for the fine outer casing of the pyramids. Other material also had to be delivered: raw foodstuff for the workers, fuel for preparing bread and beer, the immense quantities of gypsum mortar used in building the Giza pyramids, copper tools, etc. Geological borings in the area indicate a quay or revetment of some kind just a little further north of the mouth of the Main Wadi in front of the Khafre Valley Temple. A colossal wall built of stones as large as those in the pyramids extends 200 meters from the south side of the wadi (fig. 2).

Figure 2: Map of the Oriental Institute excavations, with the "Main Wadi" in the lower left.

We know that after the completion of the three pyramid complexes, the area north of the stone boundary wall was a harbor district since the three valley temples of the three pyramids would each have had access to a harbor. It seems that the southern areas of Giza, in the sandy bowl on top of the Maadi Formation (fig. 2, Area B) and the low desert just to the south of the stone wall (Area A), would be the place to look for workers' accommodations.


When the "Giza Plateau Mapping Project" was launched for the 1988-89 season from Yale graduate school, a collaboration began with Dr. Zahi Hawass, General Director of Giza and Saqqara for the Egyptian Antiquities Organization (EAO). This collaboration and the support of David Koch, Bruce Ludwig, and the William K. and Marilyn M. Simpson Endowment for Egyptology made it possible to investigate ideas about the settlement and ancient logistical support of pyramid building.

At the very beginning of the 1988-89 season, a walking survey showed that we could forget about domestic architecture in Area B since clay-like marly limestone (tafla ) could be exposed by simply scraping the loose sand in the bottom of this sandy bowl with the edge of one's shoe. It looked like the bowl may have been widened by quarrying for tafla , a major ingredient in construction ramps, roads, and embankments at Giza, and in the architecture that we found in Area A. The huge stratified dump of settlement debris --- mudbricks, fishhooks, flints, ivories, ashes, sherds, and bones --- that an Austrian team under Karl Kromer excavated in the early 1970s in Area B (fig. 2) might indicate that a settlement existed here at one time, which was probably razed and dumped in the northeast corner of the bowl so that the builders could extract yet more tafla . With little hope of finding domestic architecture in Area B, we divided the 1988-89 team in two crews, sending one crew to work in Area C, at the so-called Workmen's Barracks located to the west of the Khafre Pyramid (fig. 2), and the other crew to work in Area A, where the unique find of two bakeries was made.


Located in Area C is a great rectangular enclosure, formed by walls of broken limestone, tafla , and alluvial mud, measuring 400 meters north-south and 80 meters east-west. There are about 100 galleries, 30 meters long, arranged in a comb-like pattern formed by walls attached to the western and northern walls of the enclosure. Petrie dug two of the galleries and concluded that these were the barracks of workmen, an identification that is still recognized. Our excavations, however, have provided little evidence that people lived there. Most of the galleries had been stripped clean in antiquity, but what evidence was left indicated only craft and storage activities.


During five short weeks of excavation in 1988-89, we opened five 5 x 5 meter squares (A1-6, later designated AA in the fall season of 1991) about 250 meters to the south of a large stone wall and close to the sandy eastern slope of the Maadi Formation (fig. 3). Here we revealed the ruins of a building that was certainly not residential. It is rectangular, 9 meters long (north-south) and 6 meters wide (east-west). Its walls and floor were plastered carefully with tafla . A central wall divides the building down its axis, and on either side there is a series of low rectangular pedestals, about 50 to 70 centimeters in width and 120 centimeters in length.

Figure 3: Topographic map of Area A.

In 1991 the Giza Plateau Mapping Project had two seasons of survey and excavation, the spring season lasted from May 8 until June 18, and the fall-winter season from October 15 until December 13. Our work was made possible once again through the interest and generosity of David Koch and Bruce Ludwig, with additional funds from The Oriental Institute. Fiona Baker with Sheldon Gosline (spring) and Peter Piccione (fall) served as square supervisors.

In the spring season of 1991, Area A was already the primary focus of our interest. Since we backfill the areas we excavated at the end of each season, we were able to reopen the squares dug in 1988-89 and continue our investigation of this building. I had thought that this building in AA was a granary, but our latest investigations cast serious doubt on this early interpretation. There is now good evidence that small square compartments (as opposed to round grain silos) stood on the pedestals, albeit over the spaces between the pedestals. Several seal impressions retrieved from the alleyway between this building and a partially excavated alluvial mud building to the east mention the w 'bt of Menkaure. W'bt , derived from the root denoting "pure" in Egyptian, is a term for "embalming workshop." A w 'bt can include workshops with metal workers, joiners, painters, and draughtsmen. W'bt may refer in a more general sense to the royal administrative unit responsible for equipping the grave, including storage of offerings and craft production for funerary products.

The spring season of 1991 marked the return to excavations at Giza after a hiatus of two years. In the interim, Zahi Hawass had made many discoveries. Construction work for a sewage system for the nearby village of Nazlet es-Samman uncovered evidence of the causeway of the Khufu Pyramid as well as basalt slabs that may belong to the Khufu Valley Temple. The EAO supervised borings throughout Nazlet es-Samman and monitored a continuous trench along the Mansouriyah Canal, which runs through the center of the town. This work indicates that Old Kingdom material, probably from the context of a settlement, is very wide spread under the modern town. The EAO team began excavations in Area A immediately upslope from squares A1-6, where we found the pedestal building, and began to clear a series of unusual tombs, perhaps belonging to Old Kingdom workmen, in mudbrick and stone rubble.

In the fall-winter season of 1991, under the supervision of Augusta McMahon, we opened up a new area, designated A8 (not drawn in fig. 3), at the bottom of a sandy crater encircled by excavation dumps from the 1930s, more recently dumped sand, and horse stable cleanings. The top of the crater gives a panoramic view of the royal pyramid precinct to the north (fig. 4). Our first trench was placed against the wall within contour loop 19 (near the top of fig. 3).

Augusta patiently excavated to the base of the wall through a massive deposit of limestone chip construction debris. The wall is 10 meters in height, and the large gateway at the center of the wall, through which we would arrive at work each morning, is about 7 meters in height. We began to see that the wall was too massive for a simple functional separation of sacred from workmen's areas. Such a massive gate is certainly a statement of royal power; and there was probably a "way" leading to and from it.

Earlier, during our spring season, the EAO inspectors pointed out that a backhoe had gouged a hole, about 5 x 11 meters, at a spot 135 meters to the northeast of Area AA and 135 meters to the southeast of the large stone wall. We designated this area A7 (fig. 3). When we cleaned out the backhoe trench and peeled the sand back in a 15 x 20 meter square, we could immediately see a series of wall foundations composed of stone rubble in a compact surface. The walls, which are nicely oriented north-south, formed about a dozen rooms with doorways and living floors. The surfaces outside the rooms were built up from concentrated midden deposit consisting of so many bread mold sherds that we could call this a "breadmold-sherd gravel." In the fall-winter season (1991) we excavated a number of "rooms" within A7 and uncovered two Old Kingdom bakeries.

Figure 4: View of the pyramids from Area A8.

Michael Chazan was the general supervisor for work in A7 and lead the crew with a pedagogical approach that included daily morning briefings focused on the stratigraphy (fig. 5). In the spring season we had found it curious that the stone rubble walls exposed in the section created by the backhoe were scarcely 20 centimeters deep. It also seemed that the massive mudbrick building from which the backhoe had taken a chomp (fig. 6) was an older architectural phase because it was founded on a deeper level. When we began excavation in the two rooms designated A7d (supervised by John Nolan) and A7e (supervised by Ann Foster), we found that the stone rubble walls were deeper than those in the backhoe section, and the rooms they describe were probably in use at the same time as the large mudbrick building. Each of these rooms measures about 5.25 meters ( 10 cubits) in length and 2.50 to 2.60 meters (5 cubits) in width. Like the other architecture we have exposed so far in Area A, the walls are fairly well oriented north-south and east-west.

Figure 5: Michael Chazan briefs his crew in Area A7.

Figure 6: Stone rubble wall foundations, Area A7. (Plan by Diane Kerns)

A comparatively thin layer of mudbrick and stone rubble debris filled rooms A7d and A7e flush to the walls. In both rooms, as this layer was removed, the first feature to come to light was a large cache of typical Old Kingdom bread molds, the large bell shaped pots (fig. 7) that are shown in tomb scenes and with figurines for bread baking. They vary in size but can weigh as much as 12 kilograms. Referred to as bedja in tomb scenes, these pots have thick walls and can be flat-bottomed, but at our site they most often have a mass of clay forming a bulbous exterior bottom. Bedja always have a smooth, regular, conical interior.

Figure 7: A bedja , one of the bell-shaped breadmaking molds. (Photo courtesy of John Broughton)

Both rooms A7d and A7e have a hearth in the southeast comer, formed against an accretion to the walls of alluvial mud. These mud extensions of the walls had been burnt like fired brick or ceramic. The hearths are entirely open to the room. The hearth in room A7e had an upside down bedja as a kind of corner post. The platform of the hearths were formed of limestone slabs and bricks of calcareous desert clay (tafla or marl). As we excavated further we found the rims of large vats, about 56 centimeters in diameter, in the northwest corner of the rooms. The vats, two in room A7e and three in room A7d, were situated within the deposit that filled the room, a fine black homogenous ash that we took to calling "black velvet." Most of the bread molds rested upon this layer of ash.

In figure 8, room A7d is rendered before excavation into the ash layer, and room A7e is rendered after the ash layer was excavated down to the marl floor. Room A7e had a cache of bread molds on the west side of the room, close to the entrance, as in room A7d. Room A7d contained more of the round flat bread trays which may have been used for baking the flat bread called psn in the hieroglyphic texts.

Figure 8: Area A7, showing A7d before excavation into the ash layer, and Area A7e after excavation through the ash layer to the marl floor. (Plan by Ann Foster, John Nolan, and Mark Lehner)

As we excavated the "black velvet," we wondered where the actual baking area might be. Before we reached the original floor in either room, the answer came from A8, where Augusta McMahon was excavating a much more denuded bakery within several meters of the large stone wall. Under a cake of dry gray ash, she found eggcarton-like rows of depressions --- receptacles in which the dough-filled pots could complete the baking. When we came to the original marl floor in room A7e, we found a similar feature along the east wall of the bakery.

Here the hot pots were placed, probably handled with sticks, and filled with dough that was dipped from the nearby vats. They were covered with another pot, and hot ashes and embers were raked over the baking pits. The dough rose and was baked into the large conical loaves shown in the offering scenes.

This production activity must not have been very pleasant. The room filled with ash, homogenized from being turned over in the baking pits, until the ash reached the very brim of the vats. As we excavated the "black velvet" in room A7d, we noted very thin marl lines, indicating that the bakers had sprinkled the floor with desert clay and wetted it to hold down the fine ash. The ash under the hearth in room A7e, however, was not homogenous, but showed reddish and gray lenses, indicating an atmosphere of higher oxidation. The bakers built higher and higher hearth platforms as the ash filled the room. In figure 8, the hearth belongs to the highest (latest) floor layer, which was left standing in the corner during our excavation down to the original floor.

As soon as we found the bread molds and the vats, it was clear that these rooms were the archaeological equivalent of the Old Kingdom tomb scenes and figurines that show the baking of bread in pots (fig. 9).

Figure 9: Old Kingdom tomb scane, showing bread making activity with conical bread molds. (Georg Steindorf, Das Grab des Ti , Leipzig, 1913)

The vats in the corners must have been used for the dough, which is shown as poured liquid in the tomb scenes. The Oriental Institute has on display in its museum a fine collection of limestone figurines from the Tomb of Ny-kau-inpw that depict some of processes that took place in bread making. One of the figurines is of a woman dipping into a vat that is exactly like those in the Giza bakeries (fig. 10).

Figure 10: Tomb figurines engaged in bread making. (Courtesy of The Oriental Institute Museum. Currently on display)

The bedja were stack healed, sometimes supported on two upside down bedja (fig. 9). This was done on the open hearths in our bakery rooms; in room A7e a single bedja was split and the major half was positioned upside down, forming part of the architecture of the hearth. There is some disagreement on the purpose of the stack heating. One idea is that the massive walls of the pots retain enough of the heat to act as miniature ovens in their own right. Another idea is that the pots, stacked so that their interiors are pointed down toward the open fire, are being "tempered" to create a non-stick surface.

If stack heating serves only to temper the molds, then perhaps like pot baked bread in England in the last century, the baking was completed by putting the dough-filled pots back into the hot ashes and coals, with another pot upside down as a cover. Inside the bread would rise and bake, with a nice crust and moist crumb.

Wilma Wetterstrom, our paleobotanist, reports that the floral remains indicate that the dough must have been made from emmer or barley. It is also her preliminary impression that most of the ash is from acacia, and if so the wood must have been provisioned to the site on a large scale, considering that we now have evidence for this kind of bread baking at places spanning some 300 meters in Area A (evidence of pot-baked bread baking was also found in AA during the 1988-89 season).


In our excavations at Giza we now have opened up three areas in the tract of low desert to the south of the large stone boundary wall in Area A and have found impressive evidence of storage and production. The structure in Area AA might be part of an Old Kingdom w 'bt . The complex in A7 might be part of a pr sn' , a labor establishment associated with food and provisioning. Bakeries are called pr sn' in the Fifth Dynasty Tomb of Ty (from which I excerpted fig. 9). In the tomb scenes, bakeries are part of a larger establishment that includes grain silos and beer brewing. Bakeries and breweries were part of the same production house in ancient Egypt because lightly baked bread dough was used in the mash for the beer, and it is possible that some beer went back into the dough.

It may be natural that Area A was devoted to large scale storage and production, if the harbor district, where raw materials were delivered, was just to the north on the other side of the large wall. So far, we have not uncovered a single structure that is clearly domestic; rather, we seem to be excavating Old Kingdom royal institutions. It may be that the major domestic part of the settlement lies under the modern town, which makes the work of the Egyptian Antiquities Organization in this area of vital importance.

Further excavations and study are required if we are to understand the relationship between the complexes thus far excavated and the people who used them. Were these complexes first established for the pyramid builders and later used by the temple community? And if so, what changes can be detected when one group replaced the other?

Revised: July 30, 2007

Home > Research > Publications > News & Notes