Towns and Houses of the Middle Kingdom

A small number of sites, especially from the Middle and New Kingdoms, provide us with extraordinary amounts of evidence for the settlement archaeology of ancient Egypt. This evidence includes both the detailed archaeological recovery of the architecture of the settlement itself, of the houses within it, their contents and, if we are very lucky, textual material that describes the lives of their inhabitants. In the two following sections we shall consider this evidence.

Kahun – ‘Senwosret is Satisfied’

The most important surviving settlement of the Middle Kingdom is the town that was probably called Hetep-Senwosret (‘King Senwosret is satisfied/at peace’). The name of the town reflects both its function and its founder – it was part of the pyramid complex built by King Senwosret II near the mouth of the Faiyum at a site now known as Kahun (after Petrie’s original usage to distinguish it from the nearby pyramid site, which he designated Lahun/Illahun). The town was located immediately adjacent to the now-destroyed Valley Building of the pyramid complex; the pyramid itself is over 1 kilometre (⅔ miles) further to the west. It is clear from this positioning, and from documents recovered from Kahun, that at least part of the town was designed to be the residence of the priests who served the mortuary cult of Senwosret II. However, the size of the town makes it likely that it was a significant settlement, which had functions beyond that of being an elaborate Middle Kingdom version of the pyramid temples of the Old Kingdom. Documentary evidence from the site indicates that an office of the vizier was located here, and that at least some of the building work on the later pyramid of Amenemhat III at nearby Hawara was directed from Kahun. The town was occupied until the 13th Dynasty – at least 100 years after its foundation.

Discovering Kahun

Kahun was principally excavated by Flinders Petrie in two major seasons of work, from April to May 1889 and October 1889 to January 1890. He ‘cleared’ a large proportion of the town with a speed quite astonishing by today’s standards, but his excavation reports and plan of the site remain the most substantial body of evidence for Middle Kingdom towns, their houses and the contents of those houses.

The Layout of Kahun

Kahun was centrally planned on the same sort of orthogonal layout that is typical of known royal projects of the Middle Kingdom. Its external wall was roughly square, 384 by 335 m (1,260 by 1,100 ft). Today it is located at the interface between desert and cultivated land; as a result a substantial portion of the south and southeastern parts of the town have been lost.

Texts and Toys

Kahun is important not just because of its architectural remains, but also because of the wealth of small finds recovered by Petrie. These include the most important set of Middle Kingdom papyri, which cover a wide range of topics from the administration of local temples as well as personal letters and medical texts. From these sources we know that Kahun was an important seat of regional government and that it possessed a prison – perhaps in the now-lost part of the site. Just as important were the objects of everyday life that were left behind when the town was abandoned at some point in the 13th Dynasty. These also give a strong impression of a real and vibrant community, and include tools, items of personal religion, children’s toys and even a rat-trap.

The Western Quarter

On the western side of the town, a broad strip containing rows of back-to-back houses was divided from the rest of the settlement by a mudbrick wall. The disappearance of the southern part of this locality and the dividing wall here means that it is not clear whether there was any means of communication between this quarter and the rest of the town, nor is it obvious why this strict division was present. Earlier suggestions that this part of the town was used to house a community of non-Egyptians have now been largely discounted, although there clearly were aamu-asiatics present at Kahun, just as there were in many locations in Middle Kingdom Egypt. American Egyptologist David O’Connor has suggested that the Western Quarter might have been designed to house the priests who served in the adjacent Valley Temple of Senwosret II, perhaps on a temporary, part-time basis, while the Eastern Quarter was the ‘real’ town of Kahun, which was home to a permanent resident population, and which had its own religious focus in a temple dedicated to the god Soped. That ‘real’ town may have been partly supported by the farming of the floodplain that was close to the town, and has since encroached on the southeastern sector of Kahun. However, the Western Quarter of Kahun is important for our understanding of domestic architecture of the Middle Kingdom because it contains the period’s largest group of planned and organized ‘small houses’.

The map of Kahun produced by its excavator, Flinders Petrie, together with plans of two very different types of housing, back-to-back terraces (left) and large mansions (right). Steven Snape after Petrie.

Back-to-back Terraces

Most of the inhabitants of Kahun lived in small houses, over 200 of which were excavated by Petrie. These houses were made up of small numbers of rooms, many relatively narrow and corridor-like, and roofed with barrel-vaults. These seem to have begun life as individual units with a common ground plan, creating standardized blocks of back-to-back terraces, as part of the centrally planned, highly regularized official origins of Kahun.

Among these small houses a number of variants can be detected. The smallest are no more than 8 by 7.5 m (26 by 24 ft) and this roughly square ground plan is just enough for two main rooms, plus two smaller ones, along with the connecting space between them. A variation of this type is 8.5 by 5.25 m (28 by 17 ft), with the entrance on the broad side, and with only three rooms, all opening from a small entrance hall. The largest version of this type of house has either 7 or 8 rooms, which either opened off a central room (or rooms) or had a more complicated (and perhaps more private) internal arrangement.

However, the plan produced by Petrie shows considerable variation within these houses, as internal walls were added or demolished to adapt to the particular circumstances of the families and individuals who actually lived within them. In some cases, even the rectangular box of the house itself did not limit the inhabitants’ desire for space, as neighbouring houses could be knocked together to form larger dwellings.

The Eastern Quarter

The northern part of the Eastern Quarter is dominated by a series of large (42 by 60 m, or 138 by 197 ft) multi-roomed structures, which all seem to have been built to a similar design. Six of these structures lie side-by-side, immediately to the north of the main east–west street of the town, while three lie on the southern side of the street. These nine buildings have generally been considered to be large residential houses, for the community leaders of Kahun, but they are so extraordinarily bigger than any other non-royal housing from ancient Egypt – apart from the spacious villas of Amarna – that there has been a good deal of scholarly debate about their exact function(s).

Although each of these large Kahun houses has a different level of completeness, owing to varied archaeological survival, they are likely to have had a standard plan, the different elements of which, and their interpretation, can be summarized as:

1    Small rooms at the entrance to the house, immediately after entering the house from the street, which are most likely to be a Porter’s Lodge.

2    A set of large rooms, which have been described as stables, but which would also be suitable for a group activity such as weaving.

3    A set of small, square interconnected rooms, which are generally accepted to be a granary.

4    A series of open spaces close to the granary, which would be suitable for baking, brewing and other food-production activities.

5    A set of rooms close to the granary, which would be a suitable location for an administrative office if close attention was being paid to deposits and withdrawals of grain from the granary.

6    A large central courtyard.

7    Two integrated sets of rooms, which open onto the central courtyard. These have been identified as residential units (each one has an identifiable bedroom, with a raised bed platform), but who exactly occupied these units is not agreed.

A cut-away line drawing of the wooden model from the tomb of Meketre showing a villa and its enclosed garden. Note that the villa itself has been severely ‘compressed’ in order to give emphasis to the garden. Oxford University Press.

The ‘Core House’ at Kahun

It seems most likely that the real heart of this complex is a ‘core house’, whose entrance is behind the north-facing colonnade on the south side of the large courtyard. This is an entrance that is some distance away from the street-entrance to the complex as a whole, and suggests privacy – the workshops and offices can be visited without entering the core house. The core house has enough rooms to provide a comfortable living space for a family, and there is little doubt that this is where the head of the household resided. However, the presence of a second residential unit within the complex is more of a puzzle. Suggestions put forward to explain the presence of this second house include the idea that it was to provide accommodation for servants or a steward. Alternatively, it may have been the main residence of female members of the household, where infant children were brought up. Another suggestion is that it may have been the place where an adult eldest son of the family lived, with his own nuclear family, before taking over the main core house on his father’s death or retirement. This last suggestion is based on the idea that these complexes at Kahun were not designed primarily for comfortable, elite residence, but as administrative and economic institutions with an importance for the community as a whole.

‘Urban Estates’?

Barry Kemp has made the intriguing suggestion that the large houses at Kahun should be considered as ‘urban estates’ whose similarity to Amarna villas is not primarily in their size, but in the complexity of their functions. Just as Amarna villas seem to have had the important role of acting as economic and administrative ‘hubs’ for city districts at Amarna, so the complexes at Kahun had functions well beyond the servicing of the needs of the family (or families?) that lived in them.

Part of the evidence for this is the presence of a large granary in each of the Kahun elite houses where the level of archaeological survival has been sufficient to identify these structures. These granaries could each hold over 300 cubic metres (10,594 cubic feet) of grain, and therefore had a total aggregate capacity of over 2,700 cubic metres (95,350 cubic feet). If each of these granaries were filled to capacity at harvest time, the grain within them is estimated to be enough to feed a population of between 5,000 and 9,000 people – a figure that is close to, or exceeds, most estimates of the total population of Kahun.

A possible reconstruction of the town of Kahun. Oxford University Press.

This suggests that the most important asset of the Kahun community – its grain – was not stored in a centralized granary area, but was distributed within the urban estates of its community leaders. This emphasizes the role of these large houses as administrative centres overseen by community leaders who had immediate control over communal assets. The same is also likely to be true of other centralized production and distribution activities, including weaving.

Wah-Sut – An Enduring Place

Although remarkable, Kahun is not entirely unique. In 1902–3, Canadian archaeologist Charles Currelly began the excavation of a Middle Kingdom town at Abydos; the dig was later continued from 1994 by a team from the University of Pennsylvania led by American archaeologist Josef Wegner. A series of clay sealing impressions from the town gave its name, ‘Enduring are the Places of Khakaure True of Voice in Abydos’, the first part of which is Wah-Sut. Like Kahun, Wah-Sut was a royal foundation whose function was primarily to service a royal mortuary foundation, in this case the mortuary complex of Senwosret III (which some scholars believe is his tomb, and others a dummy ‘cenotaph’ tomb). Also like Kahun, Wah-Sut grew to a size and level of population that went well beyond what was needed for the servicing of even a royal temple.

The Elite Houses

Although the town has not yet been completely excavated, it was at least 4.5–6 hectares (11–15 acres) in area. Most of what has been excavated to date has consisted of an area filled with elite mansions, very reminiscent of the ‘urban estates’ at Kahun, although rather smaller. They were planned in typical orthogonal style, each house 27.5 by 31.5 m (90 by 103 ft), and each part of a block of four houses. The layout of the internal rooms is also remarkably similar to Kahun, consisting of a core house fronted by a colonnaded courtyard and an entrance some distance from the street, a suite identified by the excavators as a ‘secondary residential unit’ and a series of rooms that parallel the Kahun house’s production and storage areas, but on a significantly smaller scale, suggesting that they did not have the same core supply and administration function as the mansions at Kahun.

The administrative gatehouse, behind the mayor’s residence, giving access to the town of Wah-Sut. Josef Wegner.

Plan of the town of Wah-Sut, after the excavations at the site up to 2012. Josef Wegner.

The Mayor’s Residence

That centralized function was probably served by the largest elite dwelling to be discovered at Wah-Sut, the residence of the Mayor, which the excavators called Building A. This was a huge structure – at 82 by 52 m (269 by 171 ft) it was larger than any of the four-house blocks that were built nearby. It consisted of a series of distinct units, most of which were accessed from a large entrance court containing 12 sycamore fig trees in neat 4 by 3 rows. The core residential house stood on a raised platform and had a very impressive entrance consisting of an 8-columned portico 42 metres (138 ft) wide immediately followed by a shallow hall 38-metres (125-ft) long containing 14 columns. Other areas within the Mayor’s residence included storage and production areas (including a large granary) and, once again, a separate residential unit, which was at one time occupied by the ‘King’s Daughter’ Reniseneb, as suggested by many seal impressions naming her.

Ongoing excavation in the area behind the mayor’s residence at Wah-Sut. Josef Wegner.

Although the town of Wah-Sut was a deliberate royal foundation, the titles held by its community leader were similar to those of other ‘normal’ settlements of Middle Kingdom Egypt in that they brought together in one person two offices: Mayor (ḥ3ty-‘, haty-a) of the town of Wah-Sut and Overseer of the Temple (ỉmỉ-r ḥmw-nṯr, imi-r hemu-netjer) of the most important local deity – Senwosret III himself.

Qasr es-Sagha and Avaris

Although Kahun is the best-known example of a town that provided basic, standardized housing for its working population, it is not the only Middle Kingdom site to have produced this type of small house.

Snail Houses at Avaris

Middle Kingdom Avaris in the Eastern Delta was also the result of centralized planning on the orthogonal model and here, too, back-to-back terraces were built. These were in blocks 12 houses wide (i.e. 24 back-to-back houses) separated by streets over 2.5 m (8 ft) wide. In contrast to these relatively wide streets the houses themselves were tiny at only 5 by 5 m (16 by 16 ft). They also, unsurprisingly, contain only a small number of rooms, arranged in a basic, spiral layout, which leads their excavator – Austrian archaeologist Manfred Bietak – to call them ‘snail houses’. However, although the smallest Kahun houses were not much bigger than the Avaris examples, it has been suggested that the latter were so small that they may have been designed for temporary occupation (perhaps for the workforce on a specific project?) rather than for permanent residence.

Quarrymen at Qasr es-Sagha

The site of Qasr es-Sagha is located in the desert north of the Faiyum. It is one of a number of sites in Egypt’s deserts designed to house miners or quarrymen involved in the acquisition of stones, metals and minerals for the Egyptian state. Most of the settlements set up for these workers were very basic in nature – essentially campsites for short-term, seasonal occupation. Qasr es-Sagha was rather different. For one thing, it possessed a stone temple of strikingly unusual form, but whose very existence suggests an intention for permanent residence rather than occasional camping. The settlement built by the Middle Kingdom state to house their workers in this distant and desolate place also indicates a serious intent for urban development, and once more demonstrates the love of town planners of that period for regularity and the right-angle.

The settlement consisted of a rectangle 114 metres broad and 80 metres deep (374 by 262 ft), orientated north–south and with an enclosure wall. A central street running north–south, with a gateway through the enclosure wall at each end, divided the settlement in half. Although the site has suffered from erosion on its eastern side, it seems likely that the internal parts of the town were a mirror image on either side of the main street. The internal division of the town was into four blocks running north–south; the two central blocks consisted of 10 back-to-back houses, while the two outer blocks consisted of a row of five houses each. The entrances to all these houses therefore looked across the street to another house opposite them.

The 30 houses all seem to have been the same size and layout. Each had an entrance from the street giving access to a court 13 metres wide by 5.25 metres deep (42 by 17 ft). This court was used for a variety of communal, domestic activities, as indicated by the presence of ovens in some of them. At the rear of the court, five doorways gave access to five long, narrow rooms of approximately 8 by 2 metres (26 by 6 ft) each. It is tempting to see these as individual bedrooms. Although it is difficult to be certain, the evidence suggests that each of these houses was home to five individuals, who shared a certain amount of communal activity, perhaps mainly cooking and eating together. Given the nature of the site, it is difficult to think of these households as families, and they are more likely to have been five-man teams within the organization of workgangs at Qasr es-Sagha.

Reconstucted plan of the workers’ town at Qasr es-Sagha, with a detailed plan of one of the individual housing units. Steven Snape.

Big Houses and Small Houses in the Middle Kingdom?

An overview of Middle Kingdom housing strongly suggests that there was a huge disparity between large elite residences and tiny non-elite residences (including those at Avaris and Qasr es-Sagha), with not much in between. Against this it might be argued that the elite houses at Wah-Sut are not nearly so large as those at Kahun, but this is in part because of the way the large Kahun houses (and mayoral residence at Wah-Sut) had such an important role to play in administration, production and distribution at the site. However, even the basic core house at Kahun and Wah-Sut is significantly larger than any of the small houses at Avaris and Qasr es-Sagha (‘small houses’ at Wah-Sut have not yet been discovered and excavated) and this seems to suggest a stark, non-graduated distinction between the haves and have-nots in Middle Kingdom Egypt, which is not what the evidence of cemeteries of this period suggests. Perhaps the important fact is that the towns we have been looking at are all ‘official’ foundations and may have reflected a rather rigid view of society by the state, rather than the reality of social diversity that can be reflected in housing in ‘organic’ towns of the period. This diversity is reflected in the few ‘organic’ towns to have survived from Middle Kingdom Egypt.

Model Houses

The relationship between the living and the dead, their tombs and their houses, will be discussed below, for cemeteries were important and active parts of ancient Egyptian towns. An important aspect of this relationship was the need for the living to provide the dead with food offerings so that the ka-spirits of the deceased ancestors would not go hungry in their tombs. However, although this ongoing relationship between the living and the dead was an ideal, it was soon recognized that tombs, the houses of the dead, needed to be self-sufficient in the generation (using magic if need be) of everything the dead might need, especially food and drink, avoiding a reliance on living relatives to visit the tomb. A by-product of this need for self-sufficiency was the representation of parts of the city within the tomb, which provide us with a picture of urban life.

Little Boxes – Wooden Models of the Middle Kingdom

For elite tombs of the Old Kingdom, these representations took the form of an increasing level of decoration on the walls of the tomb, with scenes of food offering and, later, food production and, later still, the production of other types of goods. In the Middle Kingdom, while elite tombs with decorated walls continued, an increasingly important aspect of provincial cemeteries was the appearance of less-elaborate tombs, usually of the shaft-and-chamber type, which belonged to people – doctors, professional soldiers, middle-ranking administrators – whom we might call middle class. These tombs had little in the way of elaborate superstructures or decorated walls, but the burial chamber was usually big enough to contain both a wooden coffin and a selection of objects.

Some of these objects were wooden models that were more modest alternatives to extensive scenes painted on rock-cut tombs. Their basic repertoire included groups of little wooden figures carrying out everyday tasks: brewers and bakers are especially common, as are butchers and cow-herders, textile workers and potters. Not only are the figures themselves shown going about their daily business (as servants for the benefit of the tombowner), but they are shown with the equipment needed to carry out their occupations, and often also the premises in which they worked. Granaries are often shown, with little figures carrying sacks of grain and little wooden scribes shown recording this activity.

These models also included boats, which had a religious symbolism – allowing the deceased to journey to and from the pilgrimage town of Abydos (home of the most important cult centre of Osiris, lord of the afterlife) – but which also emphasize river travel as the prime means of long-distance communication in dynastic Egypt.

It is unclear to what extent most of the owners of these models were employers of servants when alive, but some clearly were. The most famous set of wooden models from the Middle Kingdom come from the Theban tomb of the administrator Meketre. These include modelled scenes of boats fishing with nets, and scenes of the counting of cattle, but most are standardized wooden boxes each containing a workshop of a particular activity, including baking and brewing, spinning and weaving, butchery and carpentry. Another of these boxes represents a much-abbreviated residential house with a more lovingly modelled colonnade and tree-filled courtyard. It is tempting to see in the Meketre models the component parts of an ‘urban estate’ much like those at Kahun.

Soul Houses

When Petrie was excavating the Middle Kingdom cemetery of Deir Rifeh in Upper Egypt in 1907, he recovered over 150 examples of a class of object that had only occasionally been found before. These were elaborate ceramic trays, designed to receive offerings for the dead. They were placed above ground next to fairly simple graves. Because these trays had been modelled to give them a house-like appearance, Petrie called them ‘soul houses’ and the name has stuck. The extent to which they actually resemble contemporary houses is a matter of debate. On the one hand, they needed to serve as trays in which the offering could be placed, but they do also seem to have been poorer substitutes for the large offering chapels of elite tombs of the same period, which certainly had house-like aspects to them. The tomb was, after all, the house of the ka. The need to produce a functional ceramic tray encouraged the depiction of external aspects of the house, particularly any courtyard or open area in front of it, while the internal aspects of the house would tend to be ignored or simplified. The emphasis on an external courtyard would be especially relevant if this was where food preparation generally took place, as the ‘soul houses’ themselves were especially concerned with the provision of food.

Models of granaries are often found among the range of wooden models from Middle Kingdom tombs – they were to be a source of grain for the tombowner. They might also be shown having the grain turned into flour. Left, British Museum, London. Right, The Metropolitan Museum of Art, New York/Art Resource/Scala, Florence

The surviving ‘soul houses’ show different levels of attempts to emulate houses. Some emphasize the courtyard, with perhaps a sunshade supported by a set of pillars. Others depict the front of the house, especially the columned portico that also provided a shaded area at the front of the house. Some seem to indicate that houses could have two storeys, while the roofs of many ‘soul houses’ are provided with a mulqaf, a hood-like structure over a hole in the roof designed to channel any cool breezes down into the house. Some examples move away from the open tray format and look much more like real houses, with walls, windows and a door, although the need to make these ‘soul houses’ capable of carrying offerings means that the roof is in these cases removed.

Although the value of ‘soul houses’ as accurate depictions of Middle Kingdom domestic housing might be questioned, they are extremely useful in depicting features that were, presumably, common in ‘ordinary’ Middle Kingdom houses (temporary shaded structures, mulquafs), but that do not survive in the archaeological remains of the lower parts of largely ‘official’ constructions of the period such as at Kahun and Wah-Sut.

Models of houses found within Middle Kingdom tombs included large, detailed wooden models like that from the tomb of Meketre. Left, The Metropolitan Museum of Art, New York/Art Resource/Scala, Florence. Right, British Museum, London

If you find an error or have any questions, please email us at Thank you!