Al-Shaykh Sa’id Revisited: A Reassessment of the Spatial Layout of a Monastic Community

Al-Shaykh Sa‘id (Middle Egypt) covers the southern part of the archaeo­logical concession area of the Dayr al-Barsha Project (Research Group Egyptology, KU Leuven),[1] directed by Harco Willems.[2] The site is named after the nearby tomb of a local saint. An ensemble of Old Kingdom tombs (end of Fifth —the beginning of Sixth Dynasty) was cut into the steep western cliff face of an outcrop between two wadis, Wadi Jamus to the north and Wadi Zabayda to the south (figs. 24.1—3).

Monumental tombs fine the principal ridge, with a lower and a higher level of smaller tombs and . Steep paths connect the long narrow terraces. A lower range of jagged cliffs and rifts separates the tomb complex from the floodplain.

Fig. 24.1. View of the range of Old Kingdom tombs. © C.J.M. van Loon 2013.
Fig. 24.1. View of the range of Old Kingdom tombs. © C.J.M. van Loon 2013.
Fig. 24.2. Map of al-Shaykh Said.
Fig. 24.2. Map of al-Shaykh Said.
Fig. 24.3. Map of the greater Dayr al-Barsha region. Courtesy Chr. Peeters, Dayr al-Barsha Project.
Fig. 24.3. Map of the greater Dayr al-Barsha region. Courtesy Chr. , Dayr al-Barsha Project.

At the end of the nineteenth century, Norman de Garis Davies stud­ied and documented the tombs (Davies 1901). They have received little attention since,[3] but the Supreme Council of Antiquities restored the most important tombs in 2002-2003. Davies was one of the few late nineteenth­century who also documented Christian remains.

A number of tombs show traces of reuse as , and, on the terraces in front, wall remains testify to the construction of walled courtyards and additional buildings. Tomb 25, the tomb of Werirni,[4] was remodeled as a church. Inscriptions found in some of the tombs confirm that a monastic com­munity once inhabited the site (Davies 1901; Timm 1984-92, vol. 5: 2327; Jones 1991:141-42; van Loon 2010; van Loon and Delattre 2014, with an appendix on the inscriptions by A. Delattre).

Only a handful of early travelers and scholars seem to have visited the tombs. The diplomat and antiquarian William Richard Hamilton stopped at the site in the of 1801-1802. On the tomb of Werirni, he wrote:

On the inner walls of one of the catacombs, we found some curious sculptures and paintings. It had long served as a Greek or Coptic church, either for the inhabitants of the town, or for some devout hermits who in the early ages of Christianity may have dwelt in these caverns. Whoever they were, they have given vent to their religious fury against the idols of Paganism, by mutilating all the prominent figures and by concealing the paintings under a thick coat of plaister. (Hamilton 1809: 275-76)[5]

Sir John Gardner Wilkinson visited al-Shaykh Sa‘id in 1824—25. In his famous guidebook, Modern Egypt and Thebes, he describes pharaonic reliefs in some of the tombs and adds, “Before several of the grottoes are crude brick walls, built when inhabited by the Christians, who converted one of them into a church, cutting a circular niche into the rock opposite the entrance” (Wilkinson 1843, vol. 2: 70). The Coptic inscription in Tomb 6, a series of names of monks that he copied in his notebook, was not added (Oxford, New Bodleian Library, Ms.Wilkinson dep. C 5 fol. 4, unpublished;[6] see Davies 1901: 35; edited by Delattre in van Loon and Delattre 2014).

The French Egyptologist and artist Nestor l’Hote was in al-Shaykh Sa‘id in 1838. He was the first to call this settlement “Deyr-abou-fam” (l’Hote 1840: 51; see Davies 1901: 10; Coquin and 1991f;Timm 1984—92, vol. 2: 572; al-Suriany 1990: 115; Jones 1991: 141). His unpub­lished notebook contains a sketch of the niche in the “Tombeau dont les premiers chretiens ont fait une eglise,” which he dated to the third-fourth century (, BnF, NAF 20396: 273 verso).[7]

When exactly these tombs were reused by monks is not known. So far, no historical documentation of a monastic settlement at al-Shaykh Sa‘id has been found. Dates of habitation vary. Petrie dated fragments of glass and ceramics found near Tombs 61—62 to the fourth century (Davies 1901: 4).

Jones dated ceramics to the period of the fifth to the eighth centuries (1991: 143). The themes of the paintings in the church have been known since the fourth to fifth centuries, but cannot be narrowed down to a more specific period of time (Van Loon 2010:171—75). One of the texts in Tomb 39 contains a fragmentary date, pointing to the (Davies 1901: 36; Delattre in van Loon and Delattre 2014).

One remarkable feature of this site is the remains of a large wall at the foot of the hills, to the southwest of the group of tombs. On satellite imagery, a stretch of about 270 meters of this mud-brick wall running parallel to the Nile can still be traced in the desert environment.To the north, it disappears below cultivation. Part of it is still standing (approximately 6.50 m long, 1.6 m thick, and 3.50-4 m high; figs. 24.2 and 24.4).[8]

This enigmatic wall had already intrigued Hamilton (1809:275), Edme Francois Jomard, who wrote about it in the Description de I’Egypte (1822, vol. 4: “Description de rHeptanomide,” 322-24), and Wilkinson (1843, vol. 2: 70; Oxford, New Bodleian Library, Ms. Wilkinson dep. C 5 fol. 4, unpublished). Ideas about the age and function of this once huge wall differ, but one of the theories is that it might have been part of a monastic valley settlement that was related to the community living in the tombs (Kemp 2005: 37—38, 54).

Fig. 24.4. The wail at the foot of the tomb complex, al-Shaykh Sa'id. © G.J.M. van Loon 2013.
Fig. 24.4. The wail at the foot of the tomb complex, al-Shaykh Sa’id. © G.J.M. van Loon 2013.

In all probability, the monks who settled in the tombs of al-Shaykh Sa‘id adhered to a semi-anchoritic lifestyle. They had chosen to live on their own or share a dwelling with a disciple. The community would meet once a week for mass and a communal meal. To supply their everyday needs, there might have been other services such as a bakery or kitchens (Wipszycka 2009: 288-90).

The degree of organization in such a group varied. The surroundings of their chosen location, pre-existing spaces such as tombs or, in other places, quarries or natural caves, partly dictated building activities and subsequently the size of the community and how the monks organized their Eves. The interior could, within boundar­ies, be adapted to their wishes and, if space were available, additional rooms or buildings could be constructed or a courtyard could be fenced off with a wall (van Loon and Delattre 2014; van Loon and De Laet, 2014).

This chapter explores the way monks adapted the tombs of al-Shaykh Sa‘id to a monastic community as well as the role of the wall in front of the cliffs. Was it part of the layout of the settlement or did it serve another purpose?

Living at al-Shaykh Sa‘id

In order to be able to analyze the settlement pattern, a distribution map was constructed, showing the exact location of the tombs and the road system. For each tomb, markers of habitation (budding activities [interior or exterior], decoration, and inscriptions) were noted. Using this sort of map, the diffusion of reused tombs and spaces that had a specific function began to emerge. Paths were revealed to point to particular areas, showing the internal structure of the settlement as well as Enks to the outside world.

The area was mapped using very high spatial resolution satellite imag­ery (Quickbird-2) combined with the Davies published in 1901 (pls. I—II). Remote sensing implies a spatial context of the features analyzed; a basic map can be created that serves as a starting point for field survey.

A field survey (carried out in 2012—13) included taking GPS readings of the entrances of the tombs (using a Trimble® GeoExplorer® 2005 GeoXT™ handheld GPS) and documenting characteristics related to their reuse. Paths were examined and their context recorded.

The features of each tomb were set out in Microsoft Excel and, in conjunction with the GPS readings, imported into the ESRI/ArcGIS® module ArcMap. If GPS readings for tombs were absent (mainly because of the falling away of the satellite signal caused by the steep rock faces), points were digitized in ArcGIS on the basis of photographic documentation and comparison with Davies’ maps. Data analysis was performed using selec­tion queries as implemented in the ArcGIS software.

Three general objective characteristics,‘building activities,’ ‘decoration,’ and ‘inscriptions,’ were chosen as markers for habitation. These are aU fea­tures that can be observed, as none of the tombs or the spaces in front of the tombs were excavated. Evidence for ‘budding activities’ includes, for example, stone or mud-brick walls, windows, doorways, floors, cup­boards, benches, and niches. ‘Decoration’ applies to painting and sculpture. ‘Inscriptions’ merely indicates the presence or absence of inscriptions and graffiti, not language or content.

In total, Davies numbered 102 tombs, located on the main ridge and on terraces above and below this, plus a few tombs scattered in between these levels. Some of these tombs are currently inaccessible. Examples are Tomb 18, located high up in a virtually perpendicular cliff face, or Tombs 94—102, which have now almost completely filled up with sand and debris. Moreover, Davies counted, indiscriminately, monumental tombs, small tombs, low tombs of approximately 1 cubic meter, with or without a shaft, mere niches, and ‘pit-tombs,’ single square shafts with a low vault in one of its sides.

It is clear that the latter three types were not suitable for habitation, although the low tombs could be incorporated in a complex, for example, as storerooms. All tombs were mapped and subsequently those tombs qualifying as a dwelling place were filtered out. Of the 102 tombs counted by Davies, only twenty-five were large and high enough for this purpose. Further analysis was performed in the latter group.[9]

A glance at the distribution map (fig. 24.5) makes it clear that the best possible use was made of the existing situation. The main ridge of the natu­ral outcrop houses monumental tombs, which were remodeled. Dividing walls between small tombs could be dismantled in order to create an accept­able space, as in Tombs 70—71, in which a doorway was cut and windows, seats, and steps were carved (Davies 1901: 8-9).

Another way to connect tombs was to build a communal courtyard, as in front of Tombs 61 and 62 (61 is low and very small). Davies states that the floor of the courtyard was plastered and contained two receptacles, sunk in the floor (Davies 1901: 4 and 8, pl. XVII [“Tombs arranged as a dwelling”]; van Loon 2010: 163).

The mud-brick or stone structures in front of the tombs show that full use was made of the space on the terraces. These terraces are currently very narrow and often steep, but the building remains (for example, near Tombs 37 and 39) indicate that originally these ridges must have been wider in some spots. A relatively wide and flat area in front of Tombs 24-28, where Davies excavated a small building in front of the church (Davies 1901: 4, pl. XVII [“Brick house outside tomb of Urarna”]; Jones 1991: 141-42, pl. XII-1), and extending to the southwest near Tombs 29—30, shows a concentration of mud-brick remains.

These constructions were built of bricks of various sizes laid in different bonds. Further analysis of the bricks and the scat­ter of sherds among the wall remains (planned for the mission of 2014) will give more insight into the date and function of these buildings.

Fig. 24.5. The settlement pattern at al-Shaykh Sa'id.
Fig. 24.5. The settlement pattern at al-Shaykh Sa’id.

However, it is clear that this pottery belongs roughly to either the Late Roman or the period.

Only Tomb 25, the tomb of Werirni, in which a church was installed, could be assigned a special function. The original intermediate wall divid­ing the present chamber into two was removed. A large niche was cut in the southeast wall, flanked by pilasters with capitals. The conch is painted with a winged creature and is rimmed with a white decorative border on a dark background.

To the right of the niche, fragments of a military saint (St. Phoibammon/Abu Fam?) can be distinguished, and on the southwest wall are the fragments of a unicorn, an ankh cross, and a species of deer (Davies 1901:16-17, pl.VII, XIII-B, XIV; van Loon 2010:164-75). The layout and decoration of the eastern part of the church featuring such a spacious apsidal niche fits into a series of churches installed in tombs and quarries in this region. Similar niches can be found in Amarna (Tomb of Panehsy; Davies 1905: 9, 11-13; Pyke 2008), Dayr al-Dik (Quarry 13; Martin 1971: 27-33; van Loon 2008: 182-90) and Speos Artemidos (Martin 1971: 62; van Loon, in press).

The most elaborate interior can be found in Tomb 39. It seems highly likely that this tomb was never finished or decorated. It contains a large number of wall niches (four large ones hold beds), which were painted with geometrical designs in red, black, and yellow (fig. 24.6; Davies 1901: 36, pl. XXXIII; van Loon 2010: 163—64). Apart from the church and this tomb, the decoration is sparse.

Pilasters flank a small niche in Tomb 37 and Davies noted a painted decorative pattern in Tomb 18 (Davies 1901: 31). The entrance to Tomb 30, which is low and shows no further adaptation, is flanked by engraved crosses (Davies 1901: 7). Only Tombs 6,19,25, and 39 preserve inscriptions, mainly monks’ names and a dated dedication in Tomb 39 (twelfth century; Davies 1901:16-17, 29, 35-36, pl. XIV-B/D; see van Loon 2010:176; Delattre, in Van Loon, in press a).

The infrastructure of paths in the tomb complex is the result of fre­quent use following a least-cost trajectory (fig. 24.5). None of the paths show built, paved, or cut sections. Consequently, they are almost impossible to date. The north and south paths leading from the valley, which ascend to the main ridge, were most probably the old principal entrance routes. Likewise, the paths connecting the terraces must have already existed during the time this area was used as a cemetery. Avoiding the sharp eleva­tions and deep rifts, there are some walking paths leading up directly from the floodplain to the main ridge, ending in front of Tombs 1-3 and Tombs 24-33. It is hard to say to which period the latter paths date.

Although they are steeper than the northern and southern routes, they do lead directly to the monumental tombs and they might have been in use since the Old Kingdom. These connections are all located between the floodplain and tomb complex and within the complex itself, with one exception: from the main and upper ridge, a path runs in a southeasterly direction and ends in the Wadi Zabayda. This wadi runs southeastward for about 5.5 kilometers and ends among the North Tombs of Amarna, the site of a contempora­neous Christian settlement, inhabited from about the fourth to the ninth century (Jones 1991; Kemp 2005).

Fig. 24.6. Al-Shaykh Sa'id, Tomb 39, view to the southwest, with three niches holding beds. © G.J.M. van Loon 2008.
Fig. 24.6. Al-Shaykh Sa’id, Tomb 39, view to the southwest, with three niches holding beds. © G.J.M. van Loon 2008.

At al-Shaykh Sa‘id, a small compact monastic settlement was developed. Tomb 25 was probably chosen as a church because it is located in the center of the site with easy access on all sides, including the area occupied by mud-brick structures that extends to the southwest near Tombs 29—33. So far it has been impossible to identify workshops or rooms set aside for specific use.

Barry Kemp (2005: 37) has estimated that this group of dwellings could have housed approximately thirty monks at any one time, and that is, in my opinion, a reasonable estimate. His theory of a group of hermitages dependent on a mother community in the plain, of which the wall fragment formed a part, will be examined below.

The Wall (figs. 24.2 and 24.4)

In the Description de I’Egypte Edme Francois Jomard, an engineer and cartog­rapher in Napoleons retinue of scholars who explored Egypt in 1798, noted:

Au-dessous de point, il y a encore d’autres carrieres qui annoncent que les Egyptiens ont faits des grands travaux la montagne. Un long mur egyptien, en briques d’une grande epaisseur, qu’on trouve pres du Nil, et trace parallelement au fleuve, confirme cette idee. Les briques sont enormes, et les habitans disent que c’est l’ouvrage de la plus haute antiquite.

Je crois tres probable, quoique la geographic n’en fasse point mention, qu’il a existe une position ancienne dans cet endroit; il est possible que le Nil, en abandonnant Meylaouy, et se portant a Test, ait detruit les vestiges de cette ancienne ville. (Description de I’Egypte 1822, vol. 4: vol. 4: 323—24)

In the adjoining map of the region a “mur de briques” is indicated. Around 1800, the Nile flowed much closer to the hills than it does nowadays (Jacotin 1876: feuille 13; cf. Kemp 2005: fig. 1.17).

Hamilton describes the construction of the wall. He writes that on this spot are the vestiges of an ancient [sic] Egyptian town, marked out by its wall of bricks baked in the sun, and the catacombs in the cliff. This wall is in some parts entire; but, unlike most other Egyptian walls, the bricks never present outwards their long sides; and over each layer of bricks placed edgeways on their narrow sides, are placed two layers of the same sized bricks on the flat sides. (Hamilton 1809: 275)

Both men thought this remnant was once part of a town wall and that the town itself, located to the west of the wall, had disappeared, swallowed up by the meandering of the Nile.

In his of 1824-25, Wilkinson notes “a long cr.[ude] br.[ick] wall, inclosing perhaps the burying ground, for below the T[om]bs in the rock are these of cr. br. for poorer people.”[10] However, in his later Handbook for Travellers he has this to say about this wall: “At Isbayda [= Zabayda, the name of the southern wadi] there is another portion of the Gisr el Agoos” (Wilkinson 1843: 70—71).

Norman de Garis Davies records that the wall, which could be followed for about 275 meters, had served as a defense against the . He adds: “Such walls are found elsewhere in similar situations to the north and south …, and bear the name Gisr el AgAz, ‘Dike of the Old Man,’ perhaps because the Arabs see how ill-fitted they are to serve as dams. The size of the bricks (10% x 3 x 5 inches (Mr. J. Newberry))[11] may give the clue to the age of the construction” (Davies 1901: 5). Note that he translates elAgAz as ‘old man,’ instead of the more usual ‘old woman.’

The Old Lady’s Dike or Old Lady’s Wall figures in an old and popular tale about the mythical queen Daluka, who reigned in Egypt after and his army drowned in the Red Sea while pursuing Moses and his people. According to tradition, she surrounded Egypt with a wall, either to protect the country or to protect her son.[12]

The isolated (strong) wall remains parallel to the Nile, which had no obvious context, were often identified with Queen Daluka s wall, even though there is no evidence at all that such a wall ever existed (Sijpesteijn 2011 discusses the Arabic narrative tradition but the story has older : H.Willems, personal communication).

As set out above, Kemp proposes that the remains were the outer perimeter wall of a monastic complex whose buildings were destroyed by the meandering of the Nile. This ‘parent community’ would have had satellite hermitages in the converted tombs in the cliffs. He writes that the building method “is not inconsistent with the Late Roman period” (Kemp 2005:37-38,54).

Research by the KU Leuven team confirms that the masonry of the wall fragment at al-Shaykh Sa‘id contains Late Roman ceramics (Willems et al. 2009: 293). The Late Roman date is consistent with the archaeo- botanic analysis of one of its mud bricks (Marinova et al. 2011). In 2007, the team surveyed the area to the west of the wall, which covers about 800—950 meters to the current river bank, but no building remains were found in this area.

To sum up, the wall, which once ran along in front of the cliff of al-Shaykh Sa‘id, parallel to the Nile, can be dated to the Late Roman period (c. third—sixth century). The trace is still about 275 meters long, but its original length is not known. To the west, there are no building remains, and to the east, the nearest constructions (mud-brick in front of the tombs) are located around 240 meters from the wall trace, with some steep low cliffs and a plain strip of desert in between.

As a consequence, Jomard and Hamilton’s idea of an ancient city wall, whether pharaonic or from later times, can be abandoned. Wilkinson’s hypothesis of a wall enclosing a cemetery is doubtful. The only mud-brick constructions are those at the foot of the rock-cut Tombs 25—33. Although these structures require further research, they seem to be of a much later date.

A continuous wall as a defensive structure (Davies) was almost certainly never a reality. Finally, Kemp’s theory, a wall surrounding a monastery, comes up against a similar problem to that of the lost-city theory of Jomard and Hamilton: no building remains to the west of the wall have ever been found.

The idea of a wall delimiting the area belonging to the monastic settle­ment in the cliffs springs to mind. For example, Dayr al-Bala’iza (near Asyut, abandoned in the ninth century), where a church was installed in a quarry and a large monastic complex with several churches built on the steep ascending hill, did have a surrounding wall (Grossmann 1993). However, in contrast to al-Shaykh Sa‘id, Dayr al-Bala’iza was a sizable cenobitic com­munity in which a wall to separate from worldly affairs was a necessity.

Furthermore, physically, the latter wall, like the other known contemporary monastic boundary walls, was a much more modest struc­ture than the wall at al-Shaykh Sa‘id (Grossmann 1993:173—74; Grossmann 2002: 307—309). Defensive walls in (semi-)anachronistic communities are known to have been developed from approximately the ninth century.

They were never built around a complete settlement, but around a nucleus con­taining the most important buildings (Grossmann 2002:309—13). Moreover, such a large and massive wall as in al-Shaykh Sa‘id has not been encountered in monastic architecture so far (P. Grossmann, personal communication).

A parent community in the floodplain with a satellite settlement in the cliffs has long been proposed for monastic habitation patterns on the east bank in Middle Egypt (Martin 1971; Kemp 2005: 54—55, although Kemp recognizes the risk of oversimplification). In my opinion, every site needs to be assessed individually.

For example, recent research on the monastic hill settlement and the village of Dayr Abu Hinnis, about 8.5 kilometers north of al-Shaykh Sa‘id, has led to new insights into the habitation pattern (van Loon and De Laet, 2014). Al-Shaykh Sa‘id, with its own church, most prob­ably functioned as an autonomous community. Situated not far from the river, it had ties with the outside world in the near vicinity in order to obtain necessary provisions as well as for trade, and contact with neighboring com­munities, as the southeastern path in the direction of Amarna suggests. So far, no indication of a corresponding valley settlement has been found.

Therefore, for the time being, the function of the large wall in the floodplain has to remain a mystery. Sadly, it will probably have to remain an unsolved mystery. The wall traces are in imminent danger of destruc­tion. In the hope of attracting tourists to Amarna, the Ministry of Tourism has decided to build a new road suitable for buses along the east bank. For various reasons, the trajectory does not follow the existing dirt road, but cuts through the strip of the desert at the foot of the cliffs. At al-Shaykh Sa‘id, it will run directly along the wall.

In April 2013, the new road had already reached Wadi Jamus, the wadi to the north of the Old Kingdom rock tombs, thereby destroying a First Intermediate Period cemetery that had not yet been investigated. Owners of the land adjacent to the new road had building plans. The road and other developments will probably have reached an advanced state by now, thereby forever shattering the peace and quiet that once attracted the .[13]

Gertrud J.M. van Loon


[1] The current research is part of the author’s EU-funded Marie Curie Fellowship at KU Leuven, “A View from Heaven: Exploring the Potential of Remote Sensing Techniques to Study Spatial Patterns of Monastic Habitation on the East Bank of the Nile in Middle Egypt” (FP7--IEF 254243), and is being carried out within the framework of the Dayr al-Barsha Project (Research Group Egyptology, KU Leuven), which is funded by research grants from FWO Vlaanderen and the Bijzonder Onderzoeksfonds of KU Leuven (

[2] The concession also includes Dayr al-Barsha and Dayr Abu Hinnis, which are together called the “Greater Dayr al-Barsha Region.”

[3] Most of the ground plans of the tombs have been copied from plans made on site by John Newberry in 1892 (Davies 1901: preface, xi—xii). For the history of research before 1900, see Davies 1901:10 (correction: Lepsius, Denkmaler 2,120-23).

[4] Numbering of the tombs according to Davies. Modern orthography prefers Werirni to Urarna or Ouramou as found in older literature.

[5] For this reference, I thank Christoph Peeters, Dayr al-Barsha Project.

[6] I would like to thank Colin Harris (Modern Papers & John Johnson Reading Room) for his friendly welcome and help.

[7] For this reference, I thank Marleen De Meyer, Dayr al-Barsha Project.

[8] Measurements taken in 2008. At that time, it had deteriorated only shghtly since Davies measured it (Davies 1901:5).

[9] Between Tombs 6 and 7, a tomb was used as a quarry. Davies indicated this place as “ruined tombs,’’but did not number the spot (Davies 1901: pl. I).The space has not preserved any traces of reuse for habitation and is not included in the analysis. Tombs 94—102 are not included either as they are now almost completely hidden by sand and debris. Davies (1901: 9) says that 94 and 101 were small chambers, 95—99 were very small and rough, and 100 and 102 were not cleared at that time.

[10] Oxford, New Bodleian Library, Ms.Wilkinson dep. C 5 fol. 4 (unpublished).

[11] The brick size corresponds to the measurements given by Kemp 2005: 37.

[12] In some versions of this story, was the builder of the wall (Sijpesteijn 2011: 90, 99).

[13] Special thanks go to Rosemary Robson-McKillop for correcting my English.


Leave a Reply

Your email address will not be published. Required fields are marked *