Smith, JARCE 28 (1991), Askut and the Purpose of the Second Cataract Forts

Embed Size (px)

DESCRIPTION

Smith

Citation preview

  • Askut and the Role of the Second Cataract FortsAuthor(s): Stuart Tyson SmithSource: Journal of the American Research Center in Egypt, Vol. 28 (1991), pp. 107-132Published by: American Research Center in EgyptStable URL: http://www.jstor.org/stable/40000574Accessed: 02/03/2009 21:27

    Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

    Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=arce.

    Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

    JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with thescholarly community to preserve their work and the materials they rely upon, and to build a common research platform thatpromotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

    American Research Center in Egypt is collaborating with JSTOR to digitize, preserve and extend access toJournal of the American Research Center in Egypt.

    http://www.jstor.org

  • Askut and the Role of the Second Cataract Forts*

    Stuart Tyson Smith

    The fortress of Askut lay on an island in the Batn el-Hajar, just south of the Second Cataract (fig. 1), and was one of a chain of fortresses established in the Middle Kingdom. Until the excavations of the University of California at Los Angeles, led by Alexander Badawy from 1962-64, little was known about the site and its significance. Wheeler mentioned it in passing in 1932, characterizing it as "much destroyed and rebuilt," and noting that it provided a crucial line of sight between Shalfak and Murshid, thus ultimately connecting Semna with the Second Cataract.1 It was, however, only after the Uni- versity of California excavations that Askut' s significance as a major fortress of the Middle Kingdom was realized. The site was, as Wheeler observed, extensively overbuilt, but this condi- tion actually served to protect the earlier de- posits, not destroy them. Remarkably, Askut was relatively free of modern disturbance from looters and sebakheen. The consistently deep cultural deposits reached approximately 1.2 m in the Upper fort, and 2.6 m in the Southeastern Sector outside of the main defense wall (fig. 8).2

    This excellent preservation is in stark contrast to the other Second Cataract forts, which are characterized by a high degree of disturbance and erosion.3 The potential for understanding Askut' s role in the chain of fortresses, and thus

    * The first and much shorter version of this paper was presented at the ARCE meetings in Memphis, April 1987. A substantially similar version was submitted in partial fulfill- ment of the Masters Degree in Archaeology at UCLA. I am grateful for many useful suggestions from Antonio Loprieno, Kent Weeks, Nigel and Helen Strudwick, and Bruce Wil- liams. Any faults which remain are my own. I would also like to thank the Department of Near Eastern Languages and Cultures, Institute of Archaeology and Museum of Cul- tural History at UCLA for their support. Final revisions for this paper were completed when I was recipient of the Museum's Ralph C. Altman Memorial Fellowship. 1 Noel F. Wheeler, "Excavations of the Harvard-Boston Expedition in Haifa Province, 1930-1931," Sudan Notes and Records, vol. XV (1932), 256.

    2 Alexander Badawy, "Askut: An Egyptian Island Fortress of the Middle Kingdom in Upper Nubia." MS on file,

    Museum of Cultural History, University of California at Los Angeles. At this time it is hoped that the Museum will publish the MS. 3 Only the area immediately South of the "Commandant's Quarters" was denuded at Askut, although the entire maga- zine structure was heavily disturbed by later, perhaps Meroi- tic, activity (fig. 8). While there has been some disturbance at Mirgissa, substantial areas were well preserved, including the magazine block and an armory. Still, deposits in a large portion of the interior of the main fort were less than 20 cm deep (or went unexcavated). Most of the interior was pre- served to less than 50 cm (Dows Dunham, Uronarti, Shalfak, Mirgissa. Second Cataract Forts, Vol. II [Boston: Museum of Fine Arts, 1967]; Jean Vercoutter et al., Mirgissa I. Mission Archeologique Frangais au Soudan [Paris, 1970]; and A. Vila, "L'armement de la forteresse de Mirgissa." Rd'E 22 [1970], 171-99). The inner fort of Semna South had been completely denuded, but the peripheral areas and enclosure have yielded much information from the Middle Kingdom. Final assess- ment must await publication, but for a preliminary report see Louis V. Zabkar and Joan J. Zabkar, "Semna South. A Preliminary Report on the 1966-68 Excavations of the Uni- versity of Chicago Oriental Institute Expedition to Sudanese Nubia," JARCE 19 (1982), 7-50. Apart from a section of Semna, the cultural deposits in the rest of the Second Cataract forts (Kumma, Uronarti, and Shalfak) were too heavily denuded or disturbed (or, perhaps, poorly excavated) to arrive at secure stratigraphic contexts (Dows Dunham and J. M. A. Janssen, Semna, Kumma. Second Cataract Forts, Vol. I [Boston: Museum of Fine Arts, I960]; Dunham, op. cit.). Preservation at Buhen was so poor that the excava- tors concluded that context was highly unreliable (Walter B. Emery, H. S. Smith, and Anne Millard, The Fortress of Buhen. Part 1. The Archaeological Report. Egypt Explora- tion Society Memoirs, Vol. 49 [London, 1979], 93-94). Serra East to the north was also heavily denuded prior to being overbuilt by a Christian settlement (J. Knudstad, "Serra East and Dorginarti," Rush 14 [1966], 172ff., fig. 2).

    107

  • 108 JARCE XXVIII (1991)

    Fig. 1. The Second Cataract Forts and Ramesseum Listing.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 109

    of assessing the purpose and functional inter- action of the entire Second Cataract system, is therefore greatly enhanced.

    Askut occupies an unusual position in the chain of fortresses. Semna South, Semna, Kumma, Uronarti, and Shalfak are clustered closely to- gether around the Semna Cataract. Mirgissa, Dabenarti, Kor, and Buhen are similarly focused about the Second Cataract. Askut sits by itself in the midst of these two groupings, almost half way between Semna and Mirgissa (18.9 km and 22 km respectively), closest to Shalfak (9 km). Askut has been characterized as a lonely watch post and signaling station, located in the middle of one of the most barren stretches of the Nile Valley.4

    There is no reason, however, to relegate the site to such a secondary role. Although some- what smaller, Askut compares favorably to Uro- narti and even Semna in size. It is actually larger than Semna South, Kumma, and Shalfak (Table 1). Far from being located on an inhospitable, barren stretch of river, Askut lies at the northern- most end of the wide Saras Plain (fig. 2), which contained twenty-five C-Group sites.5 Another eight C-Group sites were located in association with cultivable land on the East and West Banks north of Askut. Additionally, a large pharaonic cemetery of some 250 graves (ll-L-26) is reported on the island of Kagenarti, which is actually connected to Askut.6

    Of the sites opposite and lying to the north of Askut, there are C-Group settlement sites, five cemeteries, and one of unidentified type. Each of the C-Group sites had A-Group deposits under- lying them, demonstrating the consistent quality of the resources exploited for subsistence there. The largest site (ll-M-15) had a surface scatter of objects over an area of 500 m, and was as- sociated with three structures. Sherds of the A-Group, C-Group, Kerma Culture, and New Kingdom were recovered. A stone structure

    Table 1 . Relative Sizes of the Second Cataract Forts in Square Meters

    Site Intra-Muros Extra-Muros Total

    Mirgissa 20,000 100,000 120,000 (Main Fort)

    Askut 2,600 1,200 3,800 Shalfak 1,800 * 1,800 Uronarti 4,700 4,000? 8,700? Kumma 2,500 * 2,500 Semma 8,500 * 8,500

    All figures are approximate. * No significant extra- muros buildings extant. Based on

    Dunham (n. 3), Badawy (n. 2), and Vercoutter (n. 52).

    nearby with A-group, C-group, and Kerma pot- tery was presumably connected with this settle- ment. Another site (ll-M-7) had extensive refuse deposits up to 0.7 m thick, although no evidence was found of any structures. The largest ceme- tery held 69 graves (ll-H-5). Of the others only ll-M-5 with 15 graves has a count published.7

    The C-group habitation sites to the south on the Saras Plain were substantial, containing up to 50 cm of cultural deposits, but unfortunately the excavator gives no further details. The ceme- teries reported contained 65 (ll-Q-36), 30 (11-L- 12), 25 (ll-Q-57), and 25 (11-Q-ll) burials. There was also a substantial Kerma presence, with ten sites. An additional two sites were shared with the C-Group. One of the latter, a cemetery, contained some 65 C-Group and 255 Kerma graves (ll-Q-36). Unfortunately, Mills does not provide enough details about these sites to deter- mine whether their use by the two cultures was contemporary, or if the C-Group was replaced by the Kerma Culture as at Adindan.8 The Kerma settlement sites typically contained mud-brick structures (ll-Q-52 and ll-Q-43).9

    The Saras plain also contained a gold mine at Khor Ahmed Sherif (ll-Q-60). This activity rep- resents a substantial expenditure of effort, with

    4 Bruce G. Trigger, Nubia Under the Pharaohs. London: Thames and Hudson, 1976), 72.

    5 A. J. Mills, "The Archaeological Survey from Gemai to Dai-Report on the 1965-1966 Season," Rush 15 [1967-68], 200-201.

    6 A. J. Mills and H.-A. Nordstrom, "The Archaeological Survey from Gemai to Dal. Preliminary Report on the Season 1964-65," Rush 14 [1966], 1-15.

    7 Ibid. 8 Bruce Williams, C-Group, Pan Grave, and Kerma Re-

    mains at Adindan Cemeteries T, K, U, and J. University of Chicago Oriental Institute Expedition, Vol. V (Chicago: 1983). 9 Mills (supra n. 5).

  • 110 JARCE XXVIII (1991)

    Fig. 2. Settlement at Saras (after Mills and Nordstrom, n. 6, fig. 1; Mills, n. 5, fig. 1).

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 111

    underground as well as open pit workings. Asso- ciated with this activity were four sets of multi- room workshops, located closer to the river. The one site illustrated in the report (ll-Q-59) had over a dozen rooms. The little pottery which was found there appeared to date to Dynasty XII. Pounders and grinders, along with tail- ings of crushed quartz, provide evidence of ore reduction.10

    Clearly, Askut was by no means isolated, but located at one end of a wide plain capable of supporting a comparatively large population. It would have been in association with a signifi- cant native presence from the Middle Kingdom through the New Kingdom. Saras was an area where four cultures, Kerma, C-Group, Pan Grave, and Egyptian, could, and probably did, interact on a variety of levels. The shared C-Group/Kerma cemetery at Saras, and presence of Kerma Classique and Pan Grave pottery at Askut (fig. 3, and below) argue for significant contact, at least in the Second Intermediate Pe- riod. In fact, this is the southernmost area of C-group occupation, although a few scattered C-Group sites are present as far south as Semna. Saras was also the northernmost area with more than an ephemeral Kerma presence, even iso- lated finds being rare north of Wadi Haifa.11 Thus Askut, while not lying on a formidable natural boundary, may have lain upon a more significant cultural one.12 Still, without details of the artifacts, and especially ceramics, present on the C-Group sites, it is impossible to deter- mine whether or not they are contemporaneous with the Kerma occupation. Also, Kerma Clas- sique style wares do continue into Dynasty XVIII,13 but it would be odd if the Kerman occupying force had overlooked the substantial structures at Askut. Ceramics are an indicator of

    contact, not necessarily culture. Preliminary analysis of the Askut pottery indicates that Kerma and Pan Grave ceramics make up only a small portion of the total assemblage (perhaps 10-25%). Thus it is possible that Askut may have retained a Second Intermediate Period Egyp- tian client of the Kerman king, similar to that documented at Buhen.14

    Askut was also situated in proximity to an important resource, namely gold. Neither living quarters nor settlements were located in the vi- cinity of the ore processing workshops,15 and the paucity of ceramics found at the working sites suggests a lack of permanent occupation there. Workers might have come from the C-Group villages on the Plain. Yet virtually all of the pottery found at the workshops was Egyptian, suggesting that they were staffed entirely by personnel from either Askut or Shalfak (see be- low). The most likely administrative center for this activity is Askut, which was the larger es- tablishment, and upriver (and thus in the same direction as the gold would travel to its final destination, Egypt).

    Askut also shows some evidence of final ore processing activity. Badawy noted that pounders and grinders in the Middle Kingdom deposits at Askut were suitable for use in ore reduction.16 Some of this groundstone does indeed provide a good parallel to the spherical pounders asso- ciated with copper ore reduction at the New Kingdom workshops of Timna and copper/tur- quoise production in the Sinai.17 At least some of this groundstone, however, was surely used for grinding grain (fig. 4).18 Small plumb-bobs from Askut could have been used to balance

    10 Ibid., 204, 206, and fig. 5. 11 Torgny Save-Soderbergh, "Preliminary Report of the Scandinavian Joint Expedition: Archaeological Investiga- tions between Faras and Gemai, November 1962-March 1963," Kush 12(1964), 19-39.

    12 Georges Posener, "Pour une localisation du pays Koush au Moyen Empire," Kush VI (1958), 50ff.; Vercoutter et al. (supra n. 3), 167; Trigger (supra n. 4), 95.

    13 Manfred Bietak, Studien zur Chronologie der Nubi- schen C-Gruppe. Osterreiche Akademie zur Wissenschaften, Philosophisch-Historische Kasse. Denkschriften 97 (Vienna, 1968).

    14 H. S. Smith, Buhen II. The Inscriptions. Egypt Explora- tion Society Memoir 48 (London, 1976). 15 Mills (supra n. 5), 206. 16 Alexander Badawy, "Askut: A Middle Kingdom For- tress in Nubia," Archaeology 18 (1965), 131.

    17 Beno Rothenberg, "Ancient Copper Industries in the Western Arabah," Palestinian Exploration Quarterly 94 (1962), pl. XIII; idem, Were These King Solomon's Mines? (New York, Stein Sc Day, 1972), 180, pls. 23-24, and W. M. Flinders Petrie, Researches in Sinai (London: John Murray, 1906), 51, fig. 58.

    18 Delwen Samuel, "Their Staff of Life: Initial Investiga- tions on Egyptian Bread Making," Chapter 12 in Barry Kemp, Amarna Reports V (London: Egypt Exploration So- ciety, 1989), thanks to Mark Lehner for this identification.

  • 112 JARCE XXVIII (1991)

    Fig. 3. Second Intermediate Period Ceramics from Askut.

    scales (fig. 5).19 A fragmentary gold weight in- scribed with Snwsrt di cnh 20 was also found at Askut (fig. 5). By comparison the largest weight found at Uronarti was of only nine units. Small uninscribed weights from Askut are typical of

    the gold weights found at Uronarti. Those illus- trated fall into the following distribution: No. 772 weighs 13.05 grams (201.4 grains), exactly corresponding to the average one unit weight of the inscribed examples from Uronarti. Number 1717 measures very close to two average units at

    19 They are very similar to those found at Kahun (W. M. Flinders Petrie, Tools and Weapons [Warminster: Aris and Phillips, 1917 (reprint 1974)], pl. XLVII, B74-81). Of course, they could have been used as plumb lines and not on scales,

    although it is interesting to note that they are much smaller than the Kahun examples.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 113

    Fig. 4. Saddle-backed Quern.

    Fig. 5. Weights from Askut.

  • 114 JARCE XXVIII (1991)

    Fig. 6a. Askut "Settling System."

    Fig. 6b. Gold Washing Table, Wadi Allaqi (Belle- fonds) (after Vercoutter, n. 23, fig. 4).

    26.3 grams (405.9 grains). Number 1752, weigh- ing 18.6 grams (287.0 grains), could be a I/2 weight with a unit of 12.4 (191.4 grains), falling at the low end of the Uronarti distribution, or might have served some other purpose. This is the beqa standard, which Petrie notes is often associated with gold weights. The Askut exam- ples illustrated here are all within his Middle Kingdom "low" range of 12.2-13.1 grams.20

    The set of "settling basins," located just out- side the entrance to the main gateway (fig. 6a),21 might have been used to wash the reduced ore.22 The deep tanks at the top (a,b) would simply provide a source of water, which would be sluiced over powdered ore placed in the long, shallow channel (c). The lighter and worthless matrix would be washed away, leaving the heavier gold behind. The final deep basin (d) would serve to catch the water for potential recycling, and perhaps more importantly the ore residues, which might then be washed again to retrieve any gold missed the first time. A parallel to this system, recorded by Linant de Bellefonds, occurs in the desert along the Wadi Allaqi.23 Although differing in some details, it, like the Askut system, consists of a long, narrow wash- ing table with basins at both head and foot (fig. 6b). Unlike the example from Askut, however, the supply of water was a major difficulty. The addition of a channel and shadouf to recycle the water dictated that the first basin (a) be placed at a lower level than the washing table (b). Linant de Bellefonds notes that this waterwork was associated with gold workings of some impor- tance, as well as a group of ruined habitations, but provides no details as to their date.24 The shadouf is first attested in Egypt in the Amarna period, as seen in a representation from the tomb of Neferhotep,25 although it may have come into use earlier.26 Whatever its date, the Wadi Allaqi system provides a good functional

    20 Dunham (supra n. 3), 35f., pl. XXXV A&B, and W. M. Flinders Petrie, Ancient Weights and Measures (Warminster: Aris and Philips, 1926 [reprint 1974]), 17-19. The range of both Petrie's Middle Kingdom "high" and "low" beqa is 12.2-14.0 grams (188-215.2 grains), with a median of 13.1 grams, exactly the Uronarti standard. Fractional weights are not unknown (Arye Ben-David, "A Scarab-shaped Weight- stone," PEQ 106 [1974], 79-82). The style of the Askut inscribed and uninscribed weights is similar to those illus- trated by Petrie (op. cit., pls. VI, VII, X, XI).

    21 See Alexander Badawy, "An Egyptian Fortress in the 'Belly of the Rock': Further Excavations and Discoveries in the Sudanese Island of Askut," Illustrated London News (July 16, 1964), 86-88, fig. 3, upper right hand corner.

    22 Alexander Badawy, "Archaeological Problems Relating to the Egyptian Fortress at Askut," JARCE 5 (1966), 25. 23 Linant de Bellefonds Bey, L'Etbaye. Pays habite par les Arabes Bicharieh. Geographie, Ethnologie, Mines d'Or (Paris: Libraire de la Societe de Geographie, 1825), 27-29, illustration on p. 28, see also Jean Vercoutter, "The Gold of Rush," Rush 7 (1959), 122f., fig. 4, and William Y. Adams, "The Vintage of Nubia," Rush 14 (1966), 269.

    24 Linant de Bellefonds, op. cit., 27. 25 N. de Garis Davies, The Tomb of Nefer-Hotep at

    Thebes. Metropolitan Museum of Art Egyptian Expedition, Vol. 1, Publication 9 (New York: Metropolitan Museum of Art, 1933), pls. 46, 47.

    26 Herbert E. Winlock, The Rise and Fall of the Middle Kingdom in Thebes (New York: MacMillan, 1947), 165f.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 115

    Table 2. Magazine Capacity of the Second Cataract Forts

    Site Capacity (m3) Mirgissa 1,063.69 Askut 1,632.18 Shalfak 389.28 Uronarti (Block VI) 444.34 Kumma 574.31 Semna (max. 1000?)

    Total 5,103.80

    After Kemp (n. 43): Table 1.

    parallel, since the effort required to dig a chan- nel and lower basin would not be justified at Askut, which had a ready source of water in the Nile.

    The large block of storerooms is another fea- ture which sets Askut apart. They take up al- most half of the interior space, and alone would practically fill up the interior of Shalfak. In both proportional and absolute terms, no other of the Second Cataract Forts, including Mirgissa, had a greater capacity for storage (Table 2). Support of the mining activity might provide an explanation.27 Initial processing of the ore, how- ever, took place on site at Saras. By the time it was further processed at Askut, it would hardly require so extensive a focus on storerooms. The facilities outside the fort should have been able to account for these needs.

    Badawy sought to explain this feature by argu- ing that Askut was a major trade and produc- tion center.28 In support of this idea, he identified several structures associated with the Middle Kingdom occupation as pottery kilns.29 Kilns do appear at Mirgissa,30 and Serra East.31 Badawy's only proof for their occurrence at Askut, how-

    ever, is the presence of large deposits of ash and sherds, which would be equally appropriate for an oven or other cooking facility. Additionally, a real ceramic industry should have produced an abundance of "wasters," sherds and vessels show- ing spalling and a high degree of vitrification.32 Both "wasters" and unfired vessels do occur at Serra East.33 Badawy also overestimated the im- portance of metallurgical activity at Askut, evinced by the presence of open molds for cop- per tools (fig. 7), by misidentifying bread molds for ingot molds, a common mistake in the litera- ture.34 There is thus little evidence to support the idea that Askut produced a large enough surplus of goods above its own needs to have served as a major craft and production center.35

    Production, however, is not a prerequisite for mercantile activity at a given site, and Askut could have served as a center of localized trade. The Middle Kingdom C-Croup, far from evinc- ing a passion for Egyptian products, is noted for the retention of native material culture despite close proximity to Middle Kingdom Egyptian settlements. The only Egyptian import during this period36 at the Adindan C-Group cemeteries are faience beads.37 Of course, some of the Egyp- tian goods may have disappeared with the ex- tensive looting which is characteristic of their cemeteries. Yet, copper weapons appear so abruptly with the beginning of Kerman Control in the lib phase, that Bietak uses them as a diagnostic feature.38 Williams suggests that this pattern almost certainly indicates that trade of this important metal was actually restricted by the Egyptians.39 Perishable goods, especially

    27 Cf. Badawy (supra n. 21), 88. 28 Ibid., Badawy (supra n. 16), 127, 131, and idem (supra

    n. 2), 66. 29 Alexander Badawy, "Preliminary Report on the Exca-

    vations by the University of California at Askut," Rush 12 (1964), 51; and idem (supra n. 2), 26ff., 54. 30 Vercoutter et al. (supra n. 3), figs. 23-24. 31 Bruce Williams, "The Nubian Publication Project," The Oriental Institute 1986-1987 Annual Report (1987), 57-60.

    32 Owen S. Rye, Pottery Technology: Principles and Re- construction. Manuals on Archaeology 4 (Washington, D.C.: Taraxcum, 1981), 110, figs. 2, 91-95. Some unfired vessels do occur at Askut, but need not represent more than small scale production for internal consumption (cf., ibid., 97-99). Addi- tionally, at least some of these date to the New Kingdom.

    33 Williams, op. cit. 34 Helen Jacquet-Gordon, "A Tentative Typology of Bread Moulds," in Altdgyptische Keramik, edited by Dorothea Arnold, 1 1-24. Deutsches Archaologisches Institut, Abt. Kairo (Mainz am Rhein: Von Zabern, 1981), 23.

    35 Cf. Badawy (supra n. 16), 131. 36 Bietak's Ha (Bietak, supra n. 13). 37 Williams (supra n. 8), 117. 38 Bietak, op. cit. 39 Williams, loc. cit.

  • 116 JARCE XXVIII (1991)

    Fig. 7. Molds: a. Adze; b. Bread.

    foodstuffs, are hard to trace. Mills and Nord- strom,40 however, found a distinct absence of Egyptian ceramic types in the C-Group habita- tion and cemetery sites in the area. The pattern at the Adindan cemeteries was similar, with the presence of Egyptian ceramics dropping off sharply with the lib phase (late Middle King- dom).41 Additionally, it has been argued that the C-group's lack of reliance on Egyptian foodstuffs was partly responsible for their independence.42 The pattern reflected in the archaeological record is one of aloof contact and limited trading.

    A more likely explanation for the magazines is Kemp's very plausible suggestion that this

    kind of room block, found at all of the Second Cataract forts, actually represents a granary com- plex.43 Kemp draws a parallel between these sets of rooms and the granary model from the tomb of Meketre. In the latter, scribes sit in an outer room, off of which depend a set of contiguous square rooms. As each sack is recorded, the bearer ascends a stair leading to the top of the wall, pouring his grain into one of the chambers through its open roof.44 Each of these features, the scribe's area, stair, and set of rooms with

    40 Mills and Nordstrom (supra n. 6); Mills (supra n. 5). 41 Williams, loc. cit. 42 Trigger (supra n. 4), 79-80.

    43 Barry J. Kemp, "Large Middle Kingdom Granary Build- ings (and the archaeology of administration)," ZAS 113 (1986), 120-36. 44 Herbert E. Winlock, Models of Daily Life in Ancient Egypt. Metropolitan Museum of Art Pub. 18 (Cambridge, Mass., 1955), 25-27, 87-88, pls. 20, 62-63.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 117

    limited access, are present at the Second Cataract forts, as well as in the mansions of the town of Kahun. The placement of pillars shows, at least at Askut and Mirgissa, however, that the rooms in Nubia were covered. Presumably the grain was poured in through holes in the roof, as suggested in Kemp's reconstruction.45 Grain could then be extracted through the doorways, each room emptied in succession. Each of these structures, excepting those at Kahun, were also associated with granary peg sealings from doors and/or boxes (misidentified as bag sealings by Reisner).46 Askut was no exception. Sealings of the htm snwt Dct ntr nfr nb tEwy Snwsrt ("Seal of the Great Granary of the Perfect God, Lord of the Two Lands Senwosret"), along with that of the "Granary of Uronarti," were recovered.

    Kemp estimates the number of yearly rations that the granary at each of the forts represents by dividing the reconstructed maximum volume at each facility by an estimate of the average an- nual per capita ration. According to this analy- sis, the granary at Askut, when full, represents the potential to feed 3264 to 5628 individuals for one year, over three and one-half times the ca- pacity of Uronarti (889 to 1532), and one-third above that of the huge fortress of Mirgissa (2127 to 3668). 47 Because the combined total capacity of all the granaries of the Second Cataract forts far exceeded the needs of their combined garri- sons, Kemp hypothesized that they served as a secure base of supply for the periodic campaigns of the Middle Kingdom.48 Askut's large propor- tional emphasis on granaries (22% of its total area49), coupled with its secure location on an island well behind the frontier, suggests to Kemp that its primary function was as a fortified grain store.50

    It is odd, therefore, that Askut, important in terms of settlement pattern, natural resources, and military strategy, goes unnamed in the his- torical record. One possible name is contained in the Ramesseum Onomasticon, dating from the late Second Intermediate Period. This docu- ment contains a list of the fortresses of Lower Nubia, running from south to north.51 The first eight of these place names correspond to the Second Cataract region. The location of seven of these names has been determined through the discovery of inscriptions or seal impressions at each site. Following the order of the Onomasti- con they are (fig. 1): 1. Ddit sti - Semna South; 2. Shm HckSwRcmSc-hrw- Semna; 3. Itnw Pdwt- Kumma; 4. Hsf Iwniv - Uronarti; 5. Wcf Hlsivt - Shalfak; 7. ikn - Mirgissa; and 8. Bwhn - Bu- hen.52 Only a number 6, originally read by Gardiner as Dr WtiwQ), but better read as Dr Mtiw(?),53 has yet to be assigned. Since the list follows a strict geographical order from south to north, Dr MtiwQ) must fall between Shalfak (Number 5) and Mirgissa (number 7). Only four substantial Egyptian sites exist between these two points, Askut, Murshid, Gemai, and Dabe- narti. Murshid and Gemai can easily be elimi- nated since they were merely small watchposts, with a few rooms to accommodate the men stationed there, who were no doubt attached to the garrisons at Askut and Mirgissa. In his dis- cussion of the list, Gardiner argued that Dabe- narti or Mirgissa were the most likely candidates for Dr MtiwQ) because of their prominence. Mirgissa has since been identified with Ikn, leav- ing the large enclosure at Dabenarti as the chief candidate. But Gardiner's argument was made

    45 Kemp, op. cit., fig. 5. 46 Martha H. Weincke, "Clay Sealings from Shechem, the Sudan, and the Aegean," JNES 35 (1977), 127-30.

    47 Kemp, op. cit., Table 2. 48 These figures do not include Kemp's proposed recon- struction of a granary complex at Semna. If that building filled all of the available space, it could have supported 3,448 to 2,000 individuals for one year (ibid.). Even supposing that this was the case, Askut's granary would exceed its capacity by one-third. Depending on the size of Semna's granary, Askut thus accounts for from one-quarter to one-half of the combined capacities of Askut, Shalfak, Uronarti, Semna and Kumma.

    49 Ibid., Table 1. 50 Ibid., 134.

    51 Alan H. Gardiner, "An Ancient List of the Fortresses of Nubia," JEA 3 (1916), 184-92; idem, Ancient Egyptian Onomastica, 3 vols. (Oxford: Oxford University Press, 1947), 10-11,263, 266, pl. II.

    52 Gardiner, op. cit., 1916; Dunham and Janssen (supra n. 3); Dunham (supra n. 3); Jean Vercoutter, "La Stele de Mirgissa IM.209 et la Localisation d'Iken (Kor ou Mirgissa?)," Rd'E 16 (1964), 179-91; Knudstad (supra n. 3); H. S. Smith, "Kor: Report on the Excavations of the Egypt Exploration Society at Kor, 1965," Rush 14 (1966), 187-243; Louis Zabkar, "Semna South: The Southern Fortress," JEA 61:42-44.

    53 Vercoutter, op. cit., 186, n. 4. Gardiner himself notes that the reading of m is preferable to w (op. cit., 1916, 185), as an examination of the original confirms (idem., op. cit., 1947, pl. II, cf. Georg Moller, Hieratische Paldographie, 3 vols. [Osnabriick: Otto Zeller, 1927 (reprint 1965)], Number 196).

  • 118 JARCE XXVIII (1991)

    before the discovery of a substantial Middle Kingdom fortress at Askut. Thus we are left with the dilemma of two fortresses with only one name to divide between them.54

    In the absence of textual material (but see below), one must turn to more circumstantial evidence. For either Askut or Dabenarti to corre- spond with Dr Mtiw(7), the site must fit the following four criteria, although the last is of the least importance:

    1) The fortress was built in the Middle King- dom (when, presumably, the list was originally compiled).

    2) The fortress was comparable in size and appointments to the other sites included in the Onomasticon.

    3) The site was actually occupied during the Middle Kingdom.

    4) The site shows continuity of occupation through the Thirteenth Dynasty and into the Second Intermediate Period (when the Onomasti- con was actually written).55

    Evidence supporting these four points for each site will be assessed, considering Askut first, followed by Dabenarti.

    Askut

    1: A rock inscription (Askut 1, fig. 12a) reads:56

    1. rS n(i) hcp rnpt 3 hr hm n(i) 2. niswt bity shmkBRc cnh dt r nhh 3. hft wnn smsw n(i) hkS sbk sd ib

    4. hr ts m mnw ir.n 5. snwsrt mjc-hrw

    "Water level, Year 3 under the Majesty of the King of Upper and Lower Egypt Sekhem-ka- Rec, may he live forever and ever, when the royal follower Sobek's son Ib was commanding in the Fortress which Senwosret, t.v., built."

    The king mentioned in this inscription is Amenemhet-sonebef, the second king of Dynasty XIII. The measurement of the water level dur- ing his reign demonstrates that Askut was still active at the end of the Middle Kingdom. Addi- tionally, the fortress is described as being built by an unidentified Senwosret, presumably Sen- wosret III, who is credited with the construction of all of the Middle Kingdom forts south of Mirgissa except Semna South. In any case, the reign of Senwosret III provides a terminus ante quern for Askut' s erection, since no king named Senwosret reigned after him. Additionally, a sample of wood from one of the transverse rein- forcing beams of the northern fortification wall was dated by Radiocarbon testing to 3610 80 B.P., yielding a calibrated interval of c. 1700- 2100 b.c. (UCLA-1656C).

    2: An architectural analysis of Askut confirms this attribution. In plan it most resembles Uro- narti and Shalfak, both of which, like Askut, were suited to the peculiarities of their loca- tions. It has all of the components of the other forts, including timber reinforcement of the walls at regular intervals, a system of towers or sup- ports for an overhanging platform,57 protective spur walls, a massive entrance gate, a block of magazines, a "commandant's mansion" very similar to the "temple" within the walls of Uronarti, the typical three room "barracks" quar- ters, and a covered stair providing access to water in case of siege (fig. 8). Thus, Askut, by no

    54 Cf. Vercoutter, op. cit., Gardiner, op. cit. (1916), 190, and for Murshid and Gemai, Mills and Nordstrom (supra n. 6), 11.

    55 The date of the Onomasticon is not certain. Gardiner suggests Dynasty XIII to XIV (op. cit. [1947], 6). A late Middle Kingdom date is thus possible if the document falls within the earlier part of the range. 56 Note the difference in Vercoutter' s translation of the name and titles ("Semna South Fort and the Records of Nile

    Levels at Kumma," Kush 14 [1966], 139). The version pre- sented here is taken from a hand copy of the inscription included in Badawy, supra n. 2. The title smsw n(i) hk.D is certain, both from the copy and a (rather poor) copy of the photograph of the original. The reading of the name itself is less certain. The reading hr ts m, "upon commanding in," is to be preferred to hi>ty-c m "foremost in."

    57 A. W. Lawrence, "Ancient Egyptian Fortifications," JEA 51 (1965), 75-76.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 119

    Fig. 8. Askut.

  • 120 JARCE XXVIII (1991)

    Fig. 9a. Vessels typical of the Middle Kingdom from Askut.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 121

    Fig. 9b. Askut Middle Kingdom Cups.

    means the smallest of the Second Cataract forts, includes all of the appointments found at the other fortresses. In fact, Vercoutter has already tentatively identified Askut with Dr Mtiw be- cause of this similarity.58 As Kemp has suggested (see above), Askut may have served as an impor- tant granary depot, supplying the needs of the armies used in the Nubian campaigns of late Dynasty XII. Clearly, Askut was eminently suited for inclusion on the Ramesseum Onomasticon list.

    3: The archaeological record at Askut shows clear evidence of a Middle Kingdom occupation. Examples of a variety of ceramic types restricted to the Middle Kingdom were recovered (fig. 9a), along with a large corpus of sealings with types very similar to those found at Uronarti. The latter included examples of the Middle Kingdom official seals of Semna, Uronarti, and Buhen.59

    58 Vercoutter et al. (supra n. 3), 187, n. 154.

    59 Stuart Tyson Smith, "Administration at the Egyptian Middle Kingdom Frontier: Sealings from Uronarti and Askut," in Aegean Seals, Sealings and Administration, edited by Thomas G. Palaima. Aegaeum 5 (Liege: Universite de Liege, 1990).

  • 122 JARCE XXVIII (1991)

    4: Dorothea Arnold has shown that regular changes in the Vessel Index (Width divided by Height) of hemispherical cups are a good chro- nological indicator for the Middle Kingdom.60 The distribution at Askut reflects a strong Dy- nasty 13 occupation (fig. 9b).61 There is also good reason to believe that Askut was in use during the Second Intermediate Period when Kerma controlled Lower Nubia. A Kerman ceme- tery with 20 graves (ll-M-19) was located within 6.8 km of Askut. "The graves are mostly east- west oriented . . . , in most cases, on a bed, the head to the west, and often accompanied by an animal sacrifice of a gazelle or a ram."62 The authors go on to suggest that, since no Kerman settlements are located nearby, Askut is the most likely place of habitation. Sherds of Kerma Classique black topped ware, along with incised Pan Grave sherds were, in fact, recovered from Askut (fig. 3), and are associated with contexts just above Middle Kingdom deposits.63 Larger amounts of Second Intermediate Period Egyp- tian style pottery, including combed and Yahu- diya wares (fig. 3), also occur in these contexts. The last is the most complete of two examples of a type which dates from c. 1710-1650 b.c.64 In light of this evidence, it is more likely that Askut was occupied at this period by an Egyp- tian client of the Kerman king, as is well docu- mented at Buhen.65 Although a final conclusion must await the thorough analysis of the ceramic

    distribution, the presence of these wares does suggest a Second Intermediate occupation.66

    Dabenarti

    The island site of Dabenarti has been exam- ined by L. Borchardt, Noel F. Wheeler, under the supervision of George Reisner, and Jay Ruby during the Aswan High Dam salvage campaign.67 The last examination was the most thorough, and included test excavations.

    1 & 2: There is little evidence for the date of Dabenarti's construction. There are no rock in- scriptions, and the site's architecture is atypical of the Middle Kingdom design both in plan (see fig. 10), and in construction technique, which consisted of "Gewoehnlicher Ziegelverband ohne Matten und Holzeinlagen."68 Borchardt found "keine Reste von Bauten sichtbar" in the in- terior,69 and does not report the presence of any artifacts. He concludes by saying: "Die Anlage weicht vollstaendig von der der uebrigen Fest- ungen ab, nicht nur in der Ausfuehrung, die weniger sorgfaeltig ist, sondern auch in der all- gemeinen Anordnung. " 70 Hesse's analysis of brick size indicates that Dabenarti was also atypi- cal in this respect.71

    3 8c 4: Occupation at Dabenarti cannot have been more than ephemeral. Borchardt does not re- port the presence of any artifacts,72 and Wheeler's later survey could recover only four potsherds

    60 Dorothea Arnold, "Pottery," Chapter XII in The Pyra- mid of Senwosret I, edited by Dieter Arnold. Metropolitan Museum of Art Egyptian Expedition Publication XXII (New York, 1988), 140ff. 61 Note that this distribution is still preliminary. 62 Mills and Nordstrom (supra n. 6), 10. 63 Cf. Brigette Gratien, Les Cultures Kerma. Publications de l'Universite de Lille III (Lille: Presses de l'Universite de Lille, 1978), fig. 62; and Barry Kemp, "An Incised Sherd from Kahun, Egypt," JNES 36 (1977), 289-92. Other sherds could be from C-Group domestic wares. 64 I am indebted to Dr. Bietak for this identification (April 30, 1990). The style is Piriform lb, and similar examples have been found in Egypt, Nubia, and the Levant (Manfred Bietak, "Archaologisher Bufund und historische Interpreta- tion am Beispeil der Tell El-Yahudiya-Ware," Akten des Vierten Internationalen Agyptologen Kongresses Munchen 1985. Band 2 [Hamburg: Helmut Buske], 1 If., Abb. 2, 9). 65 Smith (supra n. 14), 66ff.

    66 Gratien, op. cit., 220ff.; Kemp, op. cit.; and R. S. Merrillees, "El-Lisht and Tell El-Yahudiya Ware in the Archaeological Museum of the American University of Bei- rut," Levant \0 (1978), 95.

    67 L. Borchardt, Altaegyptische Festungen an der zweiten Nilschnelle. Veroeffentlichungen der Ernst von Sieglin- Expedition, III (Leipzig, 1923); Dunham (supra n. 3); and Jay W. Ruby, "Preliminary Report of the University of California Expedition to Dabnarti, 1963," Kush 12 (1964), 54-56.

    68 Borchardt, op. cit., 10. 69 Ibid. 70 Ibid. 71 Albert Hesse, "The measurement of ancient bricks and

    its archaeological interest," in Mathematics in the Archaeo- logical and Historical Sciences, edited by F. R. Hodson, D. G. Kendall, and P. Tautu, 432-35 (Chicago: Aldine- Atherton, 1971), 433, fig. 1.

    72 Borchardt, op. cit., 9-10.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 123

    Fig. 10. Dabenarti [after Ruby (n. 67)].

    (of unspecified date). The report concludes that the fort was never fully occupied.73 The final systematic survey of Dabenarti by Ruby revealed only Late Christian sherds on the surface, and his excavations in the interior of the fort un- covered "no evidence of human occupation."74 He goes on to suggest that this fact implies that the fortress does not date to the Middle King- dom at all, since an existing unfinished fortifi- cation would at least have been temporarily occupied during the New Kingdom reconquest of Nubia.

    From the above discussion, it is clear that Askut meets all of the criteria to be chosen as the location of Dr MtiwQ). It was (1) built in the Middle Kingdom, (2) sufficient in size and ap- pointments to be included on the list, (3) oc- cupied during the Middle Kingdom, and (4) probably in use during the Second Intermediate Period. Dabenarti, on the other hand, is highly

    unsatisfactory. There is no evidence even to at- tribute its construction to the Middle Kingdom. Every archaeologist who has visited it has con- cluded that it was never occupied. Even if Dabe- narti was begun in the Middle Kingdom, the author of the Onomasticon would hardly be likely to skip an important and active fortress, larger than Shalfak, Kumma, and Semna South, over the memory of a proposed fortress which was never occupied.

    The list also provides internal evidence for the association of Dr MtiwQ) with Askut. The scribe occasionally placed the heading mnw n(i) ("for- tress of") opposite a group of entries. Since this occurs at uneven intervals, it is not unreasonable to suppose that each heading represents an ad- ministrative and/or geographical unit. The first of these headings begins with DSir Sti (Semna South) and ends with Dr MtiwQ) (fig. 1 nos. 1- 6). The next heading includes Ikn (Mirgissa), Bwhn, and Ink tlwy (fig. 1 nos. 7-9).75 If Dr

    73 Dunham (supra n. 3), 177. 74 Ruby (supra n. 67), 56. 75 Gardiner (supra n. 51, 1916), 186.

  • 124 JARCE XXVIII (1991)

    a. Askut's Institutional Seals

    b. Onomasticon c. Sti after Moller Fig. 11. Askut's Name and the Ramesseum Onomasticon.

    Mtiw(?) was located at Dabenarti, one would expect the name to be included in the same heading as Mirgissa, since the two forts are in such close proximity. Askut, on the other hand, is associated with the forts of the first grouping both by proximity and architectural similarity. Thus on all counts, Askut provides the best location for Dr MtiwQ).

    Several seal impressions found at Askut come from the Upper Fort, Hurt, and Treasury of the otherwise unknown Fortress of Dr Stiw {% )> "fending off" or "destroying the Nubians" (fig. lla). Such a name would fit in very well with the other fort names in Lower Nubia. Yet it does not appear on the Onomasticon either for Nubia or the First Upper Egyptian Nome, known as Tj-Sti, "Nubian Land," where one would ex- pect it. Could Dr Stiw be Askut's real name? Were it not for the onomasticon, this conclusion would have been inevitable, if necessarily tenta-

    tive. But can the two names be reconciled? There are objections. While the papyrus is broken at this point and the reading is difficult, the critical part of the name is discernible. The only charac- ter in question is the m. It definitely begins with dr (S ii)> not dr ( 5> ) and the determinative ^ is missing on the seal, but present on the papy- rus (fig. lib). As Erman and Grapow note, how- ever, the two words are virtually synonyms, dr meaning "fernhalten von jem," or "einer Zu- stand (Hunger u.a.) beseitigen," and dr meaning "entfernen" or "vertrieben." Both use the same determinatives, ^ or v-i .76 A scribe quickly copy- ing a list of names, or especially having the list read to him, might easily have substituted the one like-sounding and like-meaning word for

    76 Adolf Erman and Hermann Grapow, Worterbuch der dgyptische Sprache, 7 vols. (Leipzig, 1927-63), vol. V, 473, 595.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 125

    the other. The determinative would have been omitted on the seal due to the limited space available. It is also well known that d becomes d during the Middle Kingdom.77 Earlier, in Dy- nasty 12, when the fort was founded and the seal carved, we might still expect dr. By the late Middle Kingdom, when the Onomasticon was written, the change to dr could have taken place. This leaves only the conversion of the m(?) to sti. In hieratic both the m and sti resemble an 8, but with the addition of two strokes for the sti sign (=i)).78 A single fat stroke does indeed ap- pear after the m(?), presumably acting as part of the following tiw bird (fig. lib). This is an administrative text, not a carefully written lit- erary exercise. If the scribe were in a hurry, especially if writing from dictation, he might have simply subsumed the two strokes of the sti into one fat stroke with the tiw. This recon- struction, only made possible by the discovery of the seal impressions, is preferable to that used formerly. No longer do we have to reconstruct an unlikely and/or otherwise unattested Mtiw, Wtiiv or WhStiw. The use of Stiw in the name is very appropriate considering the other fort names. The name of Semna South even makes reference to Sti land. It also explains why no local name had been previously identified on the over 250 seal impressions found at Askut, a large number of which originated from official door or box sealings.79

    The three institutions named in the sealings are also consistent with conditions at Askut. It has a clear upper-lower dichotomy between the main fort and southeast quarters, justifying the use of an Upper Fort seal (fig. 8). There would be ample room for a treasury, which might have stored the gold from Saras. The Hurt was a kind of labor prison, where criminals, especially those who had run away from the corvee, might be sentenced to work for the state. The institution would have had several branches at various places in Egypt, with a Great Hurt at Thebes.80

    If Askut controlled the gold mining activity at Saras, then it might have required its own branch of the Hurt in order to supply labor. The use of Egyptian forced labor for this activity would also explain the absence of C-Group ceramics at the ore reducing sites. These individuals might have been housed in the southeast quarter where the sealings were found, presumably under less than luxurious conditions. Hayes suggests that the Hurt might have served as a location for hearing court cases. The Great Hurt held a criminal record.81 It seems likely that each Hurt would have retained the criminal records for its local area. Thus Askut might have served as the administrative center for hnrt-labor and court documents throughout the Second Cataract Fort system, at least those forts south of Mirgissa.

    Identifying Askut as Dr Stiw places it firmly within its functional system, the fortresses south of the Second Cataract. Most of these establish- ments were probably erected as a strategic unit, or at least as complements of one another over a period of years, by Senwosret III. These massive monuments represent an incredibly large expen- diture of man-hours and materiel. Providing for their maintenance and garrisoning would also constitute a considerable drain on the royal re- sources. The primary purpose of these monu- ments has been the topic of much debate. The majority argue that the fortresses should be taken at face value, as the military response to a very powerful enemy lying to the south. The most likely candidate is the growing Kerman polity,82 although Trigger has since revised his opinion, arguing that they were aimed at potentially hos- tile desert and Nilotic groups which might pose a threat to the important trade route between the Second Cataract and Kerma.83 Kemp has added a new dimension to this argument by emphasiz- ing how this system might have functioned as a military tool to support the periodic campaigns of the era,84 in addition to serving as a static defense. Adams has taken an entirely opposite

    77 Jiirgen Osing, Die Nominalbildung des Agyptischen. Deutsches Archaologisches Institut, Abt. Kairo (Mainz am Rhein: Von Zabern, 1974), 790ff.

    78 Moller (supra n. 53), vol. I, 41 [#437], see fig. lie. 79 Smith (supra n. 59). 80 William C. Hayes, A Papyrus of the late Middle King-

    dom in the Brooklyn Museum (Brooklyn, 1955), 37ff.

    81 Ibid., 38-39. 82 Walter B. Emery, Egypt in Nubia (London: Hutchin-

    son, 1965), A. J. Arkell, A History of the Sudan (London: Athlone Press, 1961), Trigger (supra n. 4). 83 Idem, "The Reasons for the Construction of the Second Cataract Forts," JSSEA 12 (1982), 1-6. 84 Kemp (supra n. 43), 133f., fig. 6.

  • 126 JARCE XXVIII (1991)

    view.85 He argues that no native polity could pose enough of a threat to require such massive fortifications. He sees them as fulfilling two roles, the Egyptian need for material hypertrophy in the creation of a symbolic monument far larger than any practical requirements might necessitate, as with the construction of the Great Pyramid in the Old Kingdom and Temple of Abu Simbel in the New Kingdom, and the need to police and facilitate trade along a very bad stretch of river.

    These two theories need not be mutually ex- clusive.86 A more complete picture of the Egyp- tian activities in Lower Nubia, and especially at the Second Cataract, can be derived from a syn- thetic view, in which both the militaristic and economic roles are represented as a systemic whole, in which no one trait takes supremacy. Thus the forts functioned not only as an offen- sive and defensive military tool, but also eco- nomically with respect to natives coming from the south, the exploitation of local resources, and control of the local population. Before such a reconstruction can be proposed, however, one must consider the location of Hh, the southern boundary established by Senwosret III in Year 8 of his reign, a border which lasted well into Dynasty XIII.

    The location of Hh has profound implica- tions for the role of Askut, and, indeed, of all the other Second Cataract Forts. Any theory regarding the placement of Hh must rely upon an interpretation of the Semna Stela of Sen- wosret III. The pertinent parts of this read:87

    Southern Boundary made in Year 8 under the Majesty of KhakauRe, may he be given life for ever and ever; in order to prevent all Nhsi passing it in travelling downstream by water or by land with a ship or with all cattle of the Nhsiw, except when a Nhsi will come in order that trading might be done in Ikn or on a commission. Any good thing may be done

    with them; but without allowing a boat of the Nhsiw to pass in travelling downstream by Hh, forever.

    Two sites have been proposed for this boundary (fig. 1), the Semna cataract (favored by the ma- jority of Egyptologists88), and Abu Sir.89 If the border was located at Semna, then the main point of contact would reside there, and Askut would lie well protected behind the frontier. If the Nhsiw were allowed to travel by boat as far as Abu Sir, however, Semna decreases in impor- tance. A Nubian fleet, or even a single ship, which looked suspicious might be stopped, but most would no doubt be allowed to continue on to Ikn for trading. Askut would be located in the midst of an important area of contact between the Egyptians, Kermans, and C-Group, and must have had a significant role in policing the ri- verine traffic from the south and regulating con- tact between the C-Group and the southern Nhsiw.

    Circumstantial evidence supports the location of Hh at the Semna cataract. The Semna com- plex of forts represents the southernmost Middle Kingdom military presence, and its cataract forms a very defensible natural point of control. Addi- tionally, the only mention of it occurs on the stelae which were placed there.90 Vercoutter,

    85 William Y. Adams, Nubia: Corridor to Africa (London: Penguin, 1977), 183ff. 86 I do reject, however, Adams' idea of "material hyper- trophy."

    87 Koenigliche Museen zu Berlin, Aegyptische Inschriften aus dem Koeniglichen Museen zu Berlin (Leipzig, 1913), 255f.

    88 Cf. James Henry Breasted, Ancient Records of Egypt, vol. 1 (Chicago: University of Chicago Press, 1906), 294, n. b; Torgny Save-Soderbergh, Agypten und Nubien (Lund: Hakan Ohlsson, 1941), 76; John A. Wilson, The Burden of Egypt (Chicago: University of Chicago Press, 1951), 137; Arkell (supra n. 82), 75-76; Emery (supra n. 82), 157; Miriam Lichtheim, Ancient Egyptian Literature, vol. 1 (Berkeley: University of California Press, 1973), 118, 120, n. 1; Trigger (supra n. 4), 67. 89 Karola Zibelius, Afrikanische Orts- und Volkernamen in hieroglyphischen und hieratischen Texten. Beihefte zum Tiibinger Atlas des vordern Orients. Riehe B, Nr. 1 (Wies- baden: Reichert, 1972), 147, who follows Vercoutter (supra n. 52), 187ff.

    90 Another possible mention of Hh is contained in the biographical inscription of Sihathor, an official of Amenem- het II, although Breasted reads the name as hc, near Abu Simbel (op. cit., 274). The account mentions that Sihathor had conquered the area, and gone around its islands. This sheds little light on the location of Hh, since both the Second Cataract and the approach to Semna have islands which provide an obstacle to navigation.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 127

    however, makes several objections to the equiva- lence of Hh and Semna which cannot be dis- missed out of hand.91 As the Middle Kingdom frontier, Hh would have been an important place to the Egyptians. One would expect to find some mention of it in one of the many private inscriptions recorded at the Semna cata- ract, and as an epithet of one of the deities worshiped in the temples there. Yet Hh appears nowhere other than on the stelae. He also ar- gues, less convincingly, that Hh must, in fact, be located to the north of ikn, Abu Sir being the most likely point north of Mirgissa that might qualify.

    At the crux of this disagreement are differing interpretations of the wording of the Semna stela. The standard reading of this text has been set forth by Lichtheim.92 The first clause, begin- ning "in order to prevent . . . ," sets up an abso- lute boundary and point of control. The sec- ond clause, however, beginning with "except when . . . ," sets up a special circumstance, ex- empting Nubians who come either to trade at Ikn, or upon a commission. The last clause expands on this point, and adds a further stric- ture, that no boat of the Nubians (even if its occupants wish to trade at Ikn or are on a commission) can travel downstream past Hh. Thus, anyone from the south wishing to travel farther north was obliged to either transfer their goods or themselves onto Egyptian ships, or mount an overland expedition.

    Vercoutter, however, sees an internal contra- diction in the text when Hh is equated with Semna.93 He finds it unlikely that native boats would be refused access to the principal trading center, Ikn. The absolute border, at Hh must necessarily lie beyond that point (Mirgissa), to allow the merchants and emissaries access. Semna might still serve to restrict traffic, but those with legitimate business would be allowed to travel by land or by boat at least as far as Ikn, but probably not past the Second Cataract. He there- fore suggests the prominent point of Abu Sir, with its many Middle Kingdom inscriptions of

    expeditions into Nubia, as the most suitable alternative to Semna.

    In this argument, however, Vercoutter fails to apply the same test to Abu Sir as he does to Semna. The absence of Hh in the multitude of texts at Abu Sir is just as remarkable as its absence at Semna. Furthermore, as Lichtheim has pointed out, the wording of the stela is quite clear. A specific exception is made for Nhsi going to Ikn to trade or on a commission. They may travel north, but only by land (or, pre- sumably, on an Egyptian boat). While this might indeed prove inconvenient to the native mer- chants, and even to the Egyptians, it is certainly not beyond the realm of possibility. In fact, this exception for merchants trading at Ikn actually supports a location to the south of Mirgissa.94 Hh, however, cannot be made equivalent with Semna fort itself, as Lichtheim suggests.95 Semna, Kumma, and Semna South already have names well attested from the Middle Kingdom. If the frontier were actually defined by a single fortress, one would expect it to carry the place name, even as Buhen and Ikn carry local names.

    Still, locating Hh directly at the Semna cata- ract is problematical because of Vercoutter' s objection that the name appears in no other inscription there. This problem can be over- come, however, if Hh is seen as a more general geographical term (as suggested by Gardiner96), covering a comparatively large area extending from the Semna cataract northward to some point South of Ikn (fig. 1). That Hh extended at least as far as Uronarti is implied by Senwosret Ill's stela of Year 16. Although it does not specifically mention Hh, being chiefly intended to commemorate the founding of the fortress itself, it does include the same poem used on the almost duplicate stela from Semna, also of Year 16, which commemorates the re-establishment of the boundary at Hh. Both monuments con- clude with "Now as for every son of mine who will strengthen this boundary which my Majesty (l.p.h.) has made, he is my son. . . . Now my Majesty caused that a statue of my Majesty be

    91 Vercoutter, loc. cit. 92 Lichtheim, loc. cit. 93 Vercoutter, loc. cit.

    94 Lichtheim, loc. cit. 95 Also implied by Breasted, op. cit., 294, n. b. 96 Gardiner (supra n. 51, 1916), 190.

  • 128 JARCE XXVIII (1991)

    made upon this boundary, so that you might prosper for it, and fight for it."97 Janssen, in his commentary, had some difficulty reconciling the duplication of the last passage, establishing, in effect, the same statue at the same border in two different places. If, however, we understand "this boundary" (Hh) to include Uronarti, then there is no contradiction at all, since the statue could be placed anywhere from Semna to Uronarti and still be on "this boundary."98 A Hh cover- ing this area would also explain why it was not mentioned in any other local inscriptions. The ancient authors would naturally tend to empha- size the specific location over a more general region.99 It is also not surprising that Hh fails to appear in the epithet of a deity. The region between Semna and Shalfak contained little evi- dence of settlement from any period, thus reduc- ing the likelihood of a syncretistic association of a local deity with a member of the Egyptian pantheon such as Horus or Hathor.

    The location of Hh in this region, or indeed at Semna itself, has very specific implications for the role of the forts. Now that this question has been settled, we can proceed to place Askut in its functional context in a reconstruction of each fort's role in a defensive system devoted to the regulation of trade, the native population, and natural resources. This system (fig. 12) is in- tended to supplement, not replace Kemp's re- construction, which is largely oriented towards their role in the offensive policy of the Middle Kingdom Pharaohs.

    Nubians from the south or desert would arrive at the boundary, encountering a substantial set of fortifications blocking both the water and overland trade routes,100 and from which issued regular patrols into the desert. Even small groups would be stopped, by show of force if necessary, and questioned. Nhsi not on legitimate business

    would be turned away.101 Any attempt to cir- cumvent the decree would no doubt be dealt with harshly.102 Semna's plan, in fact, is domi- nated by "barracks" style rooms, giving it the largest garrison of all the forts south of Mir- gissa. Combining these troops with the garrisons from Kumma and Semna South, the Egyptian commander could place a large force in the field (or, presumably, on water if necessary). If the threat was overwhelming, he could retire into the safety of the fortifications, and signal directly for help to Uronarti. Uronarti' s commander could then send a message to Mirgissa and ulti- mately Buhen using a pre-arranged visual signal relayed through Shalfak, Askut, Murshid, and Gemai.103

    The actual size of the garrisons has been sub- ject to some debate. I reject Vila's low estimates for the manning of Mirgissa.104 He arrives at a contingent of 35 archers and 35 foot soldiers based upon several large caches of weapons found in rooms of a well preserved building at Mirgissa. Such numbers could hardly effectively man the extensive inner fortifications, let alone the walls of the outer fortified town. Inherent in his analysis is the assumption that the weapons recovered at Mirgissa represent the entire reserve stock held there. Considering the poor preserva- tion of much of the interior of the main fort, this can hardly be justified. One of the other elite/administrative blocks might originally have held more arms. Additionally, his argument does not even consider the fortified town outside the walls of the main fort, which was not extensively excavated. Any estimate must be considered as a minimum, and should be correlated with other (i.e., architectural) data. Vila also uses overly generous divisors. It seems highly unlikely that each archer would require ten bows and eighty arrows each in reserve. Similarly, approximately two pikes and nine javelins per foot soldier also seems excessive. More reasonable numbers might be 200 or more archers (especially considering the 300-400 bows), and 150 or more pikemen.

    97 Jozef M. A. Janssen, "The Stela (Khartoum Museum No. 3) from Uronarti," JNES 12 (1953), 54, and Koeniglich Museen (supra n. 87), 257f. 98 It was presumably placed at Semna, where it would necessarily have more impact. 99 Most of the inscriptions on the rocks below Semna and Kumma do not mention a place. Those which do refer to a specific location use the names of the respective forts (Dun- ham and Janssen [supra n. 3], 130ff.). 100 Ibid., 2.

    101 Paul C. Smither, "The Semna Despatches," JEA 31 (1945), 3-10.

    102 Trigger (supra n. 4), 74-75. 103 Wheeler (supra n. 1), 255-56. 104 Vila (supra n. 3), 198-99.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 129

    Fig. 12. Function of the Fortress System.

  • 130 JARCE XXVIII (1991)

    Even this assumes that a reserve was kept for each man. Actual reserves might be less, assum- ing losses of weaponry were relatively low per person. If, for example, a reserve was kept suffi- cient to resupply one-half of the garrison, then we might reach numbers similar to those sug- gested by Emery, with 600-800 archers and 800 javelin- and pike-men. A total garrison of 2,000 or more is not beyond the realm of reason, assum- ing some losses due to disturbance, etc.105 Using Dunham's estimate of four to ten men per "bar- rack" style and three-room complex,106 the gar- rison at Mirgissa might number from at least 600-1500,107 Uronarti from 1 12-280 men, Kumma approximately 40-100 men, and the West Wing of Semna alone from 216-540 men (exceeding Reisner's rather conservative estimate of a maxi- mum of 300 men108). The garrison of Semna South, which in any case must have been small,

    would presumably be isolated by any serious assault on the border. A total of from about 368 to 920 men might have been available to the commander at Semna. A reasonable estimate of the force which the Egyptians could place in the field at need, leaving a small number to man the fortifications, might be some 400-500 men, a sizable body of troops for the period.109

    If, on the other hand, the Nhsi were found to be engaged in trade to tkn, or on official busi- ness, they could proceed along the overland route. Native cargo vessels would be required to stop and transship their goods (and/or person- nel) to Egyptian vessels, or to Egyptian or native overland expeditions. The first and seventh Semna Dispatches, originating from Semna fort, record direct trade with native merchants sailing up from the south.110 Thus, Semna should also have facilities for the transfer of goods from native cargo vessels to Egyptian bottoms, or to overland expeditions. The lower pool at Semna could hold several small vessels, as seen in a photograph of the date fleet taken in 1928. m Semna, in fact, must have been a bustling center of trade itself,112 although not as prominent as Ikn, which had access to a much broader market. Since the number of vessels arriving from the south might very well outstrip the capacity for immediate shipment north, Semna must have had facilities to hold the goods until arrange- ments could be made. Although no block of storerooms was found in the excavations there, evidence of thick walled structures, similar to those of the official buildings at the other forts, was found underneath the temple mound in the north Wing.113 Kemp has rightly suggested that this area probably contained the fort's granary, which might have been tapped for trading pur- poses.114 Every transaction recorded at Semna, in fact, included a gift of bread and beer before the trader(s) departed.115 This structure might also

    105 For example, arrows with simple wooden points, well known from military contexts (cf. Herbert E. Winlock, The Slain Soldiers of NebhepetReD Mentuhotpe. Metropolitan Museum of Art Egyptian Expedition 16 [New York, 1945]), might have been poorly preserved or overlooked during the excavations. The absence of copper or bronze weapons does not necessarily mean that none were used. We would expect copper to have a high degree of conservation and re-use, especially during the Middle Kingdom. In case of siege or abandonment, these would be one of the first items for retreating soldiers or looters to take. Less costly flint weapons were abandoned. Even so, we cannot rule out the possibility that some of the flint weapons were also taken as booty or by retreating Egyptians, and that the collection left in place represents only a portion of the total originally present. Shields would be the least well preserved, and the low numbers are not convincing. In any case, shields would require much less frequent replacement and could easily be fashioned locally, as the Mirgissa lasts for that purpose demonstrate. Also, the entire building itself and was not perfectly preserved, and some weaponry may have been lost due to disturbance.

    106 Dunham (supra n. 3), 118. 107 Reconstructing approximately 150 three-room com-

    plexes in the inner fort, which seems reasonable based on the fragmentary plans recovered for the inner fort (Ver- coutter et al. [supra n. 3], figs. 37-38). The numbers gener- ated are consistent with the higher estimates from the weapons presented above. The actual force present might have been considerably more if soldiers and/or conscriptable reserves were also quartered in the outer town. These figures also do not consider potential occupants of the elite/ad- ministrative structures.

    108 George Andrew Reisner, "Ancient Egyptian Forts at Semna and Uronarti," Bulletin of the Museum of Fine Arts Boston XXVII (1929), 72.

    109 Cf. Winlock (supra n. 105). 110 Smither (supra n. 101). 111 Reisner, op. cit., fig. 2. 112 Kemp (supra n. 43), fig. 6. 113 Dunham and Janssen (supra n. 3), 7, pls. 6C, 8A, Map

    III. 114 Kemp, op. cit., 130. 115 Smither, op. cit.

  • ASKUT AND THE ROLE OF THE SECOND CATARACT FORTS 131

    have included an attached treasury complex such as that found at Uronarti.

    Even more significant in this context, how- ever, is the outer enclosure at Semna South. Surrounded by a low, insubstantial brick wall, it contained evidence of temporary occupation, but no trace of permanent structures. The excavator suggests that it was suitable for use as a com- mercial exchange base,116 exactly the sort of fa- cility necessary for the transshipment of goods required by the edict of Senwosret III.117 The upper pool of the Semna cataract, opposite this fort, was large enough to make a harbor capable of sheltering a number of vessels.118 Ancient merchant vessels might have stopped there and off-loaded their goods, putting them into tem- porary storage at Semna South. More valuable goods might be taken for safekeeping to Semna and/or Uronarti. The large plain at Semna South would, in fact, provide a good staging area for the native caravans to await official permission to leave, or the Egyptian ones to buy up native goods brought in by boat. Alterna- tively, the goods could be transported down past the cataract to the lower pool opposite Semna fort and transferred to Egyptian ships. The ex- tensive defense wall traced by Mills from a point north of Uronarti to Semna extends into this area,119 and might have been used to keep un- authorized groups out and contain this activity.

    Kumma clearly served as an adjunct to Semna, policing the East Bank road and guarding the channel, and no doubt sending out patrols into

    the desert.120 Both Semna and Kumma would have the additional role of helping any Egyp- tian traffic past the cataract, as suggested by Adams.121 Reisner notes that one or two boats in the modern date fleet were lost each year in the difficult downstream passage through the Semna Cataract.122 A permanent garrison stationed there, available in time of peace to assist the ancient merchant vessels, might help to prevent such losses.

    Uronarti, with its extensive evidence of ad- ministrative activity,123 probably served as the local command center for the forts. Its garrison and storerooms were large enough to support Semna both militarily and economically in case of need. The garrison would also, no doubt, have worked with Semna to man the wall system running along the perimeter of the West bank. Reisner has suggested, in his analysis of the distribution of sealings, that the inhabitants of Uronarti might also have carried on a small export and import business with Egypt.124 This activity would have supplied the need of the garrison, or perhaps the garrisons of all the nearby forts, for luxury goods from Egypt. It might have been connected with the general trade from the south as well (see above). Uro- narti's garrison would also aid any Egyptian river traffic past the rapids.125 A rock inscription of Senwosret Ill's expedition of Year 19 found at Uronarti actually records the difficulties of nego- tiating the shoals there.126

    Shalfak, the smallest of the Second Cataract forts, probably sent out patrols into the desert to monitor the movements of the natives.127 Its strategic location on a sharp bend in the river at the very end of the Saras Plain, where the cliffs rise precipitously to sixty meters above the river level and the valley becomes tightly constricted,

    116 Zabkar and Zabkar (supra n. 3), 9. 117 Although strangely, they disagree with Lawrence's sug-

    gestion that Semna South was likely to have been a caravan base (ibid., 30, n. 11; Lawrence [supra n. 57], 82, n. 1). Even if traces survived of the kind of ephemeral structure that might be associated with the temporary storage and en- campment necessary for such activity, they might have been overlooked considering the conditions imposed by the sal- vage nature of the project. The Zabkars date the use of the fort until the reign of Amenemhet III, based upon the distribution of seal impressions found in a very large dump nearby (op. cit., 14). This may, however, simply date the use of the dump, and does not necessarily reflect the terminal period of occupation at the fort itself, which might well have extended into Dynasty XIII, as was the case at all of the other forts.

    118 Ibid., pl. I. 119 Ibid., 12f., pl. 1.

    120 Dunham and Janssen (supra n. 3), 2, and Kemp (supra n. 43), fig. 6. 121 Adams (supra n. 85), 184.

    122 Reisner (supra n. 108), 67. 123 Kemp (supra n. 43), 134-36, and George Andrew Reis-

    ner, "Clay Sealings of Dynasty XIII from Uronarti Fort," Rush III (1955), 26-69.

    124 Ibid., 34. 125 Adams, loc. cit. 126 Wheeler (supra n. 1), 259. 127 Kemp, op. cit., fig. 6.

  • 132 JARCE XXVIII (1991)

    implies that the fort may have helped to contain the local C-Group peoples, or at least to regulate their movements. Its garrison may also have helped boats past the rapids located there.128

    Finally, Askut, as well as being the main campaign reserve grain store, would logically also serve as a reserve for the entire system, at least in time of need. Its administrative system for the control of goods was divided into at least four branches, the Storerooms, Upper fort (pre- sumably the main fort as opposed to the south- east quarter), Hurt, and Treasury.129 Isolated from the desert trade routes and nestled safely on an island well behind the frontier, its main focus was local. Soldiers from the garrison would be detailed to man the signaling post at Mur- shid, held in constant readiness in case of emer- gency. They apparently also supplemented the garrison's diet through fishing (ll-H-4).130 The Commandant of Askut would also have overseen the mining activities at Khor Ahmed Sherif, as well as the ore reduction stations on the plain. The crushed ore would be further reduced at the fort itself, and then washed in the "settling" system. This all might have been accomplished with labor supplied by the Hurt, the gold being placed in the Treasury. Askut must also have kept watch over the considerable C-Group popu- lation of the Saras plain, which might con- ceivably pose a threat to river traffic. Some exploitation of local agricultural and pastoral activities, as well as hunting, might also have taken place, although most of the grain for Askut' s granary presumably came from Egypt. The presence of a statue of the "Director of Plowings Sob[ek . . . ]" at Askut is suggestive in this context.131 The Hurt might have been in- volved here also, since it is often associated with agricultural work. This activity would reduce

    both the drain on royal stores and cost of trans- port of supplies from Egypt, giving the forts a greater measure of self-sufficiency. Considering the local environment and likely subsistence base, exploitation of pastoral resources is per- haps more likely than agriculture during the Middle Kingdom. A preliminary pass through the faunal material from Askut indicates a heavy reliance on cattle and goat/sheep. At least one pig was present. Large amounts of fish and shellfish were also found, which ties in well with the fishing station to the north of the fort. Small numbers of gazelle and at least one ante- lope were hunted.132 The rocky stretch of river around Askut might conceivably pose difficulties for navigation, and thus Askut' s garrison may also have assisted river traffic.

    Thus, the Second Cataract forts of the Middle Kingdom operated as a well planned and inte- grated system with considerable functional differ- entiation, ranging from Semna fort's apparent emphasis of garrison, to Askut' s more passive focus on support of the other forts and local activities. One need not seek to find a single, or even primary, purpose for the system. It is clear that the forts had a multi-faceted role in the Nubian policy of the Middle Kingdom Pharaohs, serving on the one hand in support of the puni- tive campaigns to the south and as a static defense to prevent violation of the boundary, and on the other to regulate and facilitate riv- erine and overland trade, monitor the local population both on the Saras Plain and in the Western desert, and exploit the natural resources of the area. The study and publication of the assemblage from Askut, with its good context and fine preservation, can help to further clarify this picture, and shed more light on the lifeways of the garrisons.

    University of California Los Angeles

    128 Dunham (supra n. 3), 115, and Adams, loc. cit. 129 See above and Smith (supra n. 59). 130 Mills and Nordstrom (supra n. 6), 11. 131 Badawy (supra n. 16), 127f. Senwosret III is mentioned

    on the statue. It was found in a disturbed context and thus could have been traded south from Egypt during the Second Intermediate Period as with the Middle Kingdom statuary from Kerma.

    132 These are impressions only, none of the faunal mate- rial has been quantified yet, and some of the remains may date to the New Kingdom. I am obliged to Dr. Sandor Bokonyi for some of the preliminary identifications.

    Article Contentsp. 107p. 108p. 109p. 110p. 111p. 112p. 113p. 114p. 115p. 116p. 117p. 118p. 119p. 120p. 121p. 122p. 123p. 124p. 125p. 126p. 127p. 128p. 129p. 130p. 131p. 132

    Issue Table of ContentsJournal of the American Research Center in Egypt, Vol. 28 (1991), pp. 1-248Front MatterA Revised View of the Narmr Palette [pp. 1-20]Some Revisions of Temple Endowments in the New Kingdom [pp. 21-39]Composition and the Artist's Squared Grid [pp. 41-54]A Bust of Amenophis II at the Kimbell Art Museum [pp. 55-74]East Karnak Excavations, 1987-1989 [pp. 75-106]Askut and the Role of the Second Cataract Forts [pp. 107-132]Sm.f, Sm.n.f and Verbs of Motion in Sinuhe: Some Reflections [pp. 133-138]Evidence for Use of a Stone-Cutting Drag Saw by the Fourth Dynasty Egyptians [pp. 139-148]New Kingdom Astronomical Paintings and Methods of Finding and Extending Direction [pp. 149-154]The Gnomon in Egyptian Antiquity [pp. 155-185]Isis and Demeter: Symbols of Divine Motherhood [pp. 187-200]A Byzantine Chapel at Marsa Matruh (Paraitonium) [pp. 201-211]"Now You See Me, Now You Don't": Point of View and the Embedded Narrator in a-ayyib li's, "Dmat wad mid" [pp. 213-221]The Narmer Macehead and Related Objects: [Correction] [pp. 223-225]Book ReviewsReview: untitled [pp. 227-228]Review: untitled [pp. 228-229]Review: untitled [pp. 229-231]Review: untitled [pp. 231-232]Review: untitled [pp. 232-233]Review: untitled [pp. 233-235]Review: untitled [pp. 235-236]Review: untitled [pp. 236-237]Review: untitled [pp. 237-238]Review: untitled [pp. 238-239]Review: untitled [pp. 239-240]Review: untitled [pp. 240-241]Review: untitled [pp. 241-244]Review: untitled [pp. 244-245]Review: untitled [pp. 245-246]

    List of Books Received, 1990 [pp. 247-248]Back Matter