(Author's Note: a computer crash in the last days has resulted in a serious disk error with some corruption of the automatic footnote numbering system. I have not been able to correct this problem easily, so I thought I should send the document off in spite of this flaw. I have tried to indicate in the text which footnotes go with which footnote number . I also send this document without the figures, which are still in preparation. Their omission should not in any way affect one's understanding of the text. Apologies to all!)
Gebel Barkal, the Mythological Nubian Origin of Egyptian Kingship, and the Formation of the Napatan State
Introduction and Abstract.
Some months ago I was asked by the conference organizers to present a paper on "Napatan temples". In looking over what has already been written on this subject - most recently the fine treatment by Laszlo Török (1997, 299-326) - I felt it would be rather pointless for me to plough this same ground. Of course, since Török's work was prepared, important new discoveries of temples have been made in the Sudan, each with rather extraordinary implications. These are the finds at Doukki Gel/Kerma (Bonnet, Valbelle, and Ahmed 2000), at Soniyat, Hugeir, and Usli (Zurawski 1998; 2001), and at Dangeil (Anderson and Ahmed 2000). Since the excavators of these sites will be reporting on them directly at the conference, it seemed to me that my own task here ought to be nothing more than to report on the recent discoveries and conclusions of my own team in the Napatan temples at Gebel Barkal.1
Our knowledge of the Napatan Period at Gebel Barkal is founded largely on the work of George A. Reisner, who spent parts of four seasons there, from 1916 to1920, excavating the temples for Harvard University and the Museum of Fine Arts (MFA), Boston. Although Reisner published in detail only the results of his 1916 season, his original diaries and photo archives in the MFA preserve a very complete record of the scope of his work, his finds, and his interpretations. Reisner's excavating capability was massive. He worked up to four months at a time with between 250 to 400 workmen and conducted huge clearing operations. He found and recorded thousands of objects, discovered a number of important historical texts and translated them, and reconstructed the history of the site through careful analysis of the texts, statues and statue fragments, and architectural remains.
My own mission at Gebel Barkal has been the beneficiary of Reisner's great work. From the beginning our concession in the sanctuary has been precisely that of Reisner's, and it goes without saying that without Reisner's record to build on, our own contributions would be very small indeed. No one today, of course, could or would excavate on the scale that Reisner did, but the fact that he did so has left us an invaluable record of what he saw on the ground that we will probably never see ourselves. My team thus has had the advantage of being able to use Reisner's superb records to become familiar with the site, to target areas that needed clarification and further excavation, and to recognize areas that had never been probed. Here and there we were also able to correct or enhance Reisner's interpretations in the light of subsequent finds and scholarship.
My own approach to the archaeological record of Gebel Barkal has differed somewhat from Reisner's. His primary response to the material was that of the archaeologist; his focus was distinguishing building phases of temples to establish their chronology and recording the preserved texts within them to anchor the temple phases to a historical framework. My response to the material has been more that of the political and religious historian. The kinds of questions of concern to me have been: "Why was Gebel Barkal important as a religious site? What was the nature of its cult? What did the various temples mean and how were they used? What is the connection here between the New Kingdom and the Napatan Period? What prompted the Napatan rulers to convert to the Amun cult, etc? "
Excavations at many Napatan sanctuaries, especially at Gebel Barkal, reveal that the Napatan temples were generally built directly over the foundations of ruined New Kingdom temples. This indicates that the Napatan rulers of the eighth century BC deliberately restored cults and cult places that had been abandoned by the pharaohs when they evacuated Nubia some three centuries earlier. Other than the Barkal Temples, the best known Napatan temples with New Kingdom antecedents are those of Kawa, Tabo, and Doukki Gel/Kerma, to which we must now add Usli and Hugeir. New Kingdom remains have not yet actually been found at Sanam temple, but Taharqa's inscriptions there allude to its foundation by the "ancestors", by which he seems always to have meant the pharaohs (Griffith 1922, 102). Such data reveal that the Kushites, in the early Napatan Period, undertook a program of reviving long dormant Egyptian cults throughout Nubia. Somehow they had acquired a passionate belief in the Amun cult, where previously they had possessed none. Somehow they had recovered - or learned - a memory of the cults of these old Egyptian centers in order to restore them, even though within Nubia the cult had apparently been absent for three centuries and the temples had fallen to ruin. Somehow, by restoring the old Egyptian cult places, especially Gebel Barkal, they were able to present themselves both within Nubia and especially at Thebes, as the true successors of the pharaohs of the New Kingdom and the direct heirs to their throne. What exactly happened here? How did an obscure dynasty of Nubian chiefs from a village near ancient Napata rise within two or three generations to become the Twenty-fifth Dynasty of Egypt? This would seem to be one of the major unsolved issues of our discipline as well as of Egyptology.
Questions about the origins of Napatan temples and cults are directly related to questions about the origins of the Napatan state itself. Both temples and state appeared simultaneously, mysteriously, almost full-blown, at the end of a nearly three-century cultural/archaeological hiatus separating them from the end of the New Kingdom. One may thus think of the Napatan temples as the archaeological "footprint" of the obscure social and political events that gave rise to the state. In this sense their remains provide critical "foundations" for understanding Napatan origins. At Gebel Barkal the connections between the Napatan Period and the New Kingdom are especially vivid, and I have chosen to use this paper to examine the archaeological and historical implications of this relationship, which are profound.
The contents of my paper may be summarized as follows: It begins with a re-examination of the archaeological record from Gebel Barkal and concludes, from this, that the site was of far greater importance to the Egyptian pharaohs of the New Kingdom than has heretofore been suspected. Extensive use of talatat blocks on the site reveals, for example, that, just like Karnak, it had been a major cult place in the early years of Akhenaten's reign. Akhenaten's structures at Barkal, as at Karnak, were dismantled by his immediate successors and the blocks were reused in the major restorations of the Amun temples that took place in the post-Amarna era to early Dynasty 19. Centuries later, the Napatan kings built their own temples as well as their palace directly over the foundations of the later New Kingdom temples and palace, which had by then fallen to ruin. Their aim was clearly to restore the Barkal sanctuary exactly as it had been during the New Kingdom. This restoration, they apparently believed, conferred on them the right to claim the kingship of Egypt (especially Upper Egypt). Since their beliefs and behaviors were modeled on those of the pharaohs, this fact leads us to the almost unthinkable conclusion that the Egyptian pharaohs, too, prior to their abandonment of Gebel Barkal, used the site as a justification for their own claim to the kingship of (Upper) Egypt.
The paper next examines the textual data regarding Gebel Barkal, which reveals that the sanctuary, from at least the time of Thutmose III (and probably Thutmose I), was called by the same name as Karnak ("Nswt-T3wy") and that its temple was called by the same name as Karnak Temple ("Ipt-Swt"). The texts make it clear that both the Egyptians and later Kushites considered Gebel Barkal to be a southern manifestation of Karnak and understood the Nubian mountain to be the residence of the "ka" of Amun of Karnak. From their first forays into Upper Nubia, it seems, the Egyptians considered Gebel Barkal to be the original Karnak and the home of the primeval aspect of Amun. Both Karnak and Gebel Barkal also shared the epithet "Upper Egyptian Heliopolis." Luxor Temple, the "Southern Sanctuary", is shown to have been founded at Thebes by Hatshepsut as a response to the Egyptian discovery (by Thutmose I?) of Gebel Barkal in Nubia. Luxor, it seems, was built as a magical substitute at Thebes for Gebel Barkal, which was the real "Southern Sanctuary."
The mountain's religious importance and presumed power was derived from its shape, which resembled many different things The Egyptians thus identified Gebel Barkal as a residence of Amun in all his forms, united in a single mysterious being. This theory is confirmed by the so-called "Nubian Chapters" (162-165, 167) of the Book of the Dead, which reveal how the "hidden" meanings of Gebel Barkal were to be understood and "read." This data leads to the conclusion that the Egyptians recognized Gebel Barkal as the ultimate home of Amun Kamutef and the source of the royal ka. It also explains why they recognized the mountain as the source of Upper Egyptian kingship and the source of the Upper Egyptian crown. The Egyptians, for their part, apparently used Gebel Barkal to prove that their royal authority as wearers of the White Crown extended southwards to the end of Upper Nubia. The land between Karnak and Gebel Barkal from Dynasty 18, thus, became "Upper Egypt." By the late New Kingdom, this same royal authority was symbolized by the cap crown with single uraeus, a crown shape that imitated the form of Gebel Barkal, which was the imagined source of the crown. When Gebel Barkal was lost to the Egyptians after Dynasty 20, the "Upper Egyptian kingship" was lost. The Kushites, by restoring "southern Karnak" and uniting it with "northern Karnak" revived this New Kingdom concept of kingship, which justified their assumption of the cap crown. The cap crown symbolized royal authority over the reunited domains of Amun, as granted by the god through Gebel Barkal. When the Kushites assumed control over "Upper Egypt," they wore a cap crown with one uraeus. When they assumed control of Lower Egypt, they added the second uraeus to their crown.
The paper next examines the historic use of Gebel Barkal as a coronation site and describe the discovery by our team in 1997 of temple B 1100, which has proven to be the Pr-wr or "coronation temple" at Barkal. Made as a hemispeos, this was the shrine entered by the king during his coronation to receive his crown. Our excavations have revealed that the temple had three phases: Egyptian (datable to Horemheb), Napatan (datable to Piye?), and Meroitic (datable to Natakamani and Amanitore). Our finds indicate a) that Egyptian coronations were held here from Dynasty 18 (even if they may only have been magical charades for real ceremonies held simultaneously at Luxor), b) that the Napatan kings renewed this temple in order to celebrate the same coronations and to assume for themselves the same kingship as the Egyptians, and c) that the Kushites, firm in their belief that they were the heirs to the ancient kingship of the New Kingdom, continued performing coronations here well into the Meroitic Period, all the while considering the pharaohs to be their "ancestors."
The results of our excavations in B 1100 in February 2002 are next presented. At this time we found evidence (albeit preliminary) that the New Kingdom Pr-wr was destroyed by a rock fall from the cliff above. This event can be shown to have occurred after the reign of Ramses II and before the early Napatan Period. The theory is proposed that the temple was destroyed near the end of Dynasty 20 and that this, more than anything else, may have been what caused the high priests of Amun at Thebes to usurp the southern authority of the king (Ramses XI) and the Upper Egyptian kingship. When they failed to take control of Nubia from Panehsy, Viceroy of Kush, they not only lost control of Nubia but also of the Nubian Amun sanctuaries, which consequently were abandoned. The restoration of Gebel Barkal and southern kingship in the eighth century BC is proposed as a collaborative effort by the Nubian chiefdom of el-Kurru and the Theban priesthood. Its purposes were to end Tanite or Herakleopolitan political domination of Thebes and to restore the kingship of the New Kingdom, in which the Theban Amun priesthood and the ruling dynasty had a symbiotic and mutually supportive relationship.
After commenting on how Gebel Barkal was enriched architecturally and symbolically by the constructions of Taharqa, the paper concludes with a brief re-analysis of the campaign of Psammeticus II, in which it is shown with virtual certainty that Gebel Barkal was its prime objective. This campaign, which devastated of the Barkal sanctuary, is presented as the event that forced the Kushites to give up their dream of ruling the Thebaid, to reconceive their kingdom in religious and political terms, and to establish a new kingdom with a more southerly orientation.
II. The Archaeology of Gebel Barkal: A Reappraisal.
Gebel Barkal lies on the right bank of the Nile at the approximate mid-point of the river's great bend, about 325 km NNE of Khartoum (aerial photo, fig. 1). It is a small isolated sandstone butte on the western edge of Karima, and stands about 2 km from the river, which it confronts with a spectacular cliff 200 m long. The mountain's height, measured from the ancient floor level of the Great Amun Temple (B 500), is 104.5 m.; the cliff is between 80 and 95 m. high. The mountain is an anomaly in the local landscape, for, apart from its sheer cliff, it stands in an otherwise flat desert plain, and it possesses an immense, free-standing pinnacle on its south corner that rises vertically 74.6 m. It was this last feature that distinguished the mountain from all others in the Nile Valley. Very early, it seems, all these unusual characteristics played on the minds of the ancients, who made Gebel Barkal the subject of intense theological speculation and identified it as a sacred site.
Doubtless long before the Egyptians had set eyes on it, Gebel Barkal had been venerated by the Nubians. Although no pre-Egyptian settlement or cultic remains have yet been identified here, unstratified Nubian pottery has been recovered from the site dating from the Neolithic, Pre-Kerma, and Classic Kerma periods.2 This confirms that the area surrounding the mountain had been occupied from at least the fourth millennium BC. The discovery on its summit of thousands of chipped stone wasters, made of types of stones that can only be found on the desert floor, suggests that people carried stones to the mountain top to work them, a practice that implies a religious motivation.3 Another indication that the mountain may have been a pre-Egyptian cult site is the undeniable similarity of its sanctuary, as it appeared in the Egyptian and Napatan Periods, with that of the Western Defuffa at Kerma as it appeared at the end of the Classic Kerma Period. It is just possible that by that time the Defuffa, a rectangular, brick built, mountain-like platform 19 m high, may have evolved to be conceived as a kind of local magical "double" of Gebel Barkal. After all, complexes of temples were built in front of each, facing the river, and each was conceived as the dwelling place of an important god (Bonnet 1990, 32, 59-67; Kendall 1997, 79).
Although the Egyptians were probably well aware of the geography and chief landmarks of Kush long before their attempts to conquer it, they first invaded Upper Nubia militarily during the reign of Thutmose I (ca. 1504-1492 BC).4 Thutmose I penetrated the great bend of the Nile to Kurgus, 235 km upstream from Gebel Barkal, in his Year 2 and left an inscription there (Davies 1998). Until now, it has generally been assumed that his forces took the desert road from Korosko to Abu Hamed in order to bypass the Dongola Reach (Morkot 2000, 72). But evidence presented below (in Section III) makes it very clear that the king not only passed by Gebel Barkal on his way to Kurgus but also immediately identified it as a sacred place of supreme importance. The meaning he and his entourage accorded Gebel Barkal would transform the way in which cult of Amun, the origin of the world, and the nature of Egyptian kingship were thereafter understood and celebrated at Thebes.
The earliest archaeological evidence for an Egyptian presence at Gebel Barkal dates to the reign of Thutmose III (ca. 1479-1425 BC). This king's Barkal Stele, dated to his Year 47 (about 1432 BC) is the first historical document known from the site and the first to name the mountain. From this we learn that the Egyptians called it Dw-w'b ("Pure Mountain" or perhaps "Mountain of Pure Waters"). This name, I speculate, may have been derived, by phonetic and hieroglyphic pun, from an earlier, non-sacred name Dw-'b ("Horned Mountain"), which would have resulted from the mountain's curious pinnacle. Thutmose's stele also speaks of a pre-existing Nubian community at "Pure Mountain" (l. 33) as well as an Egyptian fort called Sm3-h3styw ("Slaughter of the Foreigners") (l. 2). Reisner, who recovered the stele, likewise found fragments of a seated statue of Thutmose III, wearing Heb-Sed garb.5 The earliest level of the Amun Temple B 500 ("B 500-sub"= chambers 504a, b) may also date from this period (Reisner 1917, 219, pl. 43). (Map, figs. 2, 8)
At this point there is no evidence that Amenhotep II (ca. 1427-1401 BC) undertook any construction at Barkal, but a unique fragmentary statue of the king as a rising sphinx trampling Nubian foes was found there (Boston: unregistered: Reisner 1931, p.81, no.4; Dunham 1970, p. 25). It is also this king, in his Amada and Elephantine Stelae, who first mentions a town called "Napata." The context, however, suggests that the name would have been familiar to the ancient reader and that the town was probably no recent foundation. The king states that he hung a Syrian prince from its "walls" (Breasted 1906, vol. II, p. 313; Der Manuelian 1987, 94). Until now, no such "walls" have ever been known at Gebel Barkal, but last year (2001) work crews, digging a ditch for a drainage pipe on the river side of the road in front of the mountain, discovered remains of a massive mud brick wall extending nearly 200 m. parallel to the road. Although this wall is probably not the original New Kingdom wall, it must be the remains of the ancient temenos that ringed the sanctuary for much of its existence. It suggests, at least, the sort of "wall" to which Amenhotep was alluding.
The activities of Thutmose IV at Barkal are revealed by a fragment of a statue of that king found in rubble beside B 700 (Reisner 1931, 81, no.5; Dunham 1970, 25, fig. 19, pl. 24). This statue may actually have come from B 600, a small temple the king built against the cliff. Beneath the rear corners of this structure Reisner found Thutmose's foundation deposits. The temple may even have survived reasonably intact to the Napatan Period and continued in use without significant modification until it was destroyed by a rock fall from the cliff in Meroitic times. It was then restored with blocks from the old temple, some of which bore the king's cartouches and were recorded by Reisner (1918, 99-100). A stele from B 501, depicting a king called only "Thutmose" standing before the ram-headed Amun of Gebel Barkal, is probably also to be attributed to Thutmose IV (Dunham 1970, 43, pl. 47 H).
Amenhotep III is not known for any construction at Barkal. Excluding from consideration his Soleb statues, which were brought to the site in the early Napatan Period, his original presence there is attested only by a standing statue of the king wearing the White Crown and by fragments of a statuette of Mery-mose, his Viceroy of Kush.6 A head of Queen Tiye from a royal pair statue, said to have been found at Merowe, may also have been found nearby (Kozloff and Bryan 1992, 178-179).
In his published reports, Reisner noted at Barkal a peculiar type of Egyptian masonry that he observed in the foundations of several of the Napatan temples, especially in B 500 and 300. He observed it also in the remains of several small structures erected both beside and in front of temples B 500, 800/900 and 700. Since the stones used in this masonry were associated with buildings that he could date confidently to late Dynasty 18 and early Dynasty 19, he dated the blocks to this period. They were small and rectangular, and were made of grayish or whitish sandstone; they were also laid in header and stretcher courses and were bonded together with cement. They regularly had the dimensions 1 cubit x 1/2 cubit x 3/7 cubit (52.3 x 26. 3 cm x 23 cm). Today this type of masonry is well known. It is what we call talatat and recognize as an archaeoloical signature of the early reign of Akhenaten (Map of talatat, fig. 3).7 It was used in his constructions up to his Year 5 at Karnak and elsewhere (Reisner 1917, 220, 222-224; 1918, 111; Redford 1984, 63-71; Redford 1999; Vergnieux 1999). We know that Akhenaten also used these blocks in other sites in Upper Nubia: in particular at Doukki Gel/Kerma (Pnubs) and Sesebi, and perhaps also at Kawa and Tabo.8 Their abundant presence at Barkal reveals that Akhenaten considered this site to be very high in cultic importance and that he built extensively there. The distinctive white sandstone from which the blocks were cut appears to have been quarried right on the site from an outcrop on the west side of the mountain, immediately beside B 200. This outcrop formed the mountain's basement stratum.
Apart from their appearance in the foundations of some of the larger temples, talatat blocks were also found by Reisner in some very small, temporary-looking rectangular structures built near the later "south" end of the pylon of B 501 (Reisner 1917, 218, pl. 41). He noted others ("B 520" and "B 522") on either side of the nucleus of B 500 (Reisner 1917, 224, pl. 43), as well as under the portico of B 700 and immediately to the "north" of that temple ("B 700 sub-1" and "B 700 sub-2") (Reisner 1918, 111, pl. 10). Still another appeared beside B 904 (Reisner Diary, Dec. 24-27, 1919). All of these buildings looked like small, hastily constructed shrines or shelters for statues. The one beside B 904 still contained a head from a large granite statue of a uraeus, which had carried a crown (Dunham 1970, 78, fig. 49; Reisner Diary, Dec. 26, 1919). Most of these structures had been only one block thick, a feature also observed in some of Akhenaten's structures at Karnak (Redford 1984, 75). Possibly these were the remains of some of the original structures built here by Akhenaten as shrines for statues in use during his early Heb-Sed, which was perhaps celebrated here in facsimile form concurrently with the real ceremony held at Karnak (Redford 1984, 125; Gohary 1992, 29-36 and see below, Section III). On the other hand, they could have been the first structures built by Akhenaten's successors from his dismantled buildings and could have been temporary shelters for the cult statues brought from Egypt when the site was rededicated to Amun.
Akhenaten's presence at Gebel Barkal is further attested by a fragmentary statuette of his Viceroy of Kush, Thutmose, which was found beside the small talatat structures "south" of the B 500 pylon.9 Akhenaten's activities are also manifested in the erasures of the images of Amun on the Barkal Stele of Thutmose III (Reisner and Reisner 1933a, 25, pl). They are also seen in the erasures of the names of Amun on the statue of Amenhotep III from B 700 (Dunham 1970, fig. 5) and the fragmentary statue found in debris "south" of B 904 (Reisner Diary, Jan. 7, 1920).
At Barkal Akhenaten's talatat blocks were heavily reused in the construction of the post-Amarna Amun Temple B 500 (fig. 4). This suggests that the pre-Amarna Amun temple (B 500-sub=504a,b) may have been so unacceptably altered by Akhenaten that it had to be entirely rebuilt. The initial work on the renewed temple seems to have been undertaken by Horemheb, to whom the nucleus chambers B 505-507, 514-519 and the fourth pylon can probably be attributed. The first extension of this shrine (B 503), together with the third pylon and its small added "southern" chapel (B 504c) were probably all constructed by Seti I, since his Barkal Stele, dated to his Year 11, was found in the debris of the chapel (Reisner and Reisner 1933b, 73). More talatat were used in the larger "southern" chapel (B 508-510), which was added by Ramses II, as is clear from the presence of his cartouches on several talatat blocks and a large roofing stone found there (Reisner 1917, 223-224, pl. 46). Ramses also enlarged the temple with a huge hypostyle hall (B 502). By excavating beneath the Napatan level, Reisner determined that the original Ramesside colonnade in B 502 had consisted of six rows of 12, 13, or 14 columns - or between 72 and 84 columns (Reisner Diary, Boston, Jan. 26, 1919).
Talatat blocks remain in situ in two - possibly three - other temples at Barkal. For example, they form the foundations of the outer court (B 301) of B 300, Taharqa's Mut temple (fig. 5). This indicates that the temple had an original New Kingdom phase, a fact borne out by Taharqa's own inscriptions.10 This first temple, however, was not rock-cut but a free-standing tripartite shrine built entirely in front of the mountain. Additionally, in front of Taharqa's Hathor temple, B 200, there is a single talatat block that can be seen resting in a cut niche in the gebel, as if it had been used (with hundreds of others, now quarried away) in a built forecourt of that temple. This could suggest that it, too, was part of a New Kingdom version of B 200, now totally destroyed, or it could simply be a New Kingdom block reused by Taharqa's builders. A third temple utilizing talatat is the ruined Napatan-Meroitic B 1100, about 30 m to the east ("south"=right) of B 300, which has talatat foundations. The date of the earliest version of this structure is suggested by a block, bearing the throne name of Horemheb, which our Mission recovered in nearby rubble in 1999 (fig. 33) (See below, Section VIII). Finally, the 19th century gubba of Sheikh Ahmed Karsani, built in the Muslim cemetery less than 100 m west of B 200 and 300, is made almost entirely of talatat blocks, many of which bear traces of Egyptian relief and fragmentary cartouches of Ramses II (fig. 6). Obviously these blocks were taken from the nearer temples by local residents in relatively recent times. None of the blocks, however, bears any trace of relief in an overtly Amarna style.
Following the reign of Ramses II, there is no further evidence of New Kingdom building at Gebel Barkal. The only remaining datable New Kingdom object found on the site is a fragment of a statuette of a high official of Ramses IX (ca. 1131-1112 BC). Its fragments were found both in B 503 and in debris in front of B 700 (Reisner 1931, 81, no. 22; Dunham 1970, 29, pl. 28 c-f).
The above data indicate that Gebel Barkal had considerable importance and royal patronage throughout the New Kingdom. Although the Egyptian construction program appears to have ended with Ramses II, the site probably continued to function well into Dynasty 20, probably at least until the reign of Ramses IX. Not long afterwards, however, the cult seems to have been suspended; the temples were abandoned; and they rapidly fell to ruin. The evidence for this is suggested by several of Reisner's observations. He noted, for example, that the New Kingdom columns standing in B 506, directly outside the sanctuary of the temple nucleus, had been badly worn by blowing wind and sand at the time they were enclosed with new masonry by Piye (Reisner 1917, 220). This suggests that the temple had been roofless for some time. Since the king also completely rebuilt the old hyptostyle court (B 502) of Ramses II with a new configuration of 46 columns (at least 28 of them new), we can also draw the same conclusion here. The old columns must have been so worn as to be unsalvageable, requiring all new work. Finally, Reisner noted that a part of Taharqa's temple B 300 was erected directly over a collapsed wall of its New Kingdom antecedent (Reisner Diary, Boston, Jan. 28, 1919), which suggests again that the earlier temple was in a tumbled state when the latter was built. All these clues suggest that the Barkal sanctuary had been a site in decay for three centuries when the Napatan Dynasty suddenly emerged on the scene to rebuild it.
The first Napatan king who seems to have converted whole-heartedly to the Amun cult, judging by Taharqa's remarks in his Kawa stelae (FHN I 139-141, 173-174), was his great uncle Alara, who probably began his reign about 785 BC (Kendall 1999, 31-34, 58, 63-65, 97; Vinogradov 1999). It is most likely he to whom we can attribute the first Napatan temples, which were built of crude mud brick. One such structure was said to have existed at Kawa (FHN I 140); Reisner found another at Gebel Barkal that formed the first level ("B 800-sub") of the later Napatan Amun temple B 800/900 (Reisner 1920, 247-254).
B 800/900 was at first thought by Reisner to be two separate temples, hence the double numbering. As his excavations progressed, however, he realized that these structures had originally been two parts of a single temple. The first building was B 800-sub, to which was added on its "south" side B 900, a series of side rooms. This temple was built parallel to the old Amun Temple B 500, about 30 m to the "north" or downstream side of courts B 502 and 503. Although B 800/900 was smaller in scale than B 500, one can see from its plan that its builders tried to follow closely the original plan of the older temple, which would have been lying in ruins at the time the other was built. Its nucleus, like that of B 500, was a tripartite shrine, indicating sanctuaries for Amun, Mut, and Khonsu. As the temple grew forward with added courts B 802 and 801, side chambers (B 903-908) were built on its left (“south”) side. These chambers, including a hall bisected by a single row of columns, seemed designed to imitate rooms B 504a and b of the presumed Thutmosid Amun temple (B 500-sub). In its earliest mud brick phase, B 800-sub had the look of a structure built hastily as a working substitute for B 500 and a temporary shelter for three cult statues, while preparations were being made for the complete renovation in stone of B 500.
B 800-sub was built of crude unbaked mud brick, but its walls were founded here and there on pavements of reused talatat blocks (Reisner Diary, Feb. 6, Mar. 14, 27, 1920). Since a few of these stones were inscribed with the name of Ramses II, it is apparent that these pavements were constructed in early Napatan times with stones reused from nearby ruined buildings of Ramses, which themselves had been built of blocks reused from buildings of Akhenaten. Perhaps these stones were removed from B 500 in order to give the new Amun temple the sanctity of the old. Reisner noted that he was unable to find any foundation deposits associated with the new temple, so he assumed that the practice was unknown to the first builders. On the other hand, they may have believed that the old blocks gave it appropriate continuity with the old, negating its need for foundation deposits (Reisner 1917, 220-223).
No Napatan royal names were identified with B 800-sub, but Reisner speculated that it had been built by "Kashta or his immediate predecessor" (Reisner 1918, 254). Possibly the nucleus of B 800-sub (B 803-807) was built by Alara, and its outer courts B 802-801, by Kashta. Either king, however, could have built the entire temple. Later, during or after his restoration of B 500, Piye refurbished B 800-sub. In its outer court B 801, Reisner found an abacus for one of its columns bearing his name "Piye Snefer-Re" (Reisner Diary, Feb 29 and Mar. 3, 1920). Then again, in the later Napatan walls of B 900, Reisner found large reused sandstone lintel fragments bearing a building inscription of the king:
[P]('nh)y di 'nh dt ir.n.f m mnw.f n it.f Imn nb Nswt-T3wy hry-ib Dw-w'b kd.f pr.f s'h[.f sw] "[P]iye, given life forever, he made (it) as his monument for his father Amun, Lord of the Throne of the Two Lands. He built his house causing that [he] live…" (Reisner Diary, Mar. 13, 1920). Other reused stones from the same building preserved relief fragments of over-lifesize figures, including a partially preserved portrait of the king wearing an Atef crown (fig. 7).
Unique among the known Barkal temples, B 800-sub was constructed in a location where there had been no previous or underlying New Kingdom temple. During his probings beneath the earliest level of the temple, however, Reisner reported finding traces of even older mud-brick walls, some of them white-plastered (Reisner Diary, Jan. 14, Mar. 22-27, 1920). These were of a different plan than the temple and had a "north-south" orientation.
The palaces of the New Kingdom were temporary structures built of mud brick. The largest, most luxurious, were built at the capitals, but every major town had one for use as a lodging place for the king whenever he came to visit. The palaces were always built beside the most important temples and, other than living quarters for the king and his family, they were used as places where the king prepared himself for ceremonies, or rested between ceremonies and changed his ritual garb (Gohary 1992, 35). We know that they were regularly situated at a right angle to a temple's entrance. They were said to lie "on the starboard side" of the god's bark when it was carried forth from the temple (O'Connor 1989, 79). In other words, they were built immediately to the right of the entrance of a temple as one exited it (cf. Kendall 1997, 321, n. 4.). When a temple was enlarged and its entrance moved forward, the old palace was demolished and a new one erected perpendicular to the new entrance. Looking at the position of B 800/900 with respect to B 500, we realize that it occupied the very ground that would have been occupied by one or possibly two of the New Kingdom palaces associated with the earliest phases of B 500.
When the nucleus of the first post-Amarna Amun temple was built, probably by Horemheb, we may assume that a palace was built immediately "north" and perpendicular to its pylon. It may have occupied the site of an even earlier Thutmosid palace. When the entrance to B 506 was extended some 25 m by Seti I with the addition of court B 503, the old palace may have been replaced with another which would have faced the entrance of the third pylon. When Ramses II completed his hypostyle hall B 502, pushing the temple entrance forward another 55 m, the old palace would have been replaced with another, which would now have been perpendicular to the entrance of the second pylon. This, it will be seen, is exactly the position now occupied by the Napatan palace B 1200, which lies perpendicular to the entrances of both B 800/900 and B 502.
During his excavations of B 1200, Reisner actually found two blocks inscribed with the name of Ramses II in an excavation in the lowest levels of room B 1201 (the kitchen area of the Napatan palace) (Reisner Photo Register, negative C 8587). This leads us to suspect that the Napatan builders of B 800-sub and rebuilders of B 500 also labored to restore the ancient palace of Ramses II. Since the ruins of that structure would have been very conspicuous to the early Napatan rulers, we may suppose they planned the building of B 800-sub, the restoration of B 500 as far as B 502, and the restoration of the Ramesside palace as a single project.
Piye's renovation of B 500 seems to have taken place in two stages. The first, complete to court B 502, was probably initiated soon after his accession. The construction was probably completed before his Egyptian campaign of Year 20, and the decoration was added shortly afterwards. The inner walls, adorned with ritual scenes and the emergence of the bark of Amun from the temple, include a scene of the king, in overlarge scale, with his first "great royal wife," Pekereslo (Kendall 1999, 42-43, 116, fig. 19). The outer walls of the court were filled with a pictorial narrative of his Egyptian campaign (Kendall 1986, 9-20). The second stage was his addition of court B 501, which would have been initiated and completed after his campaign. The reliefs on the "north" wall depict his receipt of the tribute and homage of the vanquished Egyptian kings (Kendall 1997b, 164-165, fig. 28), while the reliefs on the "south" wall depict his building and dedication of the temples in Napata and his celebration of his Heb-Sed (Kendall 1999, 117, fig. 20, and see below Section X) (figs 47, 48).
By the end of Piye's reign, the Gebel Barkal sanctuary had at least two parallel operating Amun temples and a palace. There were surely other contemporary shrines that are now no longer extant or recognizable. The great temple of Amun of Napata (B 500) had been fully restored to its New Kingdom grandeur, and a huge new court (B 501) had been added to it (fig. 8). The work had probably been done largely by Egyptian masons and artisans supplied to the king by his sister Amenirdis, now seated at Thebes as God's Wife of Amun. B 800-sub, at first apparently thrown up by Alara (?) as a temporary shrine for use during the restoration of B 500, was now reconceptualized and seems to have become an important institution in its own right, being refurbished for this role by Piye with new stonework. These two Amun temples now apparently had the purpose of honoring the distinctive "southern" and "northern" aspects of Amun, whose separate images, criocephalic and anthropomorphic, began to grace the local monuments. It was to these dual aspects of Amun that Piye attributed his kingship at the very beginning of his reign (FHN I 57).
There may have been other standing temples at Barkal during the reign of Piye. Their existence is suggested by the number of important deities named on the surviving abaci of B 500 - deities for whom no local shrines are yet known. These gods may have had their own cult places on the site or else they may have occupied special shrines within the two temples. Among these gods are: "Onuris who is in Ta-Seti"; "Shu, son [Atum]"; "Amun-Re of Gem-Aten"; "Amun-Re of Pnubs"; "Horus, Avenger of his Father"; "Montu-Re, lord of Thebes"; "Atum-Re, lord of Heliopolis"; and "Ptah, foremost of Ta-Seti" (Dunham 1970, 55, fig. 40). Other deities, such as "Eye of Re"; "Bastet, daughter of Amun"; and "Tefnut, daughter of Re" were probably venerated in early Napatan precursors of the temples that later served the goddesses: B 200 and 300, built by Taharqa. "Dedwen, foremost of Ta-Seti" later apprently dwelt in Atlanersa's B 700, and surely there was an "Osiris" temple somewhere on the site. Piye's's mention of Weret-Hekau (FHN I 58), the crown goddess associated with temple B 1100 (see below, Section VIII and IX), suggests that some sort of a functioning sanctuary existed there also for her at the beginning of his reign. It was Weret-Hekau, after all, who crowned the king (Gardiner 1953, 15; Macadam 1955, 95, pl. 22)
One New Kingdom temple that may have survived nearly intact into the Napatan Period was B 600, built by Thutmose IV. This small shrine may also have been restored by Piye, for it is doubtful that, had it not been standing, B 700 would have been built where it was and not over that hallowed site. It appears, in other words, that B 600 was already standing when B 700 was built (in the later seventh century BC). The form of B 600 was as a raised stepped kiosk with a columned portico, the type of structure that seems to have been used by the king during his Heb-Sed and which is even pictured in Piye's reliefs in B 501 (figs. 9, 48). Inside the temple there is a stepped plastered podium for a statue or a throne; the walls are uncarved. One can hardly avoid the conclusion that this shrine was dedicated either to the living king, where he sat during ceremonies, or to his ka in the form of a statue (see below, Section VI).
Neither Shabaqo nor Shebitqo undertook any construction at Barkal, as far as we know, but Taharqa energetically beautified the site by creating the dual rock cut temples of Hathor (B 200) and Mut (B 300). The old ruined free-standing "B 300-sub" of the New Kingdom, with its shrines for three goddesses, Taharqa now turned into to a fine rock cut temple dedicated to Mut in her transformations (as "Eye of Re" from Sekhmet to Mut) (Robisek 1989). B 200 was dedicated to Hathor in her transformations (as "Eye of Re" from Tefnut to Hathor) and retained the tripartite plan of B 300-sub (Bosticco 1988). He also undertook the extraordinary task of carving and inscribing a panel on the summit of the pinnacle on Gebel Barkal and sheathing it in gold. He then lifted a small statue (probably of himself) to this great height and installed it in a niche just underneath the gilded text (Kendall 1994). The probable significance of this effort will be suggested below (Section XI). The ruined New Kingdom temple B 1100 and its now destroyed companion B 1150 were probably also rebuilt by Taharqa as part of the B 200-300 series. These, too, will be discussed at length below (Sections VIII and IX).
After Taharqa's reign, no new building was added to the Barkal site until the reigns of Atlanersa (ca. 654-640 BC) and Senkamanisken (ca. 640-620 BC). These kings, one after the other, both labored on B 700. The first king initiated it (with foundation deposits) and the second completed it. If B 600 was a shrine to the living king or his ka, as suggested above, B 700 seems to have served as a royal mortuary temple dedicated to the Osirian forms of Amun, the god Dedwen, and the deceased king as Osiris (Reisner 1918, pl. 16) (fig. 9). It is doubtful, however, that it was originally designed with this meaning; the temple probably assumed its mortuary significance only upon Atlanersa's unexpected death. Votive bronze figurines found in the sanctuary (B 704) represented Osiris (Dunham 1970, 69, fig. 47), and statues found in the sanctuary represented a ram-headed Amun, Amenhotep III, and several other Meroitic kings in poor condition (Ibid, 69, pl. 57). Fragments of a large funerary-style false door with reliefs of the canopic gods were found by our team in 1987 in B 703. Like the Theban mortuary temples, B 700 probably integrated the cult of the dead king directly into the main cult of Amun, but unlike them it seems to have served all rulers, who at death became Osiris. The fragmentary reliefs in the outer court B 702 might have been those found in any Amun temple, for they depict on the "south" wall the presentation of offerings by the king (Senkamanisken) and his chief queen to Amun and Mut. On the "north" wall the same king was doubtless pictured, followed by the royal women (Wenig 1978, 58, fig. 33), greeting the emergence of the bark of Amun from the temple. The fragments of the bark scene are still unpublished. The superb bark stand from B 703 is now in Boston (MFA 23.728: Reisner 1918; Dunham 1970, 67-74, pls 30-31).
At this point the Napatan sanctuary of Gebel Barkal was near completion (fig. 10). Under Anlamani, B 800/900 seems to have been rebuilt entirely in stone,11 and the sanctuary was now remodeled to house only one deity, doubtless "Amun of Karnak". There was no longer a need for a tripartite shrine with Mut and Khonsu, for each god now probably had its own temple. Certainly Mut was housed in B 300. B 900 (Khonsu?) became a separate small temple on the "southeast" corner of B 800.
Both Anlamani and Aspelta also completely rebuilt the palace B 1200, which in their time was probably well over a century old and in need of renewal. This palace now included, besides a new throne room, a large, elegant sanctuary for celebrating New Year 's rites (Kendall 1997a, 324-334).
During Aspelta's reign, the palace as well as B 500 and B 800/900 were destroyed by fire, which was set deliberately. The palace was so badly burned that its rooms were simply filled solid with earth by the builders of the next generation and used as foundations for the renewed palace (mid-sixth century BC). The wooden roofs of the Amun temples were also torched and the statuary within toppled and broken. It is hard to imagine any other cause for this destruction than the invasion of Kush in 593 BC by the army Psammeticus II. Since much doubt has recently been expressed that the Egyptian army ever reached Gebel Barkal, I will try once again to convince the doubters (Section XI).
Because the primary intent of this paper is to examine the relationship between the New Kingdom and the Napatan Period through the record of Gebel Barkal and to understand the mountain's cultic significance, I will suspend my description of the site here in order to summarize what conclusions we can draw from the preceding. It should be abundantly clear by now that Gebel Barkal was extremely important long before the rise of the Kushites in the eighth century BC. In fact, it was the site's very importance and meaning during the New Kingdom that led the Kushites to restore it. The Kushites, in other words, did not give the site importance. It already existed. It was the site that gave them importance. By restoring the site and revivifying its god, they were able to justify their claims to the Egyptian throne.
Surprising as it may seem, the site was probably visited by most or all of the pharaohs from Thutmose I to Ramses II, all of whom, at least from Thutmose III, patronized it and actively built temples and erected statues and monuments there. How often they visited it is unclear. It remains for me now to try to suggest what meanings the pharaohs may have assigned to Gebel Barkal, and to show what effect these meanings may have had not only on subsequent Egyptian history but also on the formation of the Napatan state in the Sudan.
Good archaeological evidence persuades us that the Barkal site ceased to function as a cult place at about the time the New Kingdom ended, sometime during the eleventh century BC. The temples were evidently abandoned and remained untended and derelict throughout the Third Intermediate Period. The Napatan revival of the site and the rebuilding of the temples did not begin until the early eighth century BC. Despite this three-century hiatus in the site's use, we see from the surviving record that there was no loss of memory of the site's cult during this period, for it was fully restored under the Napatans. For the most part, the Kushite kings built their new temples directly over the foundations of the old Egyptian temples and revived the worship of the same gods. They also restored the old Ramesside palace and reoccupied it. The Egyptian statues and stelae that had been erected on the site they re-erected in the new buildings, and they modeled their own statues and stelae after these. It was as if, through the restoration of the Barkal sanctuary, they could present themselves confidently to the wider world as the direct successors of the pharaohs of the New Kingdom.
The great imponderable in all this is that prior to the early eighth century BC there was no obvious "Egyptianization" of the Nubian royalty. As revealed by their graves at el-Kurru, the Napatan chiefs of the mid-ninth century BC were still adhering to traditional Nubian burial practices. They were interred lying on beds in side-chamber pits under tumuli. Nor apparently were they literate. They had little or no knowledge of Egyptian language and writing. They erected no stone monuments, built nothing in stone, and had no developed art that we are aware of. They had no pretensions to kingship in the pharaonic style, nor did they honor, in more than a superficial way, the Egyptian gods (Kendall 1999a; 1999b). These facts alone would seem to confirm that there were no Egyptian temples functioning in the region and that there was no active local priesthood present that was maintaining or disseminating Egyptian religious beliefs, especially among the royal family. There is no evidence, at least initially, for any royal patronage of a cult, nor for any cultic support of the dynasty. If the Nubian chiefs of el-Kurru had never set foot in Egypt, and if there was no local Egyptian community to acculturate them, we might wonder why they suddenly abandoned their native burial customs and adopted the Egyptian. Why did they adopt the Amun cult with such a passion, when it had been dead in Nubia for at least three centuries? Why did they begin to venerate the other Egyptian gods, and to use the Egyptian language and writing for their formal inscriptions? Why they adopt the Egyptian royal style and rebuild the temples at Gebel Barkal and elsewhere in Nubia with such single-mindedness? By the time of Piye - hardly more than a generation beyond the first Napatan ruler known even by name - the "Egyptianization" of the Napatan ruling family was so complete that they did not even look back to a pre-Egyptian past. All models for their state and kingship had become Egyptian, and the pharaohs of the New Kingdom had become their "ancestors."
What and who had caused these changes? And why? Obviously, if there were no existing Nubian groups that could have influenced the early rulers, the changes can only have been effected by outsiders - from Egypt (Kendall 1999a, 49-77). If Egyptians came to el-Kurru to missionize the Nubian dynasts and to convert them to the Amun cult, would it not also be correct to assume that these were the same individuals who encouraged and directed the Kushite revival of the Barkal sanctuary and the other Nubian sanctuaries of Amun? Furthermore, if the new kings of Napata believed their royal power came from Gebel Barkal and that it was this same power that had belonged the pharaohs, would it not be likely that they had learned this tradition from the same Egyptian visitors? If this tradition was Egyptian rather than Nubian, then shouldn't we assume that during the New Kingdom the pharaohs, too, had believed their royal power derived from Gebel Barkal? This last question will sound almost preposterous, but let us consider the evidence.