Middle Kingdom Studies
Edited
by
Stephen Quirke
51
.1
SIA P U B L I S H I N G 1991 Nacterlands Instituut vsor ket Nabije Oosten Laidan - Nodarland
First published in 1991 by SIA Publishing, 31, Maiden Way New Maiden Surrey KT3 6EB
British Library Cataloguing in Publication Data Middle Kingdon studies. I. Quirke, Stephen 932 ISBN 1 872561 02 0
© SIA Publishing
Printed in England by: Whitstable Litho Printers Ltd., Whitstable, Kent
Contents Preface Janine
Bourriau
Patterns of c h a n g e in burial c u s t o m s during t h e M i d d l e Mark
Kingdom
Collier
C i r c u m s t a n t i a l l y a d v e r b i a l ? the c i r c u m s t a n t i a l sdm(.f)lsdm.n(.f) reconsidered Detlef F r a n k e T h e career of K h n u m h o t e p III of Beni H a s a n and the so-called " d e c l i n e of the n o m a r c h s " James
K.Hoffmeier
T h e coffins of t h e M i d d l e K i n g d o m : the R e s i d e n c e and the r e g i o n s K.A.Kitchen N o n - E g y p t i a n s r e c o r d e d on M i d d l e - K i n g d o m Stelae in Rio de Janeiro R.B.Parkinson T e a c h i n g s , d i s c o u r s e s and tales from the M i d d l e Stephen
Kingdom
Quirke
R o y a l p o w e r in the 13th Dynasty Pascal
Vernus
Sur les graphies d e la formule "l'offrande q u e donne le r o i " au M o y e n E m p i r e et a la D e u x i è m e P e r i o d e I n t e r m é d i a i r e
Preface In April 1988 Darwin College Cambridge hosted an Egyptological colloquium on Middle Kingdom studies under the title "the Residence and the Regions", organised by Janine Bourriau. The gathering of specialists provided a forum for discussion of many points highlighted by the exhibition "Pharaohs and M o r t a l s " mounted by Janine at the Fitzwilliam Museum. Since that double triumph three years ago Janine has retired from her position as Keeper of Antiquities at the Fitzwilliam Museum; in a small way the papers that join hers in this volume testify to her energetic role in Egyptology and her untiring assistance to colleagues of whatever station. As editor and as a student who benefited and continues to benefit from her guidance I take the opportunity of this publication to offer on behalf of the others attending the colloquium and contributing to this volume joint best wishes for a long and bright future in promoting our subject and assisting colleagues. The editor and contributors owe a particular debt of thanks to Arabella Binney and to Alan May for their generous subvention of this v o l u m e . The contributors to this volume include five of the speakers from the 1988 colloquium as well as the organiser herself. In addition research fellows Dr.Mark Collier of Corpus Christi Cambridge and Dr.Richard Parkinson of University College Oxford have added articles on language and literature, two areas of Middle Kingdom studies not covered by the original colloquium. The final product forms it is hoped some fitting tribute to the industry of Janine Bourriau.
Stephen Quirke April 1991
1
Patterns of change in burial customs during the Middle Kingdom.
by Janine Bourriau
From the Instructions of Prince Harjedef: "When you prosper, found your household, take a hearty wife, a son will be born to you. It is for the son you build a house. When you make a place for yourself, make good your dwelling in the graveyard. Make worthy your dwelling in the west" 1
The twin poles of an ancient Egyptian's ambition, to establish a family and build his tomb, are nowhere more concisely expressed. It is this unchanging and determined preoccupation with provision for the afterlife which allows us to assume that changes in burial customs accurately reflect changes in society, whether social, economic or political. In other words, so much time and so many resources went into the preparation of equipment for the tomb, the arrangements for the burial, and the funerary cult, that changes in them, which we can observe, must reflect and so help us to understand wider changes in society as a whole which are otherwise more difficult to identify. From specialist studies of particular sites, tombs, or classes of object, we can begin to see an evolution in burial customs in the Middle Kingdom. I hope to suggest some ideas and approaches to the subject,without attempting an exhaustive survey of the sources, inappropriate to this volume of short studies. I shall concentrate on the material from the Memphis-Fayum region and Middle Egypt and the period from the 2
3
1
M. Lichtheim, Ancient Egyptian Literature VolJ. The Old and Middle Kingdoms (Los Angeles, 1975), 58. ^ The bibliography is extensive, and the LÄ together with the most recent issues of the PEB provides the most convenient access to iL 3 The starting point for this paper was a lecture delivered at Emory University in Atlanta,at a conference on "A Sense of Place: Regional Art and Archaeology of First Intermediate Period and Middle Kingdom Egypt" in October 1990. It was organised by Dr. Gay Robins and the University Museum of Art and Archaeology in conjunction with a loan exhibition, "Beyond the Pyramids. Egyptian Regional Art from the Museo Egizio, Turin". I should like to take this opportunity to thank Gay Robins for inviting me to give the lecture and to benefit from the discussions which took place.
3
Janine Bourriau
4
beginning of the XII to the early XIII Dynasty, for the practical reason that this has been the focus of the most recent research and therefore provides the fullest and most reliable data. When considering burial customs we always have to accept that the tangible remains are not the whole story, even in those rare cases where the burial is found intact. The preparation of the body, the procession to the tomb, words spoken, rites carried out at the entrance or inside the burial chamber: all these activities leave little or no trace and yet may have been considered essential in ensuring safe passage through death to the afterlife. Establishing which changes in society may be reflected in changes in burial customs is full of pitfalls. Until recently it was argued, following Moret's classic exposition, that the adoption for non-royal burials of Coffin Texts (derived ultimately from the funeral liturgy of the Kings of the Old Kingdom), together with the inclusion of some items of royal regalia among the painted "Frise d'Objets" on rectangular coffins, and the modelling of them on anthropoid ones, should be understood in terms of the growth of the individual's political power at the expense of that of the King.The breakdown of order, combined with the famines of the early First Intermediate Period, had (it was argued) undermined faith in the King's power to maintain the divinely ordered sequence of the seasons and the institutions of public life, and as a consequence undermined also the belief in his control over the afterlife. This led to "democratization" of the afterlife, in which the individual, independent of any King, sought his own deification through identification with Osiris, using rites and prayers previously restricted to Royalty. A re-dating of Coffin Texts to the Middle Kingdom, rather than the First Intermediate Period, that is to a time when the monarchy was strong, not weak, has prompted a reappraisal. These changes are now thought to be an expression only of the private person's desire to claim identity with Osiris after death, rather than an attempt to usurp the privileges of living Kings. It is significant that only images of or for the use of the dead adopt royal attributes: sceptres, crowns, garments, titles and modes of address. They continue to be the exclusive preserve of royalty in sculpture representing the living. Yet if the earlier hypothesis grew out of a preoccupation with the relationship between the individual and the state, there is implicit in the new one a modern assumption about the separation of the spiritual from the political side of life, a division without meaning in Ancient Egypt. There is no doubt that the change in burial customs did occur, that it was very profound, and that it reflects an equally deep change in men's views of their relationship to the King, both in this life and in the afterlife. If this change took place in the Middle Kingdom rather than the First Intermediate Period, then it is in the society of that time that explanations must be sought 4
5
6
4
However, H.O. Willems, Chests of Life (Leiden, 1988), 14Iff; 238-244 and bibliograhy there cited, interprets the ornamentation on "standard class" coffins as an "account of the ceremonies on the day of burial". This paper owes a great deal to Willems' careful analysis of Middle Kingdom coffins, as will be seen from the numerous citations to this work which follow. A. Moret, "L'Accession de la Plèbe Egyptienne aux droits Religieux et Politiques sous le Moyen Empire" in Recueil d'études égyptologiques 8 (1922), 331-360; also H.Kees, Totenglauben und Jenseitsvorstellungen der alten Ägypter (Leipzig, 1926), 164-9. 6 Willems op.cit., passim but especially 244-249. Not all scholars would agree to this redating, see J. Hoffmeier and bibliography cited in this volume. 5
Patterns of change in burial customs
5
The evolution of burial customs can be visualised as forming three interlocking patterns which relate to geography, chronology, and status and number of people being buried or commemorated. First, we can examine the geographical pattern. All the important Middle Kingdom sites had their only, or major period of excavation early in the history of Egyptology and, sadly, in the history of archaeology, between 1890 and 1914. Significant exceptions are Aswan, the Delta, and the Memphis region, but the recent excavations in these regions still await completion and/or full publication. During this initial period of excavation, and right up until the 1960's, the traditional view was that Egypt was culturally unified after the political unification by Nebhepetre Mentuhotep in Dynasty XI, and continued to be so until the end of Dynasty XII. It followed from this that graves at Qau in Upper Egypt could be compared with graves from Harageh in the Fayum and differences between them interpreted in terms of chronology rather than of regional variation. Specialist studies of individual regions or classes of object, particularly tomb biographies, stelae, coffins and pottery, have changed this view, and most scholars would now agree that Egypt's material culture in the Middle Kingdom can be observed only in terms of the cultures of the Delta, the Memphis-Fayum region, Middle Egypt (to Asyut), Upper Egypt and Aswan. From the end of the Old Kingdom onwards each of these regions - and in some cases even smaller districts within them - evolved its own craft traditions in the manufacture of funerary monuments and equipment; for each region, a separate sequence must be established. The gaps in our knowledge are such that we can often identify a coffin or stela as belonging to a particular provincial workshop but be unable to assign it to a precise date within that workshop's history. Coffins, being difficult to transport and yet not requiring, like tomb reliefs or sculpture, the special skills of masons or sculptors who could travel from one commission to another, are most striking expressions of local workshop traditions. Furthermore, if the coffin is interpreted as a "ritual machine" i.e. as a medium through which what was spoken or acted out once could continue to be efficacious for eternity, then the evolution of its inscriptions should directly reflect changes in the rites of burial. Pottery, on the other hand, suffers from the complication that while much of it may have been made locally, some specialist vessels, containers for traded commodities, were transported widely. It is not always possible to distinguish the local product from 7
8
9
7
Willems opxit., 161, suggests that coffin painters did occasionally travel and cites a coffin from Beni Hasan whose decoration matches very closely coffins from Bersheh. The art historical analysis such as W.K.Simpson has begun for stelae in Terrace of the Great God at Abydos: The Offering Chapels of Dynasties 12 and 13 (New Haven and Philadelphia, 1974) has not yet been embarked upon for coffin decoration. 8 See for example the fragments of an unprovenanced coffin in the Fitzwilliam Museum, E.W.66a,b. J.Bourriau, Pharaohs and Mortals. Egyptian Art in the Middle Kingdom (Cambridge, 1988) Cat.no.67. It has been assigned to Asyut and to the reign of Amenemhet II on the evidence of the owner's name, Wepwawetemhet, and the pattern of three vertical text columns between the panels and four vertical text columns at each end of the long sides. The design is unique to Asyut. The alignment in the published photograph should be corrected. The two pieces belong side by side with a gap the width of one panel and two vertical text columns between them. Willems op.cit., 239. 9
6
Janine Bourriau
the import on shape alone, and many excavation reports provide no other information. Centres within these regions - places like Asyut, Aswan, Bersheh, Beni Hasan, and Gebelein - were immemorially ancient Nome capitals or cult centres, inheritors of local traditions going backiperhaps even beyond the Unification under Menés. In addition to these places there were three significant urban centres: the Residence, Iit3wy, probably to be identified with Lisht; Thebes, the centre of the administration of the Southern part of Egypt and the heartland of the XI and XII Dynasties; and finally Abydos, centre of pilgrimage, the burial place of Osiris. Much of the history of the first part of the Middle Kingdom, up till the end of the reign of Senwosret HI, can be seen in terms of the interplay between on the one hand these three cities, dominated by the activities of the King and his officials, and on the other the provincial centres controlled by dynasties of Nomarchs and their dependents. Why did Amenemhat I choose to establish what was in effect a new capital city, a new royal Residence at It-t3wy, between Memphis and the Fayum, ignoring Memphis itself? It may have been because his own reign was not secure, if we accept at face value the Instructions he left for his son Senwosret I, which tell him: 10
"Trust not a brother, know not a friend, make no intimates, it is worthless. When you lie down, guard your heart yourself for no man has adherents on the day of woe." 11
The text goes on to describe an assassination attempt, recalling perhaps the event which precipitated Sinuhe's flight to Syria in his famous narrative. However, an additional reason may have been that Amenemhet I, like many founders of new Dynasties before and since, wanted a new capital.He called it, significantly, It-t3wy, "Seizer of the Two Lands". (Nearby Heracleopolis, close to the mouth of the Fayum, had been the capital of a rival Dynasty during the First Intermediate Period, thereby emphasising the importance of this region for the control of Northern Egypt). In contemporary Egyptian texts, It-t3wy is often called simply "The Residence", a phrase sufficient to designate the royal palace, the residence of the King being inevitably the seat of central government. All the officials and institutions set up there were dependent on the new King, inheritors of no ancient traditions (Memphis), or rival loyalties (Heracleopolis). The site of this capital has not been identified for certain but must lie in the cultivation near the modern village of el Lisht, close to the place chosen for the Pyramid complexes of Amenemhet I and Senwosret I, which remained an important necropolis for the rest of the Middle Kingdom. The significance of the Residence, its total identification with the State, is illustrated by the fact that the Middle Kingdom may be considered to end with the abandonment of the Residence and the withdrawal of the King southwards to Thebes. This event took place some time after the middle of the XIII Dynasty, in response (we surmise) to the increasing power of local rulers, more particularly the Hyksos based at Avaris (Tell Dab'a) in the Eastern Delta. 12
1 0
This is a problem encountered in establishing the distribution of storage vessels of Marl C fabric, made of a clay from the Memphis-Fayum region, into Middle Egypt, Nubia and the Delta, bypassing Upper Egypt; see J.Bourriau, "Nubians in Egypt during the Middle and New Kingdoms" to be published by the British Museum. Lichtheim opxit., 136; S. Quirke in Bourriau, Pharaohs and Mortals, Cat.no.60. W.CHayes, "Horemkha'uef of Nekhen and his Trip to It-towe" in JEA 33 (1947), 3-11. 1 1
1 2
Patterns of change in burial customs
7
The region around the Residence, from Dahshur in the North to and including the Fayum, was a focus for royal activity throughout the XII and XIII Dynasties. Here the kings were buried: there are Pyramids of Amenemhet II, Amenemhet III and Senwosret III at Dahshur, Senwosret I and Amenemhet I at Lisht, Senwosret II at Lahun, and Amenemhet III again at Hawara. To the cemeteries around the Pyramid complexes came the great officials of the state, the vizirs and treasurers, desiring to be buried close to the King they served. In this way they had access, by favour of the King, to all the resources of the royal workshops. Sinuhe describes the favours showered on him by Senwosret I, "A stone Pyramid was built for me in the midst of the Pyramids. The masons who build tombs constructed it. A master draughtsman designed it. A master sculptor carved in it. The overseers of construction in the Necropolis busied themselves with it. All the equipment that is placed in a tomb shaft was supplied. Mortuary priests were given me. A funerary domain was made for me. It had fields and a garden in the right place as is done for a Companion of the first rank." The attraction of such patronage in this world and the next must have been great: it was sufficient to cause the successor of Khnumhotep II to abandon his unfinished tomb at Beni Hasan in favour of a mastaba at Dahshur, close to the Pyramid of Amenemhet III. We can perhaps assume that such burials of the highest officials and their families, i.e. of those people closest to the King, would have been the first to respond to change. In the early XII Dynasty each King moved his mortuary complex to a new site, and whilst retaining the Royal Pyramid as centrepiece, rearranged the elements of Mortuary Temple, Valley Temple, subsidiary Pyramids and private mastabas. Such architectural innovation and experiment would have provided a climate encouraging to changes in burial rites and equipment It is significant that the two most popular designlayouts for the external decoration of "standard class" coffins, Willems' layouts of types IV and VI, may originate in the cemeteries around the Residence. Two Lisht coffins are among his earliest sources for type IVaa, and type VI is thought to derive from the decoration on the royal sarcophagi of Senwosret II and III and Amenemhet HI at Lahun and Dahshur. Similarly, in pottery studies Dorothea Arnold has recently shown that the Riqqeh-Lahun-Harageh corpus thought to be characteristic of Dynasty X I I originated at Lisht in the second half of the reign of Senwosret I . Given the different ceramic traditions prevailing in Egypt in the early XII Dynasty, it is possible to chart the spread of this "Residence" pottery tradition as it reached the rest of Egypt and Nubia. It is significant that it arrives in Nubia before it appears at Thebes and becomes dominant throughout the country only in the late XTJ Dynasty. Until then the degree to which it is present on a provincial site can be used as one indication of the strength of contacts between that site and the Residence. Thebes had a different history in the Middle Kingdom, in so far as we know it. It is rich in private and royal funerary monuments from the XI dynasty until the reign of Senwosret I; thereafter, fewer monuments survive. The obvious interpretation is that 13
14
15
16
17
18
*3 Lichtheim op.cit., 233. * The evidence, which is frustratingly incomplete, is summarised in Willems op.cit., 61-2, n.19. Willems op.cit., 160, n.138.. Ibid., 162, n.147. B J. Kemp and R.S. Merrillees, Minoan Pottery in Second Millennium Egypt (Mainz, 1980), 23. Dorothea Arnold in Dieter Arnold, The Pyramid of Senwosret I (New York, 1988), 143-146. 4
1 5
1 6
1 7
1 8
8
Janine Bourriau
private cemeteries become more modest, earlier cliff tombs are re-used and simple shafts without superstructures dug, until the mid-XIII Dynasty when the centre of royal power shifts southwards again. However, there is an exceptional amount of unpublished evidence of this date, especially from the Metropolitan Museum of Art excavations at Thebes, which may change this picture radically. Abydos was the sacred city, home of Osiris and the site of his tomb, identified in the Middle Kingdom with the tomb of King Djer of the First Dynasty. The Abydos necropolis was continuously used for the burials of people who lived locally or who were attached to the site's numerous shrines, but in the course of the XU Dynasty it became increasingly important to be commemorated there, regardless of where you were buried. People wished to dwell near the god in the same way as they might wish to be buried or commemorated close to the royal tomb enclosure. Statues, offering tables and stelae were set up in small chapels. These might belong to families, or to groups of people whose work drew them together, and as officials passed up and down the river secure in the stability of a unified country, we may imagine them calling in at Abydos to arrange for a commemoration to be left for them and their families, patrons or proteges at the Terrace of the Great God. Hundreds of stelae were the product of this piety; publication and analysis of this rich documentation is one of the most important recent advances in Egyptology, giving us all the more reason to regret that so few of these stelae can be restored to the monument from which they were extracted and so related to others of the same date or family group. Nor, with very few exceptions, can they be related to the contents of the extensive cemeteries of shaft tombs at Abydos. If we could relate burial equipment and stelae as we can begin to relate sculpture and stelae, Egyptian archaeology would be immensely enriched. The regional styles that can be observed in the early Middle Kingdom were not static, either in themselves or in their relationship to each other. As the XII Dynasty progresses, a general increase in wealth is apparent, judging simply by the quality of the raw materials used in private and royal monuments. Wood gives way to stone in sculpture in the round, and modest burials contain gold and hard stone jewellery and amulets. With this wealth comes a shift of resources away from the provinces towards the Fayum region, around the Residence. This is most clearly seen in Middle Egypt, at sites like Meir, Beni Hasan, Bersheh and AsyuL They had all ceased to exist by the reign of Amenemhet III, whereas the main period of use of the Harageh, Lisht North and Dahshur cemeteries came in the late XII to XÏÏI Dynasties. 19
20
21
22
23
24
1 9
A. Leahy, "The "Osiris' Bed Reconsidered" in Orientalia 46 (1977), 424-434. 20 There is a great increase in private stelae at Abydos during the reign of Senwosret I and again in the late XII-early XIII Dynasties; Simpson op.cit., passim. 21 Simpson op.cit., passim; D. Franke, Personendaten aus dem Mittleren Reich (Wiesbaden, 1984), passim. 22 For example at Elephantine, L. Habachi, The Sanctuary of Heqaib (Mainz, 1985). 23 it could be argued that coffins provide similar dating criteria to stelae: names, titles, funerary formulae, religious texts and motifs and design of decoration. However, the number of deposits where coffins and grave goods may be associated is suprisingly small.with the important exception of the shaft tombs at Beni Hasan. Nevertheless it is noteworthy that Willems hardly ever appeals to associated objects for dating coffins. He does not provide tomb numbers in his documentation of them and this emphasises that his approach isolates the coffin from the physical context of burial. This is inconsistent, perhaps, with his aim of understanding the texts on coffins in terms of the rites of burial. Bourriau op.cit., 127, Cat.no. 159, a gold catfish from Harageh Cemetery A 72. 2 4
Patterns of change in burial customs
9
The process is twofold. There is a growing cultural homogeneity expressed in the style of coffins and pottery and the selection of grave goods, and there is a movement of people, either to a different part of the cemetery or away from the cemetery altogether. What happened is not easy to understand. Why did Asyut cease to function before Beni Hasan did, for example? Despite the uneven pace of this change, there are obvious phases in the transitional period. A lesser change is marked by the end of the reign of Senwosret I, a greater one by the end of the reign of Senwosret III and the final phase ends with the abandonment of the Residence in the XHI Dynasty. It is the cultural homogeneity of the late Middle Kingdom which contrasts so strongly with the First Intermediate Period or the first phase of the Middle Kingdom, ending with the reign of Senwosret I. At the end of this first phase, local traditions in funerary equipment and sculpture begin to give way to artefacts without such a strong local character. Asyut exhibits the trend most clearly, as Diana Magee has shown in her unpublished thesis. The burial of Djefaiha'py, who served King Senwosret I, contains the latest known example of a coffin type unique to Asyut, part of a tradition of coffin decoration, defined by motifs, inscriptions and their arrangement, going back to the early Heracleopolitan era. From a period contemporary with Djefaiha'py's coffin, and a little later, coffins of Willems' type IVaa occur at Asyut itself, at Bersheh, Beni Hasan, and Lisht. This trend towards standardisation in the exterior decoration of "standard class" coffins, which seems to me to begin at this time, is reinforced by a change which also occurred a little later, in the reign of Amenemhet II, in their interior decoration. Willems' type 1 is replaced by his type 2, and in the new type the similarities between coffins from different sites is so strong that Willems suggests that it can be accounted for only by contact between coffin workshops. Willems' documentation has to be used with care because he analyses only what he calls the "standard class" coffin, and not coffin types whose occurrence is confined to a single site. His purpose is to establish dating criteria for coffins with Coffin Texts by comparing coffins from different sites, and such "deviant" types as he calls them, (or"Siutian" for those from Asyut) are a dead end in this respect. These are, however, all-important in showing the strength of local traditions in funerary ritual and coffin craft. What the sources, uneven though they are, do show in respect of coffin decoration, is that the "standard class" coffin with interior decoration was more popular 25
26
27
28
29
30
31
32
25 Bourriau op.cit., 10-12. 26 Diana Magee, Asyut to the end of the Middle Kingdom: an historical and cultural study. University of Oxford, 1990. Donald Spanel's unpublished thesis, which I have not seen, should also be consulted, see J.Hoffmeier, this volume. 2 Willems op.cit., 136-7, with fig.10. The earliest examples are listed on p.160, n.l38.See also above n.16. 281 have already suggested above, that Lisht may be the source of the layout design. Willems op.cit., 190. 30 Ibid., 49-50. This is G .Lapp's Lower Egyptian type in LÄ,V, 430-434. Lapp's account is more comprehensive in scope though necessarily less detailed than Willems' work. See also J.Hoffmeier, this volume. 3* As a consequence he lists all coffins with interior decoration, List 1, p.19-34, but only a limited number of those without, List 2, p.35-40. 32 in Willems List 1, under Asyut, there are 69 coffins, of which 46 are "Siutian". 7
2 9
Janine Bourriau
10
at Saqqara, Bersheh, Meir, Beni Hasan and Lisht than at Asyut and (not suprisingly, given the distance involved and the strength of local craft traditions) Thebes. Too little has survived from other Upper Egyptian sites for comparisons to be useful. Nevertheless it would seem that local traditions in coffin decoration were more tenacious at some places than at others. The relationship of the site to the Residence may be a factor, not just in terms of distance but also in terms of the degree of contact with the King and his officials. However, until the early XII Dynasty coffin types used in the cemeteries at Lisht are published and can be compared with those from the great cemeteries in Middle Egypt, the situation cannot be completely understood. The strength of local styles also varied from craft to craft: compare for example, the evolution of pottery style with coffin typology. Dorothea Arnold has shown how the classic Middle Kingdom pottery repertoire was a creation of the Lisht potters in the second half of the reign of Senwosret I . She has also shown how types-of the First Intermediate period continued to circulate longer at Asyut and Beni Hasan than at Sedment or Gurob. This phenomenon may explain why Seidlmayer, in his seriation of the Beni Hasan tombs, placed those groups with classic XII Dynasty pottery in his latest level, Level HI, dated by him to the reigns of Amenemhet II-Senwosret I I . The pottery from Beni Hasan has a character all its own, and moreover, includes shapes circulating in Upper Egypt as well as examples of the "Residence style". The potters whose workshops were immortalised in the Beni Hasan tomb paintings produced a funerary pottery altogether more idiosyncratic in style and technology than the products of the coffin makers. This may also have been true at Bersheh and Meir,but the evidence for the first site is unpublished and for the second lost. The evidence suggests at present that the new pottery style contemporary with the type IVaa coffins, introduced at the end of the reign of Senwosret I, was, unlike the coffins, confined at first to the cemeteries in the immediate vicinity of the Residence, i.e. Lisht itself and Riqqeh. It is the reign of Senwosret III which introduces the final phase of the Middle Kingdom. We can see changes in that reign which go far beyond burial customs and touch all aspects of the material culture, language and institutions of society. Such a profound cultural change did not come about simply through the King's political will, as the culmination of Senwosret Ill's policy of centralisation of power, but seems to be more deep seated and unselfconscious. Stephen Quirke has argued that Senwosret Ill's concern to re-define the boundaries of Egypt may have been the catalyst for this profound social change. He fixed the southern border at Semna, and cut a channel through the rocks of the First Cataract so that, for the first time, the Nile was navigable even at low water from the Second Cataract to the Mediterranean. The effect of this on communications with Nubia is easy to see. Two examples may suffice: the soldiers of the garrison of the late XTJ Dynasty at Buhen were buried with 33
34
35
36
37
38
3 3
The coffins from Lisht South and Lisht North are currently being prepared for publication by James Allen of the Metropolitan Museum of Art. See note 18 above. Willems op.cit., 144, n.333. 16 SJ.Seidlmayer, Gräberfelder aus dem Obergang vom Alten zum Mittleren Reich, (Heidelberg, 1990), 233. This book reached me too late to be fully incorporated into the present paper. 37 LBourriau, Umm el Ga'ab. Pottery from the Nile Valley before the Arab Conquest (Cambridge, 1981), 60-63. Cemetery A at Riqqeh began in the reign of Amenemhet II, Arnold op.cit., 143, n.329. 3 4
3 5
Patterns of change in burial customs
11
grave goods indistinguishable in appearance from those of their contemporaries buried at Lisht or Dahshur over a thousand miles to the north. Marl clay storage jars typical of the Memphis-Fayum region can be found all over Nubia, as far south as Kerma. Since these jars are never found in Upper Egypt, they are evidence of direct trade between the region around the Residence and Nubia. In terms of burial customs, the changes manifested themselves in two ways, both in a movement away from the provincial cemeteries which had evolved after the Old Kingdom and in changes in the type and class of object placed in the burial. After the reign of Senwosret III no evidence is forthcoming from Beni Hasan, Meir, Bersheh, and Asyut. Provincial cemeteries of people at a lower level of society those at Diospolis Parva, Qau cemetery 7000, El Kab, and Rifeh, for example - show continuity of use throughout the Middle Kingdom, in the locality if not always in exactly the same spot; those of the Nomarchs and their dependents do not. Inscriptions, scarabs and coffins, however, are relatively rare and the names and titles of few individuals survive. It seems that the majority of the people buried in these graves were not members of the official classes. By every available criterion they were of lower status than, for example, most of those buried in the cemeteries of comparable date in the Memphis-Fayum region, or at Abydos or Thebes. There are also changes to be seen in the actual contents of non-royal burial groups of the late XII to XIII Dynasty. The most characteristic product of the funerary workshops of the early XII Dynasty, apart from coffins, was wooden tomb models. They ceased to appear in the Memphis-Fayum region after the reign of Senwosret n, with the occasional exception of boat models, which continued until the mid- XII Dynasty. Brunton's original observation during his excavations at el-Lahun has been amply confirmed by Angela Tooley's comprehensive study. Like the classic XII Dynasty pottery style, this change in burial practices did not reach some provincial cemeteries like Meir and Beni Hasan until later. As a consequence we find in Beni Hasan tomb 500 - an intact burial - a coffin with outer decoration of Willems' type VI, introduced between the end of the reign of Amenemhet II and early in the reign of Senwosret HI, and inner decoration of type 2, together with pottery of similar date, two model boats and a model granary. Beni Hasan tombs 75 and 886 also contained coffins with outer decoration of type VI together with boat models. Coincidentally with the disappearance of models a new range of ritual objects begins to appear in burials: funerary figurines, developing into shabtis; magic wands and rods, accompanied by figurines which are 3-dimensional versions of the subjects 39
40
41
42
43
44
45
46
39 Bourriau, Pharaohs and Mortals, Cat.no. 136. Do. Arnold, "Kerarnikbearbeitung in Dahschur 1976-1981" in MDA1K 38 (1982), fig.8, nos.3, 5, 6, 7, 8, 10, 12; fig.9, no.15; fig.19, no.2. See note 10 above. The evidence is reviewed conveniently in Willems, op.cit., 62-104. 3 Some evidence has been lost through poor preservation. Mace's notebooks recording his work in Cemetery W at Diospolis Parva occasionally note the presence of painted coffins; see also the quotation from Hilda Petrie's diary in Bourriau, op.cit., 99. W.MJ\Petrie, G.Brunton and M.A.Murray, Lahun II (London, 1923), 34. ^ Angela Tooley, Middle Kingdom Burial Customs. A Study of Wooden Models and Related Materials. University of Liverpool, Ph.D.thesis, 1990. J.Garstang, The Burial Customs of Ancient Egypt (London, 1907), 214, 226, 243.
4 0
4 1
4 2 4
4 4 4
4 6
^«àexlands Insfctum •%5or het Nabije Oester»
12
Janine Bourriau
depicted on them; faience model offerings of food and pottery; female figurines. At the same time we begin to find a style of writing of funerary inscriptions which uses mutilated hieroglyphs. There is evidence that at least some of these changes spread out to the rest of Egypt from the cemeteries close to the Residence and the royal funerary complexes nearby. At the end of the XII Dynasty, the multiplicity of temples dedicated to the mortuary cults of Middle Kingdom Kings was sufficient to ensure the presence in this area of large staffs of priests, funerary scribes, and their dependent officials; here was the perfect setting for new rites of burial to appear and to spread. If we use the refinements in pottery dating now available from Lisht and apply them to a few key groups from the region, where the association of pottery, objects and (sometimes) coffin is reasonably secure, we may be able to establish when this change took place. I am aware of the great danger in extracting individual groups from a mass of archaeological material in this way, which is to generalise from a small and possibly unrepresentative sample of data. I have, however, chosen either groups which have already been used by others to establish the character of Middle Kingdom burials or well dated groups from the most recent work at Dahshur and Lisht by Dorothea and Dieter Arnold. (A list and brief discussion of the groups is given at the end of the paper). What they show is that a change in the selection of objects for burial coincides with the evolution of pottery types dateable to the late XH" to early XHI Dynasties. The starting point jeems to be the reign of Amenemhet in and the end, in this region, the advanced XÏII Dynasty. If we add three well dated groups from Abydos and Thebes the chronological picture is similar, but there is also some evidence to suggest that the changes occurred a little later at these centres. Most cemeteries of the mid-late XII Dynasty produced funerary statuettes in hard stones, representing the deceased as a wrapped mummy or wearing a long cloak. The name and title may be given, with or without the offering formula and other prayers, or the figure may be uninscribed. Such figures are sometimes placed in model coffins, and are found in offering chapels at Abydos as well as in tombs generally. It is consistent with what we know of Ancient Egyptians to assume a multiplicity of purposes for these figures. The statuette could function as an additional representation of the wrapped body of the deceased in funeral rites, and as a stand-in for it in the offering cult, or in the afterlife if the real body were destroyed. The statuettes seem to have evolved from the mummy figures placed in model funerary boats, illustrating belief in the posthumous pilgrimage to Abydos. The next step in the development of these statuettes is the application to them of the shabti text, and this seems to have happened first in the Memphis-Fayum area. The 47
48
49
50
51
4 7
Kemp and Merrillees op.cit., passim; B. Williams in Sarapis 3 (1975-6), 41-55; C. Liliquist in Sarapis 5 (1979), 27-8. See above, notes 18 and 40. ^ Abydos B13; the burial of Renseneb in Carter's tomb 25 in the Asasif at Thebes; and the objects from the Ramesseum tomb no.5. See tomb list. 50 W.M.F. Pétrie, G.A. Wainwright and E. Mackay, The Labyrinth, Gerzeh and Mazguneh (London, 1912), pl.xxx, two examples from Hawara dating from the reign of Amenemhet III; Bourriau op.cit., Cat.no.82, from Diospolis Parva, W38. H. Frankfort in JEA 14 (1928), 239-40, pl.XXII,3. 4 8
4
5 1
Patterns of change in burial customs
13
52
earliest examples are from Lisht. The text is an old spell, first known amongst Coffin Texts from Meir and El Bersheh. It is in the form of a legal contract, and sets out to provide the deceased with exemption from his duty to help maintain the irrigation system in the afterlife, by appointing the shabti as his substitute. This undoubtedly reflects the practice of a society in which the great division lay, as it still does in Egypt, between those who carry out manual work and those who do not. With the addition of this text the shabti figure has acquired a very specific function, but it did not thereby lose any of its earlier significance. Shabtis are still found, sometimes in miniature coffins, in offering deposits, or they may be represented in relief sculpture on stelae or in shrines. When it appears on the statuettes, the text uses mutilated hieroglyphs in which the legs of the signs representing birds, human beings, animals and reptiles have been removed. The reason for this we know, since the practice also occurs in earlier compilations, such as the Pyramid Texts: it is to prevent the power of the beings represented by the hieroglyphs from doing harm to the newly dead, as vulnerable (in Egyptian thinking) as the newly born to malign influence. The earliest burial group known to me to contain the script is the burial of the princess Neferu-ptah, daughter of Amenemhet III. From the late XII Dynasty onward, at Lisht, Dahshur and Hawara, this mutilated script was used for objects placed close to the body in burial. The tradition spread to Abydos and Thebes, the latest known example being the coffin of one of the most important kings of the Theban seventeenth Dynasty, NubkheperreInyotef. The wands and rods with their associated collections of magical figures, which appear in late Middle Kingdom burials anywhere in Egypt, can be paralleled by objects from domestic contexts at Kahun, a town dating from the reign of Senwosret II to the XIII Dynasty. Unfortunately the archaeological sequence at Kahun is unknown, so for most of the finds precise dates cannot be given. Nevertheless, the context is interesting in providing general support for the date of these objects and illustrating 53
54
55
56
57
58
59
60
61
62
5 2
H.D. Schneider, Shabtis I (Leiden, 1977), 182. For Lisht 453 in which the shabti of Ameny was found, see tomb list below. 53 Ibid., 46-7. To this discussion should be added Willems' comments on the dating of the relevant coffins, B2L, BIP, p.75-7 and M47C, p.97. 54 See shabtis of Bener and Wahneferhotep from Lisht, in tomb list. 55 P.Vernus,"Une Formule des shaouabtis sur un pseudo-naos de la XHIe Dynastie" in Revue d'Egyptologie 26 (1974), 101-114. 56 P.Lacau,"Suppressions et modifications de signes dans les textes funéraires" in ZÄS 51 (1914), 164. 57 N. Farag and Z. Iskander, The Discovery ofNeferwptah (Cairo, 1971), 48-58. 58 Abydos B13,dateable to Dynasty XIII, see tomb list. 59 Carter and Carnarvon tomb 25, which has a terminus post quern of the reign of Amenemhet IV; see tomb list. 60 in the British Museum, see H.E.Winlock,"The Tombs of the Kings of the Seventeenth Dynasty at Thebes" in JEA X (1924), pl.XIV. W.M.F.Petrie, Kahun, Gurob and Hawara (London, 1890), pLVHI. 62 The meagre information in the reports has been painstakingly examined in Kemp and Merrillees, op.cit., 57-102. 6 1
Janine Bourriau
14 63
how devices perhaps long used to protect the living, were being adopted into burial practices. The idea that the newly born and the newly dead required the same magical protection against the natural forces of disease and decay, the supernatural powers of the night and of the desert and the unquiet dead was hardly an innovation. The kinship between some spells among the Coffin Texts and the medico-magical papyri, such as those in the Ramesseum find, shows this very well. Nevertheless the notion seems to have found new expression at this time in the Memphis-Fayum region. The female figurines which also appear at this time cannot on present evidence be thought to originate in a particular region; sundry types were in circulation and represent local variants. The model food offerings and miniature cups made of faience, perhaps in origin a replacement for wooden tomb models, do seem to be more popular in the Memphis-Fayum region than in Upper Egypt but that may be a distortion due to an imbalance in the sources available to us. From the mid-XII Dynasty onwards, after the reign of Senwosret IU, changes in the decoration of coffins also occurred, coinciding with this change in the burial goods. Firstly the trend towards standardisation in exterior decoration becomes even more marked, as the local workshops of Bersheh, Beni Hasan, Asyut and Meir cease to operate. Willems shows clearly how his type VI coffin becomes the most common type throughout Egypt after the reign of Senwosret III. Secondly the frequency of interior decoration with Frise d'Objets and Coffin Texts shows a corresponding steep decline after the end of his reign. Isolated examples appear until well into the Xffl Dynasty but they are special cases: burials of the royal family and their household; ofSsnb-nf, a chief lector priest at Lisht; and of officials close to the King. Moreover, the coffins are not "standard class" coffins, but either "court" types or "deviants". The few examples imply that access to the texts was much more restricted in the late Middle Kingdom and this contrasts sharply with the situation in the early XII Dynasty, when coffins from Aswan, Abusir, Bersheh, Beni Hasan, Gebelein, Sedment, Saqqara, and Thebes, all carry Coffin Texts. 64
65
66
67
68
69
70
71
72
73
6 3
Altenmüller, see note 74 below, makes their primary function to protect living women and children clear. °4 Evidence to confirm this from other settlements in the region is lacking so far. The earliest levels above water at Kom Rabi'a at Memphis, currenüy being excavated by the Egypt Exploration Society, are late XII Dynasty (personal observation). It is doubtful that any site in the valley, such as Kom Rabi'a, can produce the range of objects in organic materials that have come from Kahun. Bourriau opxit., 110-127. 66 Stephen Quirke, personal communication. 6 CPinch, "Childbirth and Female Figurines at Deir el-Medina and el-Amarna" in Orientalia 52 (1983), 405-414. 6^ Willems lists three exterior decoration layout types, IVaa, IVba and VI, which occur during, and (in the case of IVba and VI) begin shortly before, the reign of Senwosret III and continue later. 20 coffins with these types of exterior decoration also carry the latest type of interior decoration, Willems' type 2, which includes Coffin Texts. Of these, none is certainly later than the reign of Senwosret III. At Dahshur, see coffins DalC-Da8x in Willems op.cit., 22-3. LiLi.Willems op.cit.,2A,\05; G. Lapp in SAK 13 (1986), 135-147. T6C and TlOC.Willems op.cit.,\\l\ Berlev in JEA 60 (1974), 106-113. 72 Willems op.cit., List 1, for definitions of these terms. Examples from these sites all carry Willems' type I layout of exterior decoration, which dies out after the reign of Amenemhet I; p.127 with chronological table p.121. 6 5
7
6 9
7 0
7 1
7 3
Patterns of change in burial customs
15
We can, I think, see a chronological correlation in the late XII Dynasty, between the decline in the occurrence of coffins with interior decoration and the appearance in burials of magical figurines, rods and wands. Is it possible that these objects are replacing some of the spells previously represented in the Coffin Texts? The basic study of the procession of animals, hieroglyphs and gods shown on the wands is by Altenmüller. The background myth is the battle of the sun god against his enemies. The sun god's helpers are shown carrying knives, and it is the protective powers represented by these beings which are called up to aid the dead. Represented on the wands in the Ramesseum group are deities: Seth, Re, Nekhbet, Thueris, and both male and female forms of Aha; mythical beasts: griffin with human head between its wings, snake-headed leopard, triple-headed serpent; real animals: frog, cat, baboon, turtle, hon; hieroglyphs: the wsr sign and the brazier sign for fire, ht. Spells referring to the myth of the sun god travelling through the caverns of the night are a major theme of the Coffin Texts. Willems has argued that the texts relate to episodes in the funerary rites themselves. If he is right, the relatively sudden disappearance of these texts from non-royal coffins would imply a dramatic change in private funerary ritual. Could the figures and the wands have come to replace, in a cheaper and more easily obtainable form, the solar spells in the Coffin Texts? The objects were, after all, in current use for a similar purpose in everyday life. This suggestion need not be invalidated by the fact that examples of Coffin Texts, and of Coffin Texts and Pyramid Texts together, continue to appear sporadically, painted or carved on the walls of the burial chambers of private tombs. I have so far discussed patterns of change in relation to the funerary geography of Egypt and the chronological phases of the Middle Kingdom. Finally I should like to suggest a change which I think can be seen in the social pattern of burial. This again relates to the last phase of the Middle Kingdom, the late XII - XIII Dynasties. Opportunities for burial and/or commemoration in a sanctified place close to a royal mortuary temple, or at Abydos at the "Terrace of the Great God", were open to more people than ever before. Statues and stelae were increasingly deposited in temples, both inside near the doors and in the outer courts, so that they might "dwell near the god" and benefit from the rites and offerings made there. Moreover, the texts of funerary stelae themselves change. The autobiographical texts of the early Middle Kingdom are replaced by religious ones, hymns to Osiris or Min (a change which may be related to a need to find another medium for texts relating to the Osiris myth previously provided by the Coffin Texts), and/or simple offering formulae with family genealogies and endless lists of names. A limestone stela from Abydos with a simple painted text is an example of the most modest category of memorial. Seventeen men and women are listed; only two of the men are given titles, and they are simply "steward". There is no reference to master or mistress, so we may assume their status was above that of the serving classes. Here we see, as Quirke has said, individuals from among the elusive middle ranks, between literate official and illiterate farmer. I shall both speculate and generalise to suggest that the late Middle Kingdom saw a simplification in burial customs, which made burials and/or commemorations 74
75
76
7 4
H.AItenmiiller, Die Apotropaia und die Götter Mittelägyptens, Diss. Munich, 1965. vols.I, U; id., in SAK 13 (1986), 1-27. 75 See tomb list below. 76 Quirke in Bourriau op.cit., Cat.no.41.
16
Janine Bourriau
less costly, and so a "goodly burial in the west", with its promise of eternal life, available to many more individuals. Fewer burials were furnished with both inner and outer coffins and fewer coffins had interior as well as exterior decoration; instead of specially made tomb models, what was placed in the tomb was amulets and magical figurines used in daily life, together with standard pottery vessels and food offerings. It is noticeable at Lisht North, for example, that there were many more burials of the late Xll-Xm Dynasties than of the early XII Dynasty and to judge by their inscriptions, these were not individuals of high rank; such people had moved away to set up their "dwelling" close to the King they served or, more often, his successor. The political process which marked the end of the Middle Kingdom was the reduction in the area controlled by the Pharaoh from the Residence at It-t3wy, culminating in the removal of the royal household to Thebes. This event must have been traumatic when it took place; nevertheless the evidence does not show, to date, that a correspondingly sudden or even very profound cultural change followed it. In Northern Egypt, the burial traditions of the late Middle Kingdom continue; the separate culture of the Eastern Delta flourishes until the period of the Hyksos wars. At the great mortuary temples of the XII Dynasty Kings at Lisht and Dahshur, one can observe, from the mid-XIII Dynasty onwards, an invasion of first, tomb shafts, then houses and grain silos into the sacred precincts. However, the contents of those houses and shafts show a steady evolution, not an abrupt change from the familiar types of the XII Dynasty itself. It is the advent of the XVIII Dynasty which introduces the next major cultural phase. TOMB LIST 7 7
Lisht South Wahneferhotep Group. Shabti and model coffin , both inscribed using the mutilated script, associated with pottery of the late XIII Dynasty. Closed group. Date: Advanced XJJJ Dynasty or later. 78
79
80
81
82
Bener Group. Shabti and model coffin , both inscribed using the mutilated script, associated with pottery of the late XH-early XIII Dynasty. Closed group. Date: Late XII - early XÏÏI Dynasty. 83
7 7
There are undoubtedly many other relevant groups from Lisht, but they must await full publication of the cemeteries around the North and South Pyramids. I have selected groups referred to in Kemp and Merrillees (see note 16 above), two recently published closed groups from Lisht South, and Senebtisi, because her burial is a constant source of reference. Schneider, opxit., 183. Type HIB, text version IIIA, fig.6. 79 P.Dorman in Arnold op.cit., 147-149. Arnold op.cit., 37-40. Schneider opxit., 183.Type HIA, text version ITIA, fig.6. Dorman in Arnold op.cit., 147-149. Arnold op.cit., 34-37. 7 8
8 0
8 1
8 2
8 3
Patterns of change in burial customs
17
Lisht North Tomb 315. Faience hippopotamus; lion from a magic rod; squatting man holding an offering table ; with pottery of Amenemhet III or a little later. The pottery included carinated cups, which do not occur in the early XII Dynasty corpus at Lisht South but do appear at Dahshur in the cemetery around the Pyramid of Amenemhet UP and in late XII Dynasty stratified levels at Memphis, Kom Rabi'a. No information on number of burials. Date: Late XII Dynasty 94
85
86
87
88
39
90
91
Tomb 453. An elaborate shaft tomb which contained remains of two burials, judging by the presence of two skulls and of 3 inlaid eyes from coffins. There were fragments of an anthropoid coffin; faience figurines of women; model food; magical figurines of a cat and a jerboa; a scarab and shabti of the same man, Ameny written using the mutilated script; and pottery, including hemispherical cups with an index of 150, fragments of Tell el-Yahudiyeh ware, and the neck of a wine j a r of Lisht South type shown on fig.70, no.67. The pottery is not consistent with a single period of use but nothing is earlier than the late XII Dynasty. Date: Late XU Dynasty onwards. 92
93
94
95
96
91
Senebtisi. This burial does not contain any of the grave goods whose date of introduction I am seeking to establish but it is a key burial for the period. Now dated by the pottery to the reign of Amenemhet III or a little later. The coffins support this date.lt is a point insufficiently stressed in the discussion of the date of the burial of Senebtisi that her coffins do not employ the mutilated scriptThis sets it apart from that of Neferu-ptah which it otherwise closely resembles. 98
99
E.Riefstahl, "An enigmatic faiencefigure"in Miscellanea Wilbouriana I (New York, 1972), 137143. 85 pace Riefstahl. 86 Information from tomb card, by courtesy of the Metropolitan Museum of Art. R.Engelbach, Harageh (London, 1923), pLXXXIV, 10B-10P. Arnold op.cit., 125-134. 89 Arnold, "Keramikbearbeitung", see note 39 above, fig.6,no.l2. 90 Personal observation. 91 Kemp and Merillees op.cit., 167. 92 See note 86 above. 93 G.T.Martin, Egyptian administrative and private name seals, principally of the Middle Kingdom and Second Intermediate Period (Oxford, 1971), 21, nos. 195,196. 9 Schneider op.cit, 182, type IIIA, text version IIC, fig.6. 95 Shown clearly in a photograph. 96 Arnold op.cit. At Lisht South this type comes in cluster 3, on fig.76, from a deposit dated to Senwosret Ill-early Amenemhet HI. B. Williams in Sarapis 3 (1975-1976), 41-55; C. Liliquist in Sarapis 5 (1979), 27-8; A. Mace and H.E. Winlock, The Tomb of Senebtisi at Lisht (New York, 1916), passim. 98 Arnold op.cit., 37, n.114. Her observation is confirmed by comparison made by the author, of a new drawing of the Senebtisi marl C jar, now in the Oriental Institute in Chicago, with the Dahshur example of the late XÜ" Dynasty cited by Dorothea Arnold. 99 Willems op.cit., 104-5. 8 7
8 8
4
9 7
Janine Bourriau
18 Closed Deposit. Date: Late Amenemhet DI. 100
Harageh Cemetery A No.77. Contained 3 burials in 2 chambers. Listed in Kemp's seriation in a cluster around tomb 91, which has a terminus post quern of Senwosret III. It contained a faience hippopotamus and a faience vase, pLXIV, 10,12. Date: Late XTJ Dynasty or later. m
102
Cemetery A No.55. 2 burials in 1 chamber. Pottery includes a wine jar of type 4 If with an aperture index which places it in Arnold's cluster 3 , belonging to the period between the reign of Senwosret II and the end of the XII Dynasty .The tomb also contained a faience dwarf, pl.XIV,9 and "ivory hands", not illustrated. These are presumably ivory clappers such as those found in the Ramesseum tomb, see below. Date: Dynasty XII, Senwosret II or later. 1 0 3
104
Cemetery A NoS6. Contained 10 burials in 3 chambers. In addition to pottery, the group included a faience dog figurine, pl.XIV,8; a "rough" hippopotamus and fragments of a headrest. If we can consider a deposit with so many burials as a single g r o u p , the pottery includes a hemispherical cup of approx.index 140, dated by Arnold to the advanced XIII Dynasty, and a carinated cup, a type which does not 107 occur before the reign of Amenemhet in. Date: Advanced XIII Dynasty. 105
106
108
Cemetery A No. 73. One burial in a single chamber. In addition to pottery it contained a faience dwarf,cf.pl.XIV, 11.The pottery included a carinated c u p and a hemispherical cup with index of 130, which belongs to the advanced XLU Dynasty. Closed Deposit. Date: Advanced XÏÏI Dynasty. 109
110
1 U U
I am selecting those groups listed in Kemp and Merrillees, 175. It is not possible to discuss Kemp's seriation here, but in my view problems lie in his reliance on the accuracy of the typing of pottery by Engelbach and in the assumptions made in reducing the original corpus to a size suitable for the seriation technique. Engelbach, op.cit., pLLVIII. ™ Ibid. Arnold op.cit., fig.76, p.143. Engelbach op.cit., pLLVIII. 105 Kemp in Kemp and Merrillees, opxit., 31, seriated the pottery in tombs with single chambers not known to contain more than one body, and found that the results compared well enough with the seriation from multi-burial groups to suggest there was no significant distortion. This surprising observation needs further testing on other cemeteries. Arnold op.cit., 141. It is dangerous, of course, to use a type drawing rather than a drawing of the actual vessel to measure the vessel index, because the validity of any conclusions depends on the accuracy of the typing which cannot, in this case, be checked since the whereabouts of this cup are unknown. It is also dangerous to place too much weight on the evidence of a single cup, Arnold opxit., 141. All that can be done is to suggest probabilities in the light of the evidence which remains. See notes 89 and 90 above. Engelbach opxit., pl.LVIII. See notes 89,90 above. Arnold opxit., 141. 1 0 1
2
1 0 3
1 0 4
1 0 7
1 0 8
1 0 9
1 1 0
Patterns of change in burial customs
19
ul
Cemetery A No.ll2. A single burial in one chamber. Group of faience figurines on pl.XIV.l. Among the pottery is a wine jar, 41m, which belongs to Arnold's cluster 3, of the period from the reign of Senwosret II to the end of the XII Dynasty. Closed Deposit. Date: Late XU Dynasty. 112
11
Cemetery B No.353. * No information on number of burials. In addition to pottery,the deposit contained faience figurines of a cow and a frog, pl.XIV,6,7. The pottery included a hemispherical cup with an index of 163 and Arnold has shown that at Lisht South only the pottery deposit in the South East dump, dated to Senwosret HI to early Amenemhet m, contained cups with indices below 1 7 0 . Date: Late XII Dynasty. 114
115
Cemetery S No.644. Shaft with two chambers; no bodies recorded. A full list of contents is provided by Kemp and Merrillees. Assuming this is a homogeneous deposit, they argue in favour of placing it at the late end of the Harageh sequence. The pottery includes two hemispherical cups with indices of 163 and 130 respectively, suggesting an advanced XIII Dynasty date, with which the Tell el-Yahudiyeh ware and kohl pots agree. Date: Advanced XIU Dynasty. 116
1 1 7
Hawara Tomb 58. Intact burial in a pit containing the coffin of a woman, Sitrenenutet. There was a female figurine in wood with a model bed; a scarab; a model storage jar of wood; and a faience model of a pigeon. The pottery consisted of a water jar and two hemispherical cups with indices of 165 and 175, dateable to the end of the reign of Amenemhet III. 118
119
Closed Deposit. Date : End of the reign of Amenemhet HI. Thebes Tomb 25 in the Asasif.Burial of Renseneb,"Great one of the Southern Tens". From a shaft with two chambers containing at least 5 burials. The coffin had been dragged out of one of the chambers, but the burial was intact. The coffin's inscriptions are in mutilated hieroglyphs, but only fragments of the text giving the man's name and title were recorded before the coffin disintegrated. Within the body wrappings, against the back, was a faience hippopotamus. Underneath the coffin were fragments of a toilet 120
1 1
1 Engelbach opxit., pl.LIX. H2 Arnold op.cit., 143. 11^ Engelbach op.cit., pl.LXI; Bourriau op.cit., Cat.no.88b. ! * Arnold op.cit., fig.75. Engelbach op.cit., pllXII; Kemp and Merrillees op.cit., 23,34, 36, 39,163,223, fig. 15. 116 Arnold op.cit., 141. 11 The cemetery began in conjunction with the building of the Pyramid of Amenemhet in. 118 W.M.F. Pétrie, G.A. Wainwright and E. Mackay, The Labyrinth, Gerzeh and Mazguneh (London, 1912), 36, pLXXX; Bourriau op.cit., Cat.no.88a. 119 Arnold op.cit., 141. 120 H. Carter and Carnarvon, Five Years Explorations at Thebes, 54 ff. 4
1 1 5
7
Janine Bourriau
20
box with the name of Amenemhet IV, and in the south chamber were a magic wand and an ivory crocodile. The pottery, shown in the publication on pl.LII, appears entirely consistent with the date of the box. The hemispherical cup has an index of approx.162. The pottery group as a whole shows how the classic XII Dynasty style which originated at Lisht had by now reached Thebes. Except for one Upper Egyptian type, pl.Ln,2 J5, all the pottery can be paralleled in the Harageh/Lisht/Dahshur corpus. Date: Late XH-early XIII Dynasty, after reign of Amenemhet IV. 121
Ramesseum group. Shaft with three chambers. At the bottom of the shaft a group of objects was found scattered around a box of papyri, within an area of 2 square feet. The papyri had suffered only natural decay and had not been disturbed beyond their original displacement into the shaft, and this suggests that the objects had undergone a similar fate. There are strong links between the objects in style, function and date, all of which suggests that we are dealing with a homogeneous group. This consists of 3 female figurines of faience and limestone; a paddle doll of painted wood; a wooden statuette of a woman wearing a Bes-Aha mask and carrying snake wands; model cups of faience; a model cucumber, a fragment of a large magic rod in ivory; fragments of 3 ivory wands; a pair of clappers; an ivory object of unknown use; faience figurines of two baboons and a standing lion; an ivory djed pillar, and a cobra wand. The papyri have been dated to the mid-Xni Dynasty and the objects are entirely consistent with that date. It is worth stressing that with the exception of the cobra wand and the statuette of the woman carrying wands.the objects are typical of a late Middle Kingdom burial. They have been considered to be an exceptional group, composed of instruments of magic used with the magical texts among the papyri. Their function within the rites of protective magic, with the exception of the model food and vases, is unquestioned but they are unremarkable within the general context of burial groups of the late XII to Xni Dynasties. Date: Mid-XUJ Dynasty. 122
123
124
125
Abydos Cemetery B No.13. The deposit contained the two shabtis of Renseneb with a text inscribed in the mutilated script, and pottery of advanced XHI Dynasty date. Date: Advanced XUJ Dynasty. 126
1 2 1
127
iE. Quibell, The Ramesseum (London, 1898) 3, pl.III; Kemp and Merrillees, op.cit., 166. 1 The detailed list given in Kemp and Merrillees, op.cit., 166, is not repeated here. * ^ This class offigurineis to be dated to the late Middle Kingdom according to information provided by Dorothea Arnold. Wrongly identified by Kemp and Merillees as a handle. Bourriau op.cit., 110-111, Cat.nos.62, 100. T.Peet, Cemeteries of Abydos II (London, 1914), 57-8, 113, pl.XIII,3; Bourriau op.cit., Cat.no.83, 135a. * Schneider op.cit., 183-4, type IIIC, text version IIIA. 2 2
2
1 2 4
1 2 5
1 2 6
2 7
Circumstantially
adverbial?
the circumstantial sdm(.f)lsdm.n(.f)
reconsidered*
by M a r k Collier
0. Introduction The work of Professor H.J. Polotsky has had a profound impact on the study of Middle Egyptian grammar, particularly the study of the syntax of the verb . It is now over a quarter of a century since Polotsky published his discovery of the circumstantial sdm(.f)/sdm.n(.f) in Middle Egyptian . Based on their ability to occur in environments where simple adverbial expressions such as prepositional phrases are to be found, Polotsky developed a syntactic analysis of the circumstantial sdm(.f)/sdjn.n(.f) as adverbial forms of the verb - an analysis which has become widely accepted. However, in recent work I have put forward the case that, at least in certain environments where the circumstantial sdjn(.f)lsdjn.n(.f) occur, such an analysis leads to serious difficulties which do not arise if the circumstantial sdm(.f)lsdtn.n(.f) are analysed simply as verbal verb-forms . In this paper, it is argued that the verbal analysis of the circumstantial s(hn(.f)lsdjn.n(.f) can account satisfactorily for the overall syntactic distribution of these forms in Middle Egyptian in the following seven 1
2
3
4
5
* I am grateful to Janine Bourriau and Stephen Quirke for inviting me to contribute to this volume, even though I was not able to attend the 1988 Colloquium, and to Harry Smith for comments on a draft of this paper. The research was supported by Research Fellowships from The British Academy and Corpus Christi College, Cambridge. For reasons of space, philological and grammatical notes and references have been kept to a minimum. For Middle Egyptian, see particularly: Etudes de syntaxe copte (Cairo, 1944), part III; 'Egyptian tenses', conveniently in Collected papers (Jerusalem, 1971), 71-96; 'Les transpositions du verbe en égyptien classique', Israel Oriental Studies 6 (1976), 1-50 (Transpositions'). Polotsky, 'Egyptian tenses', originally published in The Israel Academy of Sciences and Humanities, Vol. II.5 (1965), 1-25. As Polotsky notes (§41), Erman had earlier proposed a circumstantial form of the sdm(.f) restricted to the iwf sdm.f pattern (Äg. Gr? §343). 3 The fully developed position is to be found in Transpositions', section 3. In this paper Polotsky also proposed a circumstantial form of the prospective sdßi(.f) (3.6). However, this form does not share the particular distribution of the circumstantial sdjn(.f)/sd/n.n(.f) and is not discussed here. M.A. Collier, 'The circumstantial sdm(.f)/sdjn.n(.f) as verbal verb-forms in Middle Egyptian', JEA 76 (1990) 73-85; id. A grammatical analysis of sentences with iw in Middle Egyptian (University of London PhD thesis, 1989), revised version forthcoming as Verbal syntax in Middle Egyptian . This analysis is intended to cover the occurrence of the circumstantial sdjn(f)/sdm.n(.f) in all Middle Egyptian source genres. However, the exemplification focuses primarily on literary Middle Egyptian. 1
2
4
5
21
Mark Collier 6
affirmative environments in which they have been securely identified , and thus compares favourably with Polotsky's adverbial analysis of the circumstantial sdjn(.f)lsdjn.n(-fP' 1. After initial particles 2. After converters 3. Virtual' relative clauses 4. After auxiliaries 5. Adjunct/focus position 6. Coordination 7. Control The central claim of the verbal analysis is that the bare circumstantial sdjn(.f)lsdjn.n(.f) (i.e. the circumstantial sdm(.f)/sdjn.n(.f) without introductory expression), and indeed the bare sentence with adverbial predicate and the bare pseudo-verbal construction, are unconverted, i.e. non-transposed, patterns. This can best be seen in the sentence with adverbial predicate . The clause forms of this pattern can be classed : 8
9
10
" Other possible occurrences after ir (cf most recently E. Doret, The narrative verbal system of Old and Middle Egyptian (Geneva, 1986), 22 with fn 77) and after the negation n (cf F. Kammerzell GM 102 (1988), 46-50) are more contentious and are not discussed here. 7 This paper presents an overview of the verbal analysis of the circumstantial sdm(.f)/sdjn.n(.fl. For many of the sections below, details of particular analyses (where these differ significantly from the Polotskyan tradition), full exemplification and comparisons with the Polotskyan tradition are to be found in papers noted in the references below. To supply full supporting references, it has been necessary to refer to a number of papers either in press or in MS form at the time of writing. 8 The term 'bare' is intended to indicate the basic syntactic formation considered apart from any syntactically active introductory expression. Although it does not affect the syntactic status of the pattern (cf the brief discussion of the relationship between syntax and morphology in section 4), the morphological form of the pronominal subject with the sentence with adverbial predicate and pseudoverbal construction depends on the local morphological environment in which it occurs. After auxiliaries the pronoun suffixes to the auxiliary (EG §§37, 117.2, 323), after most initial particles (excepting smwn, cf EG §241) the pronoun cliticizes to the initial particle (EG §§44.2, 119, 324), whereas converters allow a complex intermixing of the two (EG §§44.2, 200, 223). In 'virtual' relative clause, adjunct, coordination and en vedette usage, the auxiliary ÎW is required to 'support' the pronominal subject (EG §§117, 323). However, Iw clearly does not behave as a syntactic converter or initializer in this usage but as morphological support. The terminology adopted here is extended in its syntactic sense to cover pronominal examples including examples with w-support, The pseudo-verbal construction is standardly treated as exhibiting the syntactic form of the sentence with adverbial predicate. However, in the present analysis, this holds only for the pseudo-verbal construction with preposition + infinitive predicate. The Stative is analysed as a verbal verb-form in much the same manner as discussed below for the circumstantial sd/n(.f)/sdjn.n(.f) (cf particularly the discussion of the inherent non-nominality of unconverted verbal expressions below). However, the Stative differs from the simple suffix conjugation forms in exhibiting SVO order (as well as mandatory agreement and in being able to occur without a subject as a predicative phrasal expression). The pseudoverbal construction with stative predicate is hence analysed as a verbal sentence exhibiting SVO order. Since this issue lies beyond the bounds of the present paper, attention is focused on the simple sentence with adverbial predicate. 10 The classification is dependent, of course, on analyses which differ somewhat from the traditional Polotskyan analysis. See the relevant sections below. 3
3
3
3
3
9
Circumstantially adverbial?
(initial) main clause 'adverbial' clause true 'adverbial' clause nominal clause adjectival clause
converter/initializer initial particle/auxiliary e.g. mkllw
unconverted clause sr m pr srm pr
prepositional converter e.g. hr-ntt nominal converter e.g. ntt/wnt/wn(n) adjectival converter e.g. nty/wn(n)
sr m pr srmpr 11
sr ïm.f
In this analysis the form traditionally analysed as the 'adverbial' clause form is analysed as the basic unconverted form of* this construction which co-occurs with initial particles, auxiliaries and converters. It would seem that the simple suffix conjugation presents an analogous paradigm. If Polotsky is correct about the role of nominal and adjectival transpositions , then the forms of the sdm(.f)Isd.m.n(.f) can be organized similarly: 12
(initial) main clause 'adverbial' clause true 'adverbial' clause nominal clause adjectival clause
converter/initializer initial particle!auxiliary e.g. mkllw prepositional converter e.g. hr-ntt
unconverted clause circumstantial $dm( f)lsdjn.n(.f) circumstantial sdm(.f)/sdm.n(,f) circumstantial s dm ( f) Is dm .n(.f)
that-form sdm( .f)lsdm.n( f) relative form
sdm(f)lsdpi.n(.f)
The circumstantial sgjn(.f)lsdjn.n( f), the forms traditionally analysed as the 'adverbial' forms of the simple suffix conjugation, are analysed as the basic unconverted forms of the simple suffix conjugation which co-occur with initial particles, auxiliaries and converters. However, in nominal and adjectival environments, the suffix conjugation verb itself undergoes conversion and exhibits a specialized form. The verbal analysis of the circumstantial sdjn(f)/sdm.n(.f) diverges from the Polotskyan tradition in that an unconverted/non-transposed verb-form clearly cannot be analysed as an adverbial
1 1
Relative clauses, of course, exhibit a relationship between the antecedent and an overt or non-overt resumptive expression within the relative clause. Cf section 3 below. This framework is not dependent on the postulation of nominal and adjectival transpositions of the verb. Indeed it opens up the possibility of alternative analyses of these forms (and hence of 'second tenses'). However, this paper is concerned with the circumstantial sdm(.f)lsd,m.n(.f) and Polotsky's analysis of the nominal and adjectival transpositions of the verb is adopted here to facilitate the presentation of this overview. 1 2
Mark Collier
24
conversion or transposition of the verb, but must be analysed simply as the verbal verbform of the simple suffix conjugation . In the present framework, the basic parts of speech are analysed as feature bundles . In particular, parts of speech are classed according to their nominal and verbal properties. It is commonly recognized that in Middle Egyptian nouns and adjectives contrast with (unconverted) verbs and prepositions in having nominal properties . At the same time verbs and adjectives contrast with (true) nouns and prepositions in having verbal properties . The Middle Egyptian parts of speech may thus be represented: 13
14
15
16
3) noun adjective verb 'adverb'
[NJ [+N] [+N] [-N] [-N]
[V] [-V] [+V] [+V] [-V]
Equally, it is commonly recognized that the behaviour of an entire expression is determined by the most important element within that expression (the 'head'). Within clausal expressions, this element is the head of the predicate (the 'predicator'). Clearly, in the verbal analysis, the circumstantial sdjn(.f)ls&m.n(.f) are verbal expressions (being headed by a verb), whereas the bare sentence with adverbial predicate is clearly adverbial (being headed by an adverbial predicate) . The feature analysis of the basic major parts of speech allows an account, where required, of the so-called 'adverbial' properties of the circumstantial sdm(.f)/sdrn.n(.f), without invoking adverbial substitution or adverbial transpositions, by reference to the sub-feature [-N] shared by both unconverted verbal expressions and prepositional phrases. In this approach, 'adverbiality' is not treated as a super-category encompassing non-nominal verbal expressions (which are thus required to exhibit an adverbial form) and prepositional phrases, but as a sub-category inherently shared by unconverted verbs and prepositions. In this manner, verbal expressions do not have to be converted or transposed into adverbial forms in order to exhibit a substitutional paradigm with prepositional phrases in certain syntactic environments. 17
13 The present paper is concerned solely with the synchronic analysis of the syntax of the circumstantial sdjn(.f)/sdm.n(f) in Middle Egyptian. However, the verbal analysis is not incompatible with the possibility of a nominal origin for the Middle Egyptian suffix conjugation verb-forms (here the circumstantial sdm(.f)lsdfn.n(.f)). It is not precluded that there was a reanalysis of an originally nominal form at some point before Middle Egyptian, as postulated, for example, for prepositions (which may well have had a nominal origin but do not behave as nouns in Middle Egyptian). 1 The feature analysis of the basic parts of speech originated with N. Chomsky, 'Remarks on nominalization', conveniently in id. Studies on semantics in generative grammar (The Hague, 1972), 11-61, and has become standard in generative grammar. Cf EG §48. Cf £G §143. The circumstantial sdjn(-f)lsdjn-n(f), the bare sentence with adverbial predicate and the bare pseudoverbal construction may be said to have the form of a 'small clause'. The small clause analysis offers interesting possibilities for an analysis of a wide range of grammatical properties ranging from tense/aspect effects to control. 4
1 5
3
1 6
3
1 7
Circumstantially adverbial?
Although the circumstantial sd,m(.f)/sdjn.n(.f) can thus be used in subordinate functions in the verbal analysis (where compatible with their nature as unconverted verbal verb-forms, either in terms of their status as unconverted forms (see section 3) or in terms of their inherent non-nominality (see section 7)), as unconverted verbal verb-forms they are not forms specialized solely to exhibit subordinate behaviour as substitutes for simple adverbial expressions such as prepositional phrases. Most importantly, the present framework also licenses the circumstantial sdm(.f)/sdjn.n( f), along with the bare sentence with adverbial predicate and the bare pseudo-verbal construction, to behave as main clause patterns in certain environments, e.g. after initial particles, auxiliaries and converters . The presence of initial particles, auxiliaries and converters allow the entire pattern initial particle + unconverted clause, auxiliary + unconverted clause, and converter + unconverted clause to exhibit a distribution licensed by the initial particle, auxiliary or converter. Initializers such as initial particles and auxiliaries allow a pattern to occur initially (the former also co-occur with initial main clauses, see section 1). This brings us to the third property exhibited by the circumstantial SQjn(.f)/sd_m.n(.f), the bare sentence with adverbial predicate and bare pseudo-verbal construction. These patterns are essentially non-initial. Non-initiality is not, in the present framework, a matter of (nor related to) subordination or predication, but of relative tense/aspect : ordinarily, these forms require some extra expression (or context) on which to 'lean' for relative tense support . Since non-initiality is treated as a property of tense/aspect rather than subordination, these patterns are not required to be syntactically subordinate to the initial particle or auxiliary, conforming with the main clause usage noted above. Rather, they can be analysed as unconverted expressions functioning as non-initial main clauses . In contrast, converters allow an unconverted construction to behave subordinately exhibiting a form specialized by the presence of the convener to occur in the relevant subordinate environment. Traditionally, such subordinate behaviour has been analysed in terms of the traditional non-verbal parts of speech as nominal, adjectival and adverbial subordination, so that subordinate clauses can function as nominal clauses, adjectival clauses and adverbial clauses respectively . As noted above, the suffix conjugation differs from the sentence with adverbial predicate and pseudo-verbal construction in that, if Polotsky is correct, the verb itself can undergo subordinating conversion and displays a form specialized for nominal and adjectival subordinate functions. 18
19
20
21
22
*8 The internal grammatical relation between initial particle, auxiliary, or converter and the unconverted clauses with which they co-occur will be the subject of the relevant section below. !9 Cf J.P. Allen, in Essays on Egyptian grammar, ed. W.K. Simpson, (New Haven, 1986), 1-21. The issue is but touched on here. The term 'relative tense/aspect' is deliberately intended to be vague in order not to prejudice compatibility with a number of interesting proposals now under development. 20 Treating non-initiality as a tense/aspect property rather than as a property of syntactic subordination has the important advantage of licensing the circumstantial sdjn(.f)lsdm.n(.f), the bare sentence with adverbial predicate and bare pseudo-verbal construction to occur syntactically initial in suitable contexts, as noted for noun + Stative by F. Junge, in Crossroad, eds. G. Englund and P.J. Frandsen (Copenhagen, 1986), 208-9 with fn 84. This account draws on the similar usage by J.H. Johnson, Serapis 6 (1980), 69-73. However, Johnson attempts to integrate non-initiality with issues of subordination and predication within the adverbial analysis of the circumstantial sdm(.f)lsdm.n(.f). Cf HJ. Polotsky, in Lingua sapientissima, ed. J.D. Ray (Cambridge, 1987), 18-9. 2 1
2 2
Mark Collier
26
This approach to the syntax of the suffix conjugation, the sentence with adverbial predicate and the pseudo-verbal construction has much in common with that of Polotsky and is clearly in his debt on many issues . However, there is a substantive difference in the treatment of the circumstantial sdjn(.f)lsdm.n(.f). In Polotsky's system, these verb-forms are analysed as adverbial transpositions of the verb on die basis of their known ability to substitute with simple adverbial expressions such as prepositional phrases and hence their distribution is to be accounted for in these terms. However, in the verbal analysis of the circumstantial sd/n(.f)/sdjn.n(.f) presented above, the behaviour of these forms can be described according to several more finelygrained grammatical oppositions which cannot easily be articulated within the Polotskyan tradition. Thus the circumstantial sd/n( f)/sdjn.n( f) are unconverted forms and are capable of aligning in a paradigm with other unconverted expressions (particularly unconverted clauses) to contrast with converted expressions. They are also non-nominal expressions and can align in a paradigm with other non-nominal expressions to contrast with nominal expressions. Utilizing these resources, the verbal analysis of the circumstantial sdjn(.f)/sdjn.n(.f) can account satisfactorily for the distribution of these forms, not only in the core environments for Polotsky's adverbial analysis (i.e. where the circumstantial sdm(.f)/sdjn.n(.f) occur in a paradigmatic environment with prepositional phrases), but also in environments which do not easily reduce to an analysis based on adverbial substitution . 23
24
1. After initial p a r t i c l e s
25
With the exception of an important paper by Janet Johnson, constructions introduced by initial particles have received little detailed attention within the Polotskyan tradition . The exemplification provided here focuses on mk, the most common initial particle, although the discussion below can be applied to a range of other initial particles . The particle mk is exhibited before the following patterns: 26
27
2 3
This work is also indebted to recent developments in generative grammar. However, the present paper is concerned with Egyptological issues and linguistic references have been avoided where not integral to the argument. The term 'unconverted clause', coined to relate to the transpositional terminology introduced by Polotsky, is essentially the equivalent of the Linguistic category S. This paper is concerned solely with the syntactic form and function of the circumstantial sdm(.f)lsdm.n(.f) and does not deal with their morphology, the essentials of which were established in 'Egyptian tenses', §§11, 41, nor with predication, for which see Collier, JEA 76 (1990), section 4. Tense/aspect (and thus non-initiality) and topic-focus/theme-rheme organization are not discussed in order not to prejudice compatibility with other proposals currently under development (cf fn 19). 25 For details, full exemplification and comparison with the adverbial analysis, see Collier, JEA 76 (1990), 73-85. In the text examples are drawn where possible from the simple affirmative examples of each pattern; negation and anticipatory emphasis are not normally recorded. 26 J.H. Johnson, in Festschrift Westendorf. Studien zu Sprache und Religions Ägyptens, I, ed. F. Junge (Göttingen, 1984), 71-85. Cf Polotsky, 'Egyptian tenses', §§41-5. Other particles which exhibit this behaviour are: ïs£, ti (see Collier, op. cit. 74-7); h3 (see id. 'Constructions with h3 revisited', GM 120 (1991) forthcoming); smwn,hr, nhmn (see id. 'More on initial particles' MS). The claim of the present section is that these initial particles form a natural class and co-occur with the same type of grammatical pattern (the unconverted clause, see below). However, this does not preclude individual particles from exhibiting specific collocational restrictions which may reduce or increase the number of patterns (drawn from the class of unconverted clauses) with which they may co-occur in comparison to those exhibited after mk. 2 4
2 7
Circumstantiaiiy adverbial?
27 2
a) sentence with nominal predicate * 4) Sh.S 159-60: mkh.rt.ipw Im.k See, it is my due from you. b) sentence with adjectival predicate 5) Leb 86-7: mk b h rn.i See, my name is detested. c) initial prospective sdjn( f) 6) Hek 2,29: m&i ir.iSmw 3 See, I shall spend the summer here. d) second tenses 7) Peas. B2,123-4: mk trr.k r irt hn .i See, it is with me that you will arrange to act. e) participial statement/ïn+noun+sajn(.f) 8) West 6,6-7: mt Ink db3[.i\ sw See, I am the one who will replace it. f) existential sentence 9) Peas. B 1,203-4: mknnkmn dd n.k st See, there is no profit for the one who says it to you. g) sentence with adverbial predicate 10) Sin B263: mk wi m-b3h.k See, I am before you. h) pseudo-verbal construction 11) Peas. Bl,70: mk wl 3tp.kw(i) See, I am burdened. i) circumstantial sdjn(.f)/sdm.n( f) 12) Sin B181: mk in.t(w) n.k wdpn n nsw r rdh rh.k ntt.... See, this decree of the king is brought to you to cause that you know that 13) Sh.S 2-5: mk ph.n.n hnw ssp hrpw hwi mnit h3tt rdï.t(i) hr t3 See, we have reached home, the mallet has been taken, the mooring-post struck, and the prow rope placed on land. The initial main clause patterns a)-e) are patently unconverted forms: initial main clauses do not behave as specialized substitutes for non-verbal parts of speech. Equally, in the present framework, the circumstantial sdm(.f)/sdjn.n(.f), the bare sentence with adverbial predicate and the bare pseudo-verbal construction are also the unconverted/non-transposed forms of their respective constructions. Hence the patterns c
c
c
29
30
2
° Since it is the behaviour of the entire pattern which is of concern here, this pattern is not divided according to the well known sub-categories which have been investigated in recent years. Cf most recently E. Doret Rd'E 40 (1989), 49-63. 9 For the presence of the circumstantial sdm(f)/sd/n.n(.f) in these examples, and for corroborative evidence from negation, passives and the behaviour of verbs of motion, see Collier, JE A 76 (1990), 75-77 (2h). The syntax of the existential sentence (0 is in need of detailed discussion and is not considered here. However, the wn form found after initial particles (EG §107.2) is probably to be analysed as a noninitial main clause form. 2
3 0
3
28
Mark Collier
in paradigm a)-i) form a natural class as unconverted clauses. In this analysis, then, initial particles simply co-occur with unconverted clauses : 31
PRT
mk
unconverted clause a) initial main clause sentence with nominal predicate sentence with adjectival predicate prospective sdm(.f) second tenses participial statement b) non-initial main clause existential sentence bare sentence with adverbial predicate bare pseudo-verbal construction circumstantial sdm(.f)/sdm.n(.f)
Most importantly, none of the members of paradigm 14) behave subordinate^ in relation to mk . The initial main clause patterns are clearly not subordinate, since they can stand independently as main clauses without supporting expressions and do not enter into substitutional relations with simple parts of speech. Equally, in the present analysis, the circumstantial sdm(.f)/sdm-n(-f)> the bare sentence with adverbial predicate and the bare pseudo-verbal construction do not behave subordinate^ after mk, their role is not to behave as the substitute for a simple part of speech. Rather, just like the initial main clause patterns a)-e), they function as main clause patterns. However, in contrast to initial main clause patterns, they are generally restricted to non-initial positions because of their contextually dependent relative tense/aspect properties (see section 0 above). In the present case, particles such as mk provide temporal support for these non-initial formations (for mk, by making the pattern it precedes relevant to, i.e. worth the attention of, the addressee and hence to the temporal reference point of the addressee). Separating out form and function, by form mk simply co-occurs with the basic clausal pattern for each construction it precedes; that is to say trie unconverted form. By function, these patterns behave as main clause formations; they do not behave 32
1
3 In contrast, mk does not co-occur with converted clauses, whether nominally or adverbially converted, cf Collier, op. cit 77-78. Clearly, this paradigm as it stands presents a fundamental problem for the adverbial analysis of the circumstantial sdjn{.f)Jsd/n.n(f). In this environment the circumstantial sdm(.f)lsdm.n(.f), along with the bare sentence with adverbial predicate and the bare pseudo-verbal construction do not occur in a substitutional relationship with cither simple prepositional phrases or with indisputably adverbial clauses. Rather, they occur in a substitutional relationship with initial main clause patterns, which are clearly unconverted clauses. Of course, there are various options available, although these involve some emendation to paradigm 14) as presented here. The most plausible is that advanced by Johnson, op. cit. 80 and 83 who proposes a deleted auxiliary before the circumstantial sdm(.f)/sdjn.n(.f), along with the bare sentence with adverbial prediate and the bare pseudo-verbal construction, removing the adverbial analysis of these forms to the internal syntax of construction with auxiliaries, whilst supplying an overall main clause construction to enter into substitutional relations with the other initial main clause patterns exhibited in the paradigm. Hence mk behaves syntactically as a free 'présentatif. Cf Polotsky, *Egyptian tenses', §45 and Johnson, op. cit. 83. 3 2
29
Circumstantially adverbial?
subordinated to the initial particle (or some other expression) and they do not behave as the substitute for a non-verbal part of speech, either nominal, adjectival or adverbial. 2. After converters
33
34
As has been recently re-emphasized by Janet Johnson , a similar paradigm to that found after initial particles is exhibited after preposition-rctt converters in Middle Egyptian . Preposition-rttf converters occur before the following patterns : a) sentence with nominal predicate 15) Peas. Bl,61-2: ph.k m 3pd dd3 hr-ntt ntk h n nmh You shall attain fatted fowl because you are a father for the orphan b) sentence with adjectival predicate 16) Kahun 3,33: hr.k dd.k n.fiw.0 ksn r.l hr kd hr-ntt dns tw r.l Thus you will say to him, Tt is altogether too irksome to me because you heavier than me'. c) initial prospective sajn(.f) 17) Berlin 10012, 18-19: dd(.I) rh.k r-ntt hprprtSpdtm 3bd4prtsw 16 I am speaking that you might know that the rising of Sothis will occur in the fourth month of prt day 16 . d) second tenses 18) Berlin 10023 A 1-2: dd(.i) di(.ï) rh.k r-ntt spr.nn.ï iry- 3 n hwt-ntr Snt s3 Vmny r-dd ïw3.kwï hr s3.ï I speak that I might cause that you know that the doorkeeper of the temple Senet's son Ameny has arrived saying 7 am to be taken off on account of my son'. e) participial statement 19) Kahun 2.2 29/39: .... hr-ntt ntk irr nfrt nbt .... because it is you who does everything good. f) existential sentence 20) Mill 1,5-6: sdr.k s3w.n.k lb.k gs.k hr-ntt nn wnn mr(t) n s hrw n ksnt * You should sleep having guarded your heart yourself because there is no servant for a man on the day of misfortune. 35
36
37
c
3
33 Details and comparison with the adverbial analysis in M.A. Collier, 'Preposition-n« clauses' MS. 34 Johnson, op. cit. 81-3. The discussion in this section can be extended to cover the occurrence of the bare sentence with adverbial predicate and the bare pseudo-verbal construction after ntt/wnt/wn(n) converters in nominal clauses (EG §§186-7) and nry/wn(n) converters in adjectival clauses (EG §§200, 396). The that-form and relative form sdjn(.f)lsdjn-Mf) are direct (lexical) conversions of the verb and cannot be decomposed syntactically into nominal/adjectival converter + suffix conjugation verb-form. 36 The common usage of preposition-n// clauses after introductory epistolary formulae in Middle Kingdom letters supplies a number of examples of this construction. This usage, which serves to introduce the main text of the letter, does not accord with the normal 'adverbial' usage of these patterns. However, the internal syntax of the construction would seem to be constant across both usages as recognized by EG §§223,225 and Johnson, op. cit. 81-3. But cf fn 39. 37 As Johnson, op. cit. 81 fn 30 notes, this example could also be interpreted as a second tense. 38 The variants have wn, see W. Helck, Der Text der "Lehre des Amenemhet I für seinem Sohn" (Wiesbaden, 1969), 22. 3 5
3
3
3
M a r k Collier
30
g) sentence with adverbial predicate 21) Ptah L2 1,14: m mh Ib.k hr-ntt twmrh Don't fill your heart because you are a wise man. h) pseudo-verbal construction 22) Sin B75-6: hr hm Kmt nfr.t(l)
-ntt s(y) rh.t(ï) rdf So, in fact, Egypt is happy because it knows that he prospers. i) circumstantial sdm(\f)Isajn.n(f) 23) Kahun 12.1, 34/50-51: hn na\ br[t] nbtprKbs^ [... k3w]ty Nbswmnw nbtprlkw r-ntt sdm.n(.1) n3 h3b.n.t hr.s And greet the lady of the house Kebs, [.. the porjter Nebsumenu, and the lady of the house Dot that I have listened to that about which you have written. Given the close siirülarity with the paradigm exhibited after initial particles, the analysis is essentially as in section l . The initial main clause constructions a)-e) are patently non-transposed/unconverted clauses; they are the basic patterns for their respective constructions and do not behave subordinately as the substitutes for nonverbal parts of speech. Equally, the circumstantial sa/n(f)/sa/n.n(.f), the bare sentence with adverbial predicate and the bare pseudo-verbal construction are the unconverted patterns for their respective constructions and do not behave subordinately after preposition-ntt converters as substitutes for non-verbal parts of speech, but as noninitial main clauses. Hence patterns a)-i) form a natural class as unconverted clauses: 39
c
40
converter
preposition-ntt
unconverted clause a) initial main clause sentence with nominal predicate sentence with adjectival predicate prospective sdm(.f) second tenses participial statement b) non-initial main clause existential sentence bare sentence with adverbial predicate bare pseudo-verbal construction circumstantial sdm(.ß/sdrn.n(.f)
39 Corroborative evidence for the presence here of the circumstantial sdm{.f)lsdmn(.f) from negation (n sdjn.n(.f)/n sdjn(.f)), passives (the passive sdjn(w)(f)) and the behaviour of verbs of motion is given in Collier, op. cit. There are no examples of the circumstantial sdm(.f)/sdjn.n(.f) after preposition-/i/r converters from the literary corpus. In the present state of our knowledge, it is difficult to determine whether this is a genuine difference between literary and non-literary Middle Egyptian or a mere matter of survival. In view of the relatively limited number of examples of preposition-nr/ constructions from the literary corpus and the literary examples exhibiting negation (n sd/n(.f}/sdm.n(.f)) and passives (the passive sajn(w)(./)), perhaps caution should be exercised. Notice that in either case, the discussion still holds for non-literary Middle Egyptian and, of course, for the bare sentence with adverbial predicate and bare pseudo-verbal construction after preposition-mr converters in literary Middle Egyptian. The problems for the adverbial analysis are the same as after initial particles. In this construction the circumstantial patterns are in a substitutional relationship with initial main clauses and not with simple adverbial expressions such as prepositional phrases. Once again, an adverbial analysis requires some emendation of the paradigm as it stands, perhaps along the lines suggested by Johnson. 4 0
Circumstantially adverbial?
31
This results in the maximally simple statement that the role of converters is to convert unconverted (main) clauses into a form (courtesy of the converter) suitable for specialized subordinate usage. 3. 'Virtual' relative c l a u s e s
41
Middle Egyptian exhibits two separate types of relative clause - the 'real' relative clause and the 'virtual' relative clause . These two types of relative clauses are well known to exhibit important differences in their syntax. However, of the two, the 'virtual' relative clauses has received little detailed attention . 'Real' relative clauses exhibit forms specific to relative clauses, i.e. forms which do not occur in other usages. 'Real' relative clauses headed by verbs display special verb-forms (participles and relative forms as adjectival transpositions of the verb) , whereas sentences with adverbial predicate and pseudo-verbal constructions display a converter, either nty or a form (participle or relative form) of the auxiliary verb wnn : a) participle 25) Leb 78-80: mhy.ihr msw.s saw m swht m3w hr n hnty n- nht.sn I shall grieve for her children who have been broken in the egg, who have looked on the crocodile-god before they have lived. b) relative form 26) Sin B28: nfr irrt.n.sn What they did was good. c) nty clause 27) Sin B166: nd.l hrt hnwt-t3 nttm h.f Mayl greet the mistress of the land who is in his palace. d) wn(n) clause 28) Peas B 1,256: n rh.n.tw wnnt m lb One does not know what is in the heart. 'Real' relative clauses exhibit typical adjectival properties: adjectival agreement with the antecedent and the ability to stand independently without (overt) antecedent (nominalized relative clauses). They also exhibit a strict strategy concerning whether the resumptive pronoun is mandatorily oven or non-overt . 42
43
44
45
c
c
46
4 1
For details, full exemplification and comparison with the adverbial analysis, see M.A. Collier, The relative clause and the verb in Middle Egyptian', submitted for publication to JEA . 42 The term 'virtual' relative clause is due to EG §195-6 (cf §182) and is retained here (in quotes) due to its familiarity, although the term is something of a misnomer (see below in the text). The term 'real' relative clause (similarly in quotes) is used to refer to the adjectivally converted relative clauses introduced by relative converters and by the participle and relative form verb-forms. 43 Perhaps because of the problems which 'virtual' relative clauses provide for the traditional definition of the relative clause as an adjectival clause. Equally, the 'virtual' relative clause provides problems for the adverbial analysis of the circumstantial sdm(.f)/sdm.n( f) in terms of adverbial substitution. See Collier, op. CiL for discussion of both of these issues. EG §§353, 380 and Polotsky, 'Transpositions' 2.1. EG §§199, 200, 396 and Polotsky, Transpositions', 3.4. For a recent discussion within the Polotskyan tradition, cf H. Satzinger, in Festschrift Westendorf 125-56. 3
4 4
3
4 5
3
4 6
Mark Collier
32
In contrast, 'virtual' relative clauses do not display forms specialized for relative clause usage. Rather the forms of the suffix conjugation, sentence with adverbial predicate and pseudo-verbal construction found are the circumstantial sajn(.f) /sdm.n(.f), the bare sentence with adverbial predicate and the bare pseudo-verbal construction : a) circumstantial sdm( f)/sdm.n( f) 29) Eb 91,3: kt nt msdr dlf mw3 hw3 Another (remedy) for an ear which gives off foul water. 30) Peas B 1,231-2: m wn hwrw hr htffn rh.n.k sw Do not rob a poor man of his possessions, a feeble man whom you know. h) bare sentence with adverbial predicate 31) Eb 51,19-20: ir m33.k s stwt m nhbt.f If you see a man on whose neck are swellings, c) bare pseudo-verbal construction 32) Peas B l , 171-3: mktwm mtnty d3 nb hmt cJc3 cjßy.ffdkw See you are the ferryman who (only) ferries the possessor of the fare, a straight man whose straightness is splintered. 'Virtual' relative clauses do not exhibit adjectival agreement with the antecedent, cannot stand independently as nominalized relative clauses and do not exhibit the same strategy concerning whether the resumptive pronoun is overt or non-overt. Indeed, in examples 29) and 30) above, the resumptive pronouns are overt in contexts where in the equivalent 'real' relative clause the resumptive would be non-overt (participle and relative form respectively). In the present framework, the differences between the two types of relative clause result from the form that the clause is forced to take; i.e. to properties of c o n v e r s i o n . The choice of relative clause seems to be linked with issues of dellniteness: 'real' relative clauses are required with definite antecedents and Virtual' relative clauses are required with indefinite antecedents . Definite antecedents require that the relative clause be converted into a specialized form exhibiting adjectival properties, including adjectival agreement. Since they exhibit adjectival conversion, 'real' relative clauses are restricted to environments compatible with such adjectival properties. In contrast, indefinite antecedents do not require that the relative clause be converted into a specialized form exhibiting adjectival conversion. Lacking conversion, Virtual' relative clauses must, then, exhibit the form of the basic unconverted pattern 47
c
48
49
4 7
3
For other possible examples of 'virtual' relative clauses, see EG §196. 48 Regardless of their conversion properties, the two forms of relative clauses share in common the properties of restrictiveiy modifying nouns (despite unsubstantiated claims to the contrary) and an antecedent-resumptive dependency (although the form of the resumptive pronoun is affected by the presence or absence of adjectival agreement, see below in the text). Hence these properties are not dependent on adjectival conversion and are not to be defined in terms thereof. Rather, adjectival conversion is to be treated as a contingent, not a necessary, property of relativization, see Collier, op. cit. for discussion. As noted in EG §§198-9 and J.G. Griffiths, JEA 54 (1968), 60-66, there are some (apparent, see Griffiths, op. cit. 66), exceptions especially with generic antecedents. However, the precise articulation of the opposition awaits a study of the interaction of definiteness, referentiality, specificity, and quantification in Middle Egyptian and so the traditional definite-indefinite distinction is retained for expository convenience. 4 9
3
Circumstantially adverbial?
33
for the respective constructions involved in the relative clause. In the case of the suffix conjugation, the sentence with adverbial predicate and the pseudo-verbal construction, the basic unconverted forms are, of course, the circumstantial sd.m(.f)/sdrn.n(.f), the bare sentence with adverbial predicate and the bare pseudo-verbal construction respectively. As unconverted forms, these forms are just those used in other environments where specialized conversion is not required and hence exhibit a wider range of usages than the forms found in 'real' (converted) relative clauses. Equally, this framework allows a unified approach to the form of the resumptive pronoun within the relative clause. In the current framework, the form of the resumptive pronoun in relative clauses (and other unbounded dependency constructions ) is integrated into a general account of pronoun omission in Middle Egyptian. Middle Egyptian exhibits two strategies of pronoun omission: omission under relevance and omission under agreement. Omission under relevance is optional and is sensitive to certain semantic/pragmatic parameters, particularly specificity and animacy. In terms of specificity, pronouns with non-specific or contextually clear reference are the more likely to be non-overt. In terms of animacy, pronouns with impersonal/inanimate reference are more likely to be non-overt than pronouns with personal reference (as a consequence, pronouns with specific personal reference tend to be overt): 33) Fish A2,3: din 0 hr ht n Sbk nb S We shall put (something) on the fire for Sobek, lord of She. 34) Meri C4,3-4: mk sp hsy hpr m k3w.î.... hpr.n.0 n-ïs m irt.n.i See a wretched incident happened in my time .... that it came about was not through what I had done. In contrast omission under agreement is mandatory in that where agreement occurs, an overt coindexed pronoun or noun cannot occur locally to (in Middle Egyptian, immediately following) an agreement-carrying expression . Since the 'real' relative clause exhibits adjectival agreement, resumptive pronouns must be mandatorily orriitted within the local domain of agreement (e.g. where the subjects of ntyjwn(n) clauses immediately follow the agreement carrier). If placed beyond the local domain, they are generally overt, being sensitive only to omission under relevance (including examples where the subjects of nty/wn(n) clauses do not immediately follow the agreement carrier ). The interaction of omission under agreement and omission under relevance thus provides the foundation for an account of a wide range of resumptive effects in 'real' relative clauses . In contrast, 'virtual' relative clauses do not exhibit adjectival agreement and hence the resumptive pronoun can only be subject to omission under relevance. In 'virtual' relative clauses, resumptive pronouns, especially with personal reference, will 50
51
52
53
50 On unbounded dependency constructions (i.e. constructions exhibiting an antecedent-resumptive pronoun dependency), see Collier op. cit Another example is the topicalization construction considered below in section 4. 1 The locality is stated here in the most informal of terms. A more precise formulation requires a proper study of the incorporation and cliticization effects of pronouns noted briefly in section 4. For omission under agreement outside adjectival environments, see the discussion of the inflected preposition ('prepositional adverb') in Collier op. cit. 52 Contrast the mandatory omission of resumptive pronoun subjects in standard nty/wn(n) clauses with the optional omission under relevance exhibited by nty n sdjn(.f) in Peas B2,80 (overt coindexed personal pronoun subject) and Nef 26 (non-overt non-specific coindexed impersonal pronoun subject). 53 See Collier, op. cit. for discussion. 5
34
M a r k Collier
tend to be overt. Thus, in addition to examples 29)-32) above, consider the following classed according to the function of the resumptive pronoun: a) subject (cf 29 above) 35) Eb 102,3-4: hbs ib.fpw ml s wnm.n.f k3w nw nht
It means that his heart is covered like a man who has eaten the unripe fruit of the sycamore. 36) Peas B 1,173-4: mktwm hry-$n w n rdi.n.fsw3 $w hrSee you aire as a chief of the storehouse who does not let one who is lacking pass in areas/immediately(?). c
37) Sh.S 61-2: gm.nX hßw pw iw.fm
c
ih
(And) I found that it was a snake which was coming. b) object (cf 30 above^ 38) Sh.S 147-8: mi Irrt n tip- mrr rmt mt3w3 nrhsw rmt As is done for a god whom men love in a far land which they do not know, c) prepositional
object
39) Sin B159-60: ptr wrt r ^bt h3t.i m t3 ms.kwl im.f What is that which is greater than burying my body in a land in which I was born. d) possessor of noun (cf 31 and 32 above)
40) Meri 5,4: m sm3 s îw.k rh.tî 3hw.f Do not kill a man whose worth you know. However, omission under relevance can also occur. Compare 36) with: 41) Adm 7,1: mtn Is ir ht n p3.0 hpr See, now, things have been done which never used to happen. In this framework, the 'virtual' relative clause is as much a true relative clause as the 'real' relative clause. Indeed the only difference between the 'real' and the 'virtual' relative clause seems to be connected with defmiteness: definite antecedents require a specially converted form exhibiting adjectival agreement (and its attendant effect on the form of the resumptive pronoun); whereas indefinite antecedents do not require a specialized converted form and unconverted forms are used. A better term for the 'virtual relative clause would be the unconverted (paratactic or juxtaposed) relative clause which contrasts with the converted relative clause (the 'real' relative clause). Separating form and function, 'virtual' relative clauses may be said to function 'adjectivally' in that they modify nouns. However, this does not necessitate that they be adjectival by form . 1
54
4. After a u x i l i a r i e s
55
Constructions with auxiliaries have been amongst the most widely discussed of Middle Egyptian constructions . The most common auxiliary element is iw, which 56
5 4
It is thus preferable to distinguish strictly between terminology of form and terminology of function. 5$ For detailed discussion, full exemplification and comparison with the adverbial analysis, see M.A. Collier, Verbal syntax in Middle Egyptian (in preparation), a revision of id. A grammatical analysis of sentences with iw in Middle Egyptian (University of London PhD thesis, 1989). 56 The literature is now too extensive to cite here. The most influential discussion remains Polotsky, Transpositions', 3.8.
Circumstantially adverbial?
35 57
will form the focus of the exemplification provided here . iw co-occurs with the following (basic) patterns: a) iw + noun/suffix pronoun + 'adverbial'expression * i) Iw + noun (suffix pronoun* prepositional phrase/simple adverb 42) Peas B2,65: iw Sdw.k m sht Your land-plots are in the countryside. ii) iw + noun/suffix pronoun + Stative 43) Sin B307-8: iw twt.i shr m nbw My statue was overlaid with gold. iii) iw + noun/suffix pronoun + preposition + infinitive 44) Sh.S 119-20: iw dpt r iit r hnw A boat will come from the residence. b) Iw + suffix conjugation verb-form + noun/suffix pronoun i) iw + circumstantial sa/n(.f) 45) Kag 2,1: ïw gr.k m r.k You are silent with your mouth. ii) iw + circumstantial sdm.n( f) 46) West 8,8: try w.s. nb.i iw in.n.iQdi Sovereign l.p.h., my lord, I have brought Djedi. iii) iw + passive sdm(w)(f) 47) Sin B297-8: iw in n.i$3b m hsp3sp4 n hrw Meals were brought to me from the palace 3, 4 times a day. At first sight, this might suggest an analysis of the tw-construction with iw as an initializer before the bare sentence with adverbial predicate, the bare pseudo-verbal construction and the bare circumstantial sajn(-f)/sd/n.n( .f) (and their passive equivalents). Indeed this will be the analysis suggested below. However, there are two well known issues which have complicated the discussion of this construction and have led to a rather different view dominating the field. The first is the common occurrence of the pattern: iv) iw + nounisufix pronoun + circumstantial sajn(f) 48) Sh.S 17-8: îw r n s nhmfsw The mouth of a man, it can save him. This pattern would sometimes seem to behave as a simple construction with iw (indeed it is more common than iw sdm( f)). In particular, iwf sajn.f can exhibit a functional paradigm with the SVO patterns of the simple pseudo-verbal construction with iw . A good example of this is: 5
c
c
59
51 The discussion can be extended to cover the other auxiliaries which exhibit a similar behaviour noted in EG §§469-83, e.g. h .n, u.n, pr.n, sdr.n, g/.n, and forms of wnn. 5% Constructions with iw are initial main clause patterns. However, in environments where there would be no other morphological support for a pronominal subject, e.g. in 'virtual' relative clause, adjunct/focus and coordinate usage of the patterns in a), the auxiliary Iw is invoked to supply the necessary morphological support (cf £G §§ 117.2, 323). Patterns with nominal subjects and patterns, such as those in b), where the pronominal subject follows the verb, do not exhibit ïw-support in such usages (cf fn 78). The internal syntax of cases with w-support is, of course, that of a construction with ïw. For homogeneity of function, the representative examples of the patterns in a) have been chosen with nominal subjects. Cf P. Vemus, in Festchrift Westendorf, 197-212 and id. in Crossroad, 379-80. Indeed a number of researchers, following W. Westendorf, MIO 1 (1953), 337-43, have proposed that the pattern noun + sdm(.f) is itself a pseudo-verbal construction. 3
c
c
3
5 9
36
M a r k Collier c
c
49) Sin Bl-2 « R24-5: ist wi h sajn.n.î hrw.f
Sin B2: îw.fhr mdt Sin R25: iwfmdw.f Now I was stood (nearby) and I heard his voice as he was speaking. The second issue is the morphological evidence that suffix pronouns suffix to the auxiliary (cf 49) and hence are standardly treated as being grammatically dependent on the auxiliary, rather than on the 'adverbial' expression which follows. Indeed both issues clearly influenced Polotsky in arriving at his most developed analysis of the construction with iw : 60
nominal subject lw noun/suffix pronoun lw nounlsuffix pronoun iw noun/suffix pronoun lw nounlsuffix pro nounl. 0 iw.0 IW.0
adverbial predicate prepositional phrase Stative preposition + infinitive circumstantial sa}n(.f) circumstantial sajn.n(.f) passive sgjn(w)(.f)
Although the details of this analysis have been much discussed, this remains the most influential analysis of the construction with lw. The basic noun-adverb division of the construction has been followed by all subsequent research within the same tradition . Moreover, most of these analyses include iw.f SQjn.f as a basic construction with lw and this has lead to general acceptance of an omitted pronominal element after the auxiliary in the patterns lw sajn(.f)/sajn.n(.f) and lw passive sajn(w)(.f) . Equally, there has been wide acceptance of the proposed syntactic-morphological division. However, the volume of subsequent discussion reflects the continued disquiet concerning these central problems and their role in the analysis of the construction with 61
62
The analysis of constructions with Iw (and other auxiliaries) in the present framework differs somewhat from the traditional analysis, most significantly in terms of the proper analysis of iwf sdmf (and hence of lw sdm.n(.f) with or without an 6 0 Polotsky 'Transpositions', 3.8-9. 6 1 Polotsky assigned the circumstantial sdm( f)/sdm.n( f) adverbial status on the basis of substitution with prepositional phrases and pseudo-verbal predicates. Some subsequent work has maintained the noun-adverb binary division of the construction, but places the circumstantial sd (f)/sdjn.n(f) in a substitutional relationship with the entire bare sentence with adverbial predicate/bare pseudo-verbal construction after iw. However, such a substitutional relationship is not necessarily indicative of adverbial behaviour, as noted in sections 1-3 above. Cf below in the text 6 2 A notable exception is F . Junjp, Syntax der Mittelägyptische Literatursprache (Mainz, 1978), 76-8. Junge proposes an analysis with iw as the nominal subject to a following adverbial clause predicate (cf last footnote). Although this analysis shares the basic noun-adverb division of the construction, it does not require an omitted pronominal element in a pattern such as lw sdjn.n(.f). iw.f sdmf is analysed as lw + [noun + sdjn.fl construction with noun + sdjn(.f) analysed as a pseudo-verbal construction, cf F . m
Junge Rd'E 3 0 (1978), 96-100.
6 3 cf the three alternative proposals for the analysis of the construction with lw proposed in W . Schenkel, Materialien zur Vorlesung. Einführung in die klassisch-ägyptische Sprache und Schrift, Wintersemester 1989/90 (Tübingen, 1989), 4.4.2.3 note and 6.4.2 note 2 . The most recent addition to the discussion is H. Satzinger, GM 115 (1990), 99-102.
Circumstantially adverbial?
omitted pronoun after iw) and the proper analysis of the ability of suffix pronouns to suffix to auxiliaries. In the present analysis, there is no place for the iw.f sdmf pattern in the analysis of the basic construction with iw; rather it is analysed as a topicalization or anticipatory emphasis construction. As a consequence, there is no place for a ïw.0 sdm.n(.f) pattern: this formation is analysed simply as Iw sajn.n(.f). Equally, the construction is not divided syntactically into the [iw.f] [m pr] division which its morphology would seem to require - rather a morphology-syntax mismatch is suggested. Since these issues have played a major part in recent discussions of this pattern, the view presented here requires some defence. Topicalization is a common, although little studied, property of Middle Egyptian . As is well known, a wide range of patterns allow an extra expression (usually, but not necessarily, a noun phrase) to be placed in a special position (usually at or near the front of the construction) which is not available in the simple nontopicalized pattern and which is usually resumed by a pronominal expression in an appropriate position ordinarily available within the construction : 51) Sh.S 151: ntyw n.ï-lm(y) sw ntyw, it belongs to me. The basic word order in this pattern is adjectival predicate-nominal subject . Here anticipatory emphasis licenses an expression in a special position at the front of the construction serving as the antecedent of the resumptive pronoun which occupies a position ordinarily available within the construction (here as the subject). Unfortunately the discussion of topicalization in relation to the suffix conjugation has been obscured by the status of the noun + sdm(.f) pattern in Middle Egyptian noted above for iw.f sdm.f. However, there is a clear paradigm of form (as opposed to function, see below) exhibited by the suffix conjugation (in common with other major Middle Egyptian patterns) in the following contexts, where an expression occurs in a fronted position not available with the simple pattern and which is linked to a construction-internal position as the antecedent to a resumptive pronoun. For comparison, the suffix conjugation forms are compared with the pseudo-verbal construction: 64
65
66
c
c
67
6 4
The term 'topicalization' is in common usage and is retained here, although this is to be taken to indicate the presence of a special position, rather than to indicate an analysis of its communicative function. On the "emphasis' standardly exhibited by TOPICs, see J.F. Borghouts, in Crossroad, 51-3 (under 'focus'). Borghouts adopts a similar analysis of topicalization in the ivv-construction to that presented here, although he does not follow through the repercussions of this analysis. To distinguish this special position, which is not available in the simple construction, from die discourse function of topic, the former will be termed the TOPIC. Topicalization is not be confused with fronting with ir. in contrast to TOPICs, ?r-fronted expressions ordinarily precede auxiliaries and initial particles (although, as 'unbound particles', the latter can also precede an ir expression or both the ir expression and the following construction (cf Hek 1,16-7)). Cf£G §§146-8. 66 This dependency exhibits the standard antecedent-resumptive dependency also found in relative clauses and certain other constructions, see Collier, 'The relative clause and the verb' MS for discussion. As an exception to the normal antecedent-resumptive dependency, topicalized adjuncts do not exhibit a dependency with a resumptive expression. EG §114.4. 6 5
6 7
3
3
Mark Collier
38
a) circumstantial sdjn(.f) i) bare * 52) Sin R20-2: n sp slnfrssy bik hfhn Smswfnn rdlt rh st ms .f He did not delay at all. The falcon, he flew with his followers, without causing that his expedition know it. ii) after initial particles 53) Sin R15-6: tisw hm ly.f In.n.f skr(w)- nh(w) n thnw mnmnt nbt nn drw.s Now, in fact, he was returning, having brought away living captives of the l/i/tw-people and all the cattle without limit. iii) after auxiliaries 54) Sin B174-6: wn.ln hm.f h3b.f n.l 3wt- nt hr-nsw s3w.fib n b3k-im mi hk3 n h3st nbt So his~person, he sent me gifts of the king's favour, so that he might gladden the heart of this servant just like any foreign ruler. b) circumstantial sajn.n(.f) i) bare 55) Sin B168-9: -ntt .fï3w h3w wgg 3s.n.fwl Because, it, old age has descended and feebleness it has hurried upon me. ii) after initial particles 56) Adm 7,7: mtn t3 ts.n.fhr sm3y See, the land, it has been knotted with confederacies. iii) after auxiliaries 57) Kag 2,7-8: h .n hm n nsw-blty Hwl-ny mni.n.f Then the person of the nsw-blty Huni he died (lit. moored). c) passive sd.m(.w)(.f) i) bare 58) Kha 12: hr nb twt hr iw sfyt rdlw s3 r.s (And) everyone alike is burdened by wrongs; respect, the back has been placed to it ii) after initial particles: no examples in the literary material iii) after auxiliaries: no examples in the literary material. d) n sdm(.f)/n sdm.n(.f) i) bare 59) Leb 104: hnmsw nw min n mr.ny 'fhe friends of today, (they) do not love. ii) after initial particles 60) Sin B223: is w rt tn irt.n b3k n hmt(.i) s(y) Now this flight which the servant made, (J) did not plan it. e) pseudo-verbal construction i)bare 6
c
c
c
c
c
c
c
c
69
6 8
For discussion of the form and functions of this pattern, see E. Doret, JNES 39 (1980), 37-45. For Doret, noun + circumstantial sdm(.f) is a pseudo-verbal construction which has the additional function of optionally placing 'emphasis' on the subject, cf fn 77. 69 in the pseudo-verbal construction the subject is ordinarily placed before the other members of the bare construction. Hence we would not expect tofindexamples exhibiting a TOPIC-subject dependency except under exceptional circumstances.
Circumstantially adverbial?
39
61) Adm 7,7: kn hsy hr nhmw [htj.f The brave man, the coward takes away his property. ii) after initial particles 62) Adm 7,4: mîn sSt3 n î3 hmm drwf sh3w See the secrets'of the land which were unknown, its limits are revealed. iii) after auxiliaries 63) West 9,12: wn.in hm.f ib.fw3 r d.wt hr.s So his person, his heart fell into sadness because of this. Notice that the TOPIC precedes the suffix conjugation or pseudo-verbal construction but is itself preceded by both initial particles and auxiliaries. In each of these cases, these patterns are to be carefully distinguished by form (and indeed function) from vocatives. Vocatives address the interlocutor directly and can occur at the beginning, throughout or at the end of constructions . Most interesting to note are the occurrences of vocatives before initial particles, which thus contrast with the position of TOPICs: 64) Nef 5-6: dd.in hm.f .w.s. n.sn rhw min rdl.n.i i3 s".tw n.tn r rdit.... So his person l.p.h said to them, 'Companions, see I have had one call to you to cause 65) Peas B 1,219-20: wh3 mktw ph.t(ï) Fool, see you are reached. When we turn to the construction with iw, then the pre-auxiliary expressions found are vocatives : 66) West 8,7-8: dd.ïn s3-nsw Hr-dd.fiîy .w.s. nb.l iw in.n.iDdî So the king's son Hordedef said, 'Sovereign l.p.h., my lord, I have brought Djedi. Equally, constructions with iw (just like the other constructions with auxiliaries noted above) exhibit patterns where an extra expression appears in a special position after the auxiliary and before the remainder of the pattern which is not available in the simple 1Q
11
72
c
c
73
c
7 0
'Admonitions' is arichsource of topicalization patterns ill-evidenced elsewhere, due to the thematic organization of the text around the object of the particular admonition and the reiterative usage of mtn is and iw ms. 71 This pattern is quite common with parts of the body subjects, as noted by Polotsky, Transpositions', 3.8.7.1 and £ G §215. Both Polotsky and Gardiner analysed these constructions as exhibiting a pseudo-verbal construction with a clausal predicate, although the latter noted the alternative possibility of analysing the pattern as a topicalization construction. However, this does not account satisfactorily for the antecedent-resumptive relations in this pattern where a resumptive expression within the pseudo-verbal partem is linked to the TOPIC antecedent. EG §87. 3 There may be prosodie restrictions on the 'size' or 'heaviness' of the TOPIC which can separate an initial particle/auxiliary from the following pattern and large TOPICs may be forced to precede the initial particle/auxiliary (cf D. Zee and S. Inkelas, Prosodically constrained syntax', in The phonologysyntax connection, eds. S. Inkelas and D. Zee (Chicago, 1990), 365-78). A possible example with tw maybe: FN1) Sin BlOl-4: h3st tn rwit.nJ r.s ïw lr,n.l hd im.s Every land against which I advanced, I made an attack against it. However, as noted by Blackman Middle Egyptian stories (Brussels, 1932) 24a n 1 la, the auxiliary was added to the manuscript later. 3
7 2
7
3
M a r k Collier
40
construction and which exhibits an antecedent-resumptive relation with a resumptive pronoun in a position ordinarily available within the construction: a) iw + pseudo-verbal construction 67) Peas B 1,284: In iw wsfwspry r h rrnpr.f As for the sluggard, will a petitioner stand at the door of his house ? b) lw + passive SQjn(w)(.f) 68) Adm 6,8-9: iw ms sX nw tm(3) drw ss(w).sn Surely, the scribes of the cadaster(?), their writings have been destroyed. c) iw sdm.n(.f) As is well known, there are no standard examples of lw noun sdm.n.f'm (most) Middle Egyptian source genres . However, important corroborative evidence is provided by adjunct topicalization of the time phrase min : i) iw + pseudo-verbal construction 69) Sin B149: iw min ib.fï Today, his heart is washed. ii) lw + sdm-n(f) 70) Peas B 1,180-1: sdmw n 3 sdm.n.k tm.k tr sdm hr-m iw min 3 hsf.n.i 3dw Hearer, you are not listening. Why don't you listen? Today, I have warded off the savage one. 71) Sin B189-90: lw mln is $3 .n.k tni Today, now, you have begun old age. d) iw sdm(.f) In terms of this paradigm of form, the iw.f sdm.f çattem would seem to find a natural place: 72) Sh.S 17-8: lw r n s nhn.fsw The mouth of a man, it can save him. If these constructions with iw are not analysed as topicalization constructions, then constructions with lw (and indeed constructions with auxiliaries in general) would exhibit an unexplained gap: alone amongst main clause formations, they would not exhibit topicalization at all, since there is no other likely candidate for such a pattern. At the same time, they would exhibit patterns which would seem to display an extra expression in a special position and which seems to participate as the antecedent in an c
c
74
75
76
c
7 4
3
After £G §148,3. D. Silverman, Interrogative constructions with jn and jn-ßv in Old and Middle Egyptian (Malibu, 1980), 44-5 with fh.252 understands "Will a sluggard who petitions wait at the door of his house?' with the comment that wsfw cannot be in anticipatory emphasis because 'we would expect it to appear outside the question'. As the discussion in the text indicates, this does not hold. Notice that the peasant has already alluded to Rensi's sluggardly properties (iw wsf.k r tht.k) in B 1,281. However, as noted by Polotsky Transpositions' 3.8.5, examples do occur in the Coffin Texts: FN2) CT I 74i: iw Wp-w3wt wp.nfn.fw3wt nfr(w)t Wepwawet, he has opened the perfect ways for him. The reason for the lack of standard iw TOPIC sdjn.n(f) patterns in other Middle Egyptian source genres remains obscure but is perhaps to be sought in terms of functional/communicative concernsratherthan in terms of syntax. ?6 Gardiner considered min to function as an enclitic particle in these examples (EG §208). However, unlike enclitics, mln can occur initially infrontedpositions (cf the difficult Adm 12,7) and in standard word order positions as subject, object and adjunct As a topicalized adjunct min is not linked to a resumptive expression (cf fh 66 above). 7 5
3
antecedent-resumptive dependency. It is difficult to avoid the conclusion that, by form, noun + suffix conjugation patterns (including iw.f s&m.f) are topicalization constructions exhibiting a TOPIC in a special position as the antecedent to a resumptive pronoun. The parallel with the initial particle construction and other constructions with auxiliaries where the TOPIC follows the initial particle and precedes the remainder of the pattern is striking. Concerning iw.f sdm.f, it would seem necessary to conclude that by form this pattern belongs to the paradigm of topicalization constructions. However by function (courtesy of the resulting 'SVO' order), iw.f sdm.f'can exhibit functional usages where it would seem to behave as a simple declarative clause like the simple SVO ordered pseudo-verbal constructions with Iw . If we adopt this analysis, then many of the traditional problems associated with the iw.f sdm.f pattern (and indeed noun + suffix conjugation patterns in general) can be finessed. As a topicalization construction by form, i.e. with an extra noun phrase in a special structural position not available in the simple non-topicalized pattern, the iw.fsdmf pattern has no place in the paradigm for the analysis of the form of the simple (i.e. non-topicalized) construction with iw. The repercussions of this re-analysis of the paradigm for the simple construction with iw are major. Firstly, along with other evidence, it removes the motivation for an analysis of the pattern iw sdm.n(.f) as iw.0 sdjn.n(.f) with an omitted pronoun after iV . Rather this pattern is to be analysed simply as iw sdm.n(.f) (see below). Secondly, this 11
78
' ' That is to say without necessarily exhibiting the informational effects commonly found with TOPICs (cf fn 64). Compare the standard Linguistic distinction between the form of a construction and its usage, where, as in many languages, constructions with the form of a simple declarative statement can be used interrogatively (£G § 491) or where constructions with the form of an interrogative statement can be used conditionally (EG §489, Silverman, op. cit. 105-8). However, we would expect to find examples where lw.f sdjnf exhibits 'emphasis' on the TOPIC and indeed examples seem to be forthcoming (cf Doret, op. cit. 43): FN 3) Peas Bl.215-7: sw3 hr sp îwf r snwy in wnm dp iw wiSdw wtb.f in sdrw m33 rswt Pass over a case and it will become two. It is the eater who tastes. He who is addressed, he answers. It is the sleeper who sees the dream. Given this 'tension' between form and function, it is not surprising, I suggest, that two of the major changes in the transition to Late Egyptian are the loss of the circumstantial sdm(.f) and the change in function of iw. 78 The postulated omitted pronoun does not abide by the standard conventions for pronoun omission (cf section 3): omission underrelevance,which is optional; and omission under agreement, which is clearly inappropriate here. The pattern lw.0 sdrn.n(f) has also been compared to examples such as: FN 4) Sin B43: lw.0 ml shr ny (It) was like the plan of god. (cf M.A. Collier, Tieview', DE 18 (1990) forthcoming for brief arguments against analysing iw as the direct subject of the adverbial predicate). However, once word order has been taken into account, it is clear that the equivalent suffix conjugation pattern is: FN 5) Herds 23: iw ir.0 mi dd.f (It) was done like he said, where the non-overt pronominal subject follows the verb-form (cf below in the text). Moreover, the purported lw.0 sdm.nif) pattern is not used in contexts where fw-support is required, although iW-support is required both with overt pronominals (cf the iw.f sdjn.f pattern in 49) and non-overt pronominals: FN6) Meri 4,5: hm-ht pw hnty iw.0 n kwy He is a fool the one who is greedy when (things) belong to others. 3
3
Mark Collier
42
discussion of topicalization undermines the analysis of the structure of the entire construction with iw as basically that of either a sentence with adverbial predicate or a second tense, with iw as the nominal subject to an adverbial predicate (cf 50 above), since in these construction types the TOPIC occurs before the entire construction, not after the nominal subject (whether simple nominal or that-form) . The second innovation follows recent developments in generative grammar. Over the last few years it has come to be realized that syntax and morphology do not necessarily stand in a transparent relationship; i.e. the morphological analysis of a construction may not agree exactly with the syntactic analysis of the construction . The ability for such 'mismatches' to occur is, however, extremely limited, as one would expect Consider the following two examples: 73) Kwakwala: k ix id-ida beg anema-x-a q'asa-s-is t'elwag ayu clubbed-the man-OBJ-the otter-INST-his club The man clubbed the sea-otter with his club. 74) West Greenlandic Eskimo: Tuttu-p neqi-tor-punga reindeer-REL meat-eat-INDIC/ls I ate reindeer's meat. In 73), the detenniners are enclitics which cliticize to the immediately preceding word, although they are grammatically related to the following noun . In 74), the noun root incorporates with the verb but is directly grammatically related to the external noun possessor. In Middle Egyptian, cliticization is important for the understanding of the behaviour of dependent pronouns, n+ pronoun, certain determiners (nb, pltn, pltf) and enclitic particles. Incorporation is important for the understanding of suffix pronouns, which can be treated as incorporated pronouns . As noted above, it is clear that the morphological analysis of the simple construction with lw with following pronominal subject is: 75) [ïw.f] m pr where the pronoun incorporates with the auxiliary. Equally, this applies when the pronoun is a TOPIC: 76) [iw.f] sdm.f However, as discussed above, the syntactic analysis of the iw.f sdm.f construction as a topicalization pattern (along with other evidence not discussed here) 79
80
w
9
w
w
81
S2
83
84
7
3
9 Cf £G §148 and 61) above (pseudo-verbal construction), Adm 9,5 (second tense). 80 Convenient discussion in J.M. Sadock, 'Auto-lexical syntax: a proposal for the treatment of noun incorporation and similar phenomena', Natural Language and Linguistic Theory 3.4 (1985), 379^439, id. "The auto-lexical classification of lexemes', in Theoretical morphology, eds. M. Hammond and M Noonan (San Diego, 1988), 271-90 and M.C. Baker, Incorporation: a theory of grammatical junction changing (Chicago, 1988). 81 From S.R. Anderson, 'Kwakwala syntax and the government-binding theory', in Syntax and semantics 16: the syntax of native American languages, eds. E.D. Cook and D.B. Gerdts (San Diego, 1984), 24 ex 1). For typographical reasons, 'e' is used in the gloss in place of Anderson's inverted 'e'. 82 From J.M. Sadock, "Noun incorporation in Greelandic: a case of syntactic word formation', Language 56 (1980), 309 ex 33). In the synonymous non-incorporated sentence (ibid, ex 32), the noun root and the possessor form a phrase. Cf Baker, op. cit. 96-7. 83 Cf J.L. Klavans, "The independence of syntax and phonology in cliticization', Language 61 (1985), 95-120. 8 The syntactic and morphological behaviour of pronouns (and expressions which exhibit a similar morphological behaviour) is the subject of my post-doctoral project for the British Academy entitled 'A generative grammar of pronominals and anaphors in Middle Egyptian'. 4
Circumstantially adverbial?
43
indicates that iw stands at the periphery of the clause and co-occurs with the bare circumstantial sdjn(.f)lsdm.n(.f), the bare sentence with adverbial predicate and the bare pseudo-verbal construction. There is thus a syntax-morphology mismatch. Syntactically, iw is an initializing auxiliary which co-occurs with the clausal patterns just noted. Morphologically, however, the auxiliary is an incorporation host and can incorporate a pronoun as long as it is locally available to the the auxiliary . From this discussion, the paradigm for the syntactic analysis of the form of the simple construction with iw reduces to: 85
initializer iw
non-initial main clause circumstantial sdjn(.f)lsdm.n(.f) bare sentence with adverbial predicate bare pseudo-verbal construction
In this analysis, the circumstantial sdm(.f)/sdjn.n(.f) do not exhibit a substitutional relationship with adverbial predicates in the construction with ïw, but with the entire bare sentence with adverbial predicate and bare pseudo-verbal construction. Along with the discussion of the position of the TOPIC above, this suggests that the grammatical relation between ïw and the following pattern is not to be analysed in terms of subordination and predication relations based on nominal and adverbial substitution, but is to be analysed in much the same way as the grammatical relation between converters and initial particles and the patterns which follow them. Similarly, then, the patterns which follow ïw and other auxiliaries are to be treated synchronically as unconverted clauses behaving as non-initial main clauses, rather than as forms specialized for adverbial subordinate behaviour . Auxiliaries impose the further collocational restriction that such patterns be headed by non-nominal predicates. However, non-nominality does not reduce to adverbial properties in this case; rather it must be treated in terms of the inherent non-nominality of (unconverted) verbal expressions and adverbial expressions (see section 0). Thus the substitutional relations between verb-form and prepositional phrase/pseudo-verbal predicate are those within the clause forms of the circumstantial sdm{.f)lsdm.n(.f) (+ other dependents) and the bare sentence with adverbial predicate and bare pseudo-verbal construction: 86
AUX ïw Iw
non-initial main clause noun/ adverbiallpseudopronoun verbal predicate sdm.n
noun/pronoun
8 5
In generative terms, there is subject-to-aux incorporation exhibiting standard ECP locality effects (cf Baker, op. cit. chapters 2 and 3). Notice that when the pronoun is not locally available to iw, as in the case of iw sdm.nf, the pronominal subject cannot incorporate with ïw (rather the pronoun incorporates with the verb in this case) and hence iw remains without a morphological affix. There is thus no morphological requirement to invoke an omitted pronominal dependent after ïw in this pattern (i.e. there is no pattern iw.0 sdjn.n.f), cf fn 78. 86 it must be stressed that this is a synchronic analysis. Synchronically, auxiliary + non-initial main clause constructions do not display the syntactic properties of second tenses (nor indeed of the sentence with adverbial predicate); however, it may be that such constructions developed diachonically from original second tense formations, cf Polotsky, Transpositions, 3.8.2 on h .n. c
c
44
Mark Collier
Here the adverbial/pseudo-verbal predicate follows its subject in the sentence with adverbial predicate and pseudo-verbal construction with iw, whereas the suffix conjugation verb-form precedes its subject (and hence there is no predicate constituent) in accordance with the well known word order properties of these patterns found in the other environments in which they occur . Thus the verb-form is not ordered according to any purported 'adverbial' properties, but is ordered simply as a suffix conjugation verb-form. 87
5 Adjunct/focus
position
The adjunct/focus position allows a rather wide range of expressions. The following exemplification is of simple (i.e. non-focused) adjuncts * a) simple prepositional phrases 79) Sin B17-8: Ssp.n.ï ksw m b3t (And) I took up a crouched position in a bush. b) prepositionally converted clauses 80) Ptah L2 1,14: m mh ïb.k hr-ntt twmrh Do not fill your heart because you are a wise man. c) circumstantial sdm(.f)lsdm.n(.f) 81) Sh.S 14-5: ih ws2>.k wSd.t(w).k Then you will answer when you are addressed. 82) Sin R15-16: ii sw hm ty.f în.n.f skr(w)- nh(w) n ihnw Now, in fact, he was returning after he had brought away living captives of the r/t/nv-people. d) bare sentence with adverbial predicate 83) Fish B4.5-6: h3.i r S" stsw hr rmn .\ w.î r s3.i I will go down to the lake with the staves upon the shoulder of my arm and my poles at my back. e) bare pseudo-verbal construction 84) Sin Bl-2: sdm.n.t [hr]w.fiw.fhr mdt (And) I heard his voice as he was speaking. f) bare time phrases® 85) Men 5,7-8: dßdßt rsy wdf s3ryw rh.n.k tm.sn sfn hrw pfn wdp m3Îr The southern d3d3t which judges the needy, you know that they are not lenient on that day of judging the poor. In adjunct/focus position, the circumstantial sdm{.f)lsdjn.n{-f) do not substitute just with prepositional phrases. They occur in a paradigm which also includes expressions which do not exhibit a form specialized for adverbial behaviour. Thus, the clause forms of the sentence with adverbial prediate and bare pseudo-verbal construction occur in adjunct/focus position, forms which we have seen can also occur in non-adverbial substitutional environments. Equally, it is clear that bare time phrases display the syntax of noun phrases . Using hrw as an example, in 85) hrw co-occurs 88
c
c
c
90
87
3
Cf£G §§27and28. 88 The discussion holds equally for focused adjuncts in 'second tenses'. 89 Bare time phrases are used as the exemplars of bare nominals used as adjuncts. Other examples include ds reflexives (Adm 2,12, Ptah 7,8) and n-ts negated nouns (Ptah 6,1 and P. Ram II vs ii 5). 90 Details, discussion and arguments against an anaysis of bare time phrases as unmarked adverbial transpositions is given in M.A. Collier, The syntax of time phrases in Middle Egyptian' MS.
Circumstantially adverbial?
45
with demonstrative and indirect genitive dependents in adjunct/focus position, a behaviour characteristic of simple nominals . Moreover, exactly the same forms can occur in typical nominal positions such as subject and object: a) subject 86) Sin B309-10: Iw.l hr hs(w)t nt hr-nsw r-iwt hrw n mnî I received the king's favours until the day of mooring came. b) object 87) Sin B190-1: sM.n.k hrw n krs (And) you have considered the day of burial. Of course, ordinary bare nominals do not occur in adjunct/focus positions. However, ordinary nominals do not carry inherent semantic roles and need to be assigned a semantic role in a situation/proposition by some other expression, such as a verb or a preposition. Since adjuncts are expressions which are added to a construction to convey additional information, simple nominals are incompatible with this function, since they would be left without a role in the resulting situation. However, time phrases do carry an inherent semantic role, indicating a temporal setting, and are thus eminently suited to adjunct usage. Hence there would seem to be little motivation in assigning bare time phrases adverbial properties of form in adjunct/focus position. In the present framework, bare time phrases are uniformally analysed as nominal expressions which can function as subjects, objects and adjuncts. By function, bare time phrases in adjunct/focus position might be said to behave 'adverbially' ; however by form they are clearly noun phrases. It would seem that, in terms of form, the adjunct position does not present a homogenous paradigm, but is rather a heterogenous collection of expressions which are suited to being adjoined to a predication to add additional information without the mediation of some other expression. Concerning the circumstantial sdjn(.f)/sajn.n(./), the bare sentence with adverbial predicate and the bare pseudo-verbal construction, it would seem that their properties of conversion and part of speech (and relative tense/aspect properties) conspire to enable them to function as adjuncts. As unconverted clauses, they describe situations, which, lacking specialized conversion for specialized subordinate distribution incompatible with adjunct requirements, can be suitably added to another expression to convey additional information . Equally, there is no doubt that adjunct usage does not naturally favour nominals, nominally headed patterns or forms converted into a form specialized for nominal usage. Thus, as unconverted inherently non-nominal clauses, the circumstantial sgjn(.f)/sdjn.n(./), the bare sentence with adverbial predicate and the bare pseudo-verbal construction are suited for adjunct usage. There is no requirement for these expressions to be analysed as forms specialized by conversion/transposition for adverbial behaviour to treat their occurrence in adjunct/focus position. By function, the circumstantial sd/n(.f)lsdjn.n(,f), bare sentence with adverbial predicate and bare pseudo-verbal construction may be said to behave 'adverbially' in adjunct/focus position; however, by form they are unconverted forms. 91
92
93
9 1
3
EG §§85, 111. 92 Once again, it would be preferable to distinguish strictly between the terminology of form and function. 93 Cf Polotsky's discussion of 'parataxis' in "Egyptian tenses', §15.
Mark Collier
46 6. Coordination
Coordination is another area of Middle Egyptian grammar which is in need of detailed investigation . Partly, this is because clausal coordination in Middle Egyptian is not overtly marked. This is particularly problematic for the analysis of coordination between main clause patterns. Indeed, unless two coordinated clauses 'share' some information which is present in only one conjunct, it is extremely difficult formally to identify coordination between most main clause types. Fortunately the patterns under consideration here can be studied, since they exhibit main clause usage both with and without introductory expressions. In this section, coordination is discussed in relation to the bare non-initial main clause patterns of the circumstantial sdjn(.f)/sdm.n(.f), sentence with adverbial predicate and pseudo-verbal construction and with these constructions introduced by the auxiliary iw. In recent treatments, two properties of coordination have not received the attention they deserve. Firstly, the ability to coordinate with main clauses is not restricted to non-initial main clauses. Examples occur between j'w-constructions (initial main clause patterns): 88) Meir II, pi 8: ïw w°b ïw hwy nn r-drw n Wsir h3ty- Wh-htp pn All these are pure and abundant(?) for the "Osiris, the h3ty- , this Wekhhotep. In 88) the first conjunct with ïw shares the subject and other dependents of the second conjunct with iw. Secondly, the circumstantial sg/n(.f)/sgjn.n(f) can occur in coordination with constructions within which they themselves cannot occur (except as adjuncts) : 89) West 11,26-12,1: S3s pw ir.n t3 wb3t wn.n.s t3 t h .n sdm.n.s hrw hsy .... The servant girl went and she opened the door. Then she heard sounds of singing In this example wn.n.s cannot be an adjunct clause with relative past interpretation (she could hardly open the door without reaching it), nor can it be a that-form in a second tense since there is no following 'adverbial' expression which can be focused. Hence wn.n.s must be a circumstantial sdm.n(.f) functioning as a non-initial main clause in coordination with the preceding initial formation. 90) Sh.S 54-6: Ut.idJ shpr.n.i ht ir.n.i sb-n-sdt n ntrw I cut a fire drill, Ï made a fire and I made a burnt-offering to the gods. Once again, regardless of the proper interpretation of Stfr.z , the two sdm.n( f) clauses must be temporally coordinate rather than relative past and hence are to be analysed as circumstantial sdm.n( f)s coordinated in non-initial main clause usage. In both examples, the sajn.n(.f) 'leans' temporally on the preceding context for tense/aspect support. However, in neither case are the circumstantial sajn.n(.f)s syntactically dependent on the preceding pattern (or any member thereof) as a 94
c
c
95
c
c
c
96
9 4
3
Cf EG §§91. 487-8, 505. For the circumstantial sdm(.f)/sajn.n(.fl in coordinate clauses, see E. Doret BSEG 2 (1979), 13-22 and J.H. Johnson, Serapis 6 (1980), 69-73. 95 Contra Johnson, op. cit. 71. 6 Whether analysed as sdjnt(f) (cf H. Satzinger, JEA 57 (1971), 58-69) or narrative infinitive (cf W. Schenkel, GM 4 (1973), 23-8. 9
Circumstantially adverbial?
47
subordinate expression; rather they are coordinated with the preceding pattern as noninitial main clauses. However, it is certainly the case that the circumstantial sdjn(.f)/sdjn.n(f), the bare sentence with adverbial predicate and the bare pseudo-verbal construction prefer coordination with constructions within which they themselves can occur: a) circumstantial sd/n( f)/sajn.n( fl 91) Meri 9,4-5: iw di.n.i hwi st t3-mhw h3k.n.l farw.sn nhm.n.i mnmnt.sn I caused that t3-mhw strike them, I carried off their inhabitants and I seized their cattle. b) bare sentence with adverbial predicate 92) Peas R46-7: iw w3t.f w t hr mw kt hr it Its one side was under water and the other was under barley. c) bare pseudo-verbal construction 93) Leb 107-8: iw sf3k nht hr h3w n bw-nb Gentleness has perished and the strong man descends upon everyone. In these cases, the coordinated non-initial main clause coordinates either with the Iwconstruction or (perhaps more likely) with the non-initial main clause which follows iw in the first conjunct. Given this discussion, it would seem inappropriate to posit an analysis of the coordinate usage of the circumstantial sdjn(.f)/sd/n.n(.f), the bare sentence with adverbial predicate and the bare pseudo-verbal construction in terms of adverbial behaviour. Rather, the present framework again makes use of the analysis of these patterns as unconverted expressions. As unconverted expressions, i.e. as forms which are not specialized for subordinate behaviour, these forms can be used as non-initial main clauses. In the present framework, non-initiality is treated as a relative tense/aspect property, rather than a property of adverbial form. Hence non-initial main clauses are syntactically independent clauses which require temporal support to satisfy their relative tense/aspect properties. An already 'tensed' initial main clause pattern is, of course, an ideal pattern for the circumstantial sdrn(.f)lsa\rn.n(-f), the bare sentence with adverbial predicate and bare pseudo-verbal construction to enter into coordination with, since they can 'share' the time point of the first conjunct. However, this does not require that they be syntactically subordinated to the first conjunct or any member thereof. In this framework, by treating the non-initiality of the circumstantial sd/n(.f)/sd_m.n(.fl, hare sentence with adverbial predicate and the bare pseudoverbal construction as a property of relative tense/aspect rather than of adverbial form, such examples can be analysed simply as sentential coordination between main clause patterns involving no gapping , subordination or predication relations between the conjuncts. c
t n e
97
y /
Doret, op. cit. 22 fn 41 refers to the coordination of the circumstantial sdjn(f)lsdjn.n(.f), the bare sentence with adverbial predicate and the bare pseudo-verbal construction as exhibiting 'gapping' of an auxiliary (cf Johnson, op. cit. 71-2). However, 'gapping' refers to a particular kind of omission in coordination of an element without which the pattern would not be syntactically well-formed (with the required meaning), cf R.A. Hudson, 'Gapping and grammatical relations'. Journal of Linguistics 25.1 (1989), 57-94. The circumstantial sdrn(.f)/sdm.n(.f), the bare sentence with adverbial predicate and the bare pseudo-verbal construction are clearly syntactically well-formed without the presence of the auxiliary and hence occur in a variety of main and subordinate clause usages. Gapping, along with other 'sharing' of material, does occur in coordination in Middle Egyptian, e.g. Sin B307-8,
Mark Collier
48 7 Control
Polotsky first isolated the circumstantial sgjn(.f)/sdjn.n( f) in the substitutional environment of the oblique complement after verbs such as gmi and m33 and after the prt pw ïr.n.f construction . The interesting case is with the gmï/m33 type of verb which will form the focus of this section : a) circumstantial sdm( f) 94) Sin B52-3: m33.t(w).f h3.f r pdj(yw) When he is seen going down against the bowmen. b) pseudo-verbal predicate 95) Fish B2,4-5: m33.l msw Hwt-ihyt hr km[(3)] wx3yt I see the children of Hwi-lhyt throwing at the wi3yt birds. 96) West 12,20-1: gm.n.f Rd-Q.dt hms'.ti tp.'s hr m3st.s (And) he found Reddjedet sitting with her head on her knee. In this environment, the circumstantial sdm( f)/sdjn.n(.f) do substitute with, and indeed only with, simple prepositional phrases and pseudo-verbal predicates. In the present framework, this is simply treated in terms of the shared non-nominality of the circumstantial sdm(.f)/sdm.n(.f) and prepositional phrases/pseudo-verbal predicates (cf section 0). However, there is another issue involved in this paradigm. It is significant that the pseudo-verbal predicates must appear in this environment without an overt subject (nominal or pronominal); the understood subject of the pseudo-verbal predicates is, of course, clearly the object of gmt/m33 . Equally, the suffix conjugation verb-forms must appear with an incorporated pronominal subject coindexed with the object of gmî/m33 and not with an overt nominal subject or non-coindexed pronominal s u b j e c t . Such a distribution, where the subject argument of a predicative 98
99
100
101
102
98 Polotsky, 'Egyptian tenses', §4. 99 As in Polotsky's original discussion, this section focuses on verb-forms in the oblique complement position after gmi and m33. Polotsky had earlier noted gmi/m33 as verbs of 'incomplete predication', where the oblique complement is required to complete the sense of the verb in this usage and hence is suited to the role of focus in a second tense (RdE 11 (1957), 110 fn 1) However, it must be emphasized that 'incomplete prédication' is a lexical property and holds equally for gml/m33 in occurrences where the oblique complement is not in focus: FN7) Sin B116-7/R141-2: rkt lb pw hr m33.fwl hr Irt wpwt.k It is ill-will because he sees me doing your commissions. 100 m33 and gmi prefer the circumstantial sdjn(-f) in oblique complement position. 101 That this is indeed the correct analysis of this construction and not gmi/m33 with [noun + sd.m.f/ps&\ido-\eibäl predicate] clausal complement is indicated by examples such as: FN8) CT II ll0i-j: Is sp sn b3.l m3 tw s pf h .t(l) hft-hr.fm bw nb nt(y).flm Go go, my b3, so that that man may see you stood opposite him in whichever place he is in. Here the object pronoun of m3 cliticizes to the verb and precedes the subject noun separating the object from the oblique complement which it controls. Equally, the object can undergo standard grammatical function changes, e.g. passivization (cf 94 where, arguably, ./ is the subject of the passive form m33.t(w)). gmi/m33 can also occur with ordinary adjuncts, where the adjunct clause is not as closely related to the main verb and merely adds to the construction rather than being a lexically determined oblique complement required to complete the sense of the verb. In this case, there is no mandatory restriction on the nature of the subject, cf section 5. c
1 0 2
c
Circumstantially adverbial?
49
complement (or adjunct) must be coindexed with an argument of the matrix predicate, is often termed 'control' . In the present case, the object of gmî/m33 may be said to 'control' the pseudo-verbal predicate/suffix conjugation verb-form . Moreover, the oblique complement is effectively restricted to being phrasal rather than clausal. Thus, the pseudo-verbal predicates clearly behave here as phrasal predicates lacking a subject expression which they find indirectly in the object of gmï/m33) . However, the morphological properties of the suffix conjugation verbform disallow it from appearing without a subject except under particular and welldefined circumstances. Outside of coordination, the only exceptions are locally licensed by omission under agreement (which is clearly not appropriate here) or omission under relevance (which is optional and once again inappropriate here), cf section 3. Hence, even in control environments, the suffix conjugation verb-form must appear with a suffix pronoun subject. However, this subject pronoun must be coindexed with the controller. Since the coindexed pronominal subject cannot behave independently, the suffix conjugation f*>rm has the behaviour of a predicative phrase - i.e. as a predicate which indirectly finds its subject in some preceding nominal element under control . As a consequence, it is not simply a matter of substitution that the circumstantial sd/n(.f)/sd/n.n(.f) occur in a strictly delimited paradigm with pseudo-verbal predicates (and indeed simple prepositional phrases) in the oblique complement position after gmï/m33; it is also a product of control . In particular, it is to be noted that the bare pseudo-verbal construction and the circumstantial sd/n/sdm.n + noun have no place in this paradigm. Within the present framework, the ability of controlled pseudo-verbal (and indeed simple adverbial) predicates to occur in a substitutional paradigm with the controlled circumstantial sdjn.f/sg/n.n.f can be readily explained by the inherent nonnominal properties of both forms and by control. In the particular case of gmi/m33, as a lexical property both gmi and m33 can co-occur in this usage with either an object clause or with an object nominal and a non-nominal object-controlled oblique complement. 103
104
105
106
107
103 j h i g e covers both the 'equi' and 'raising' constructions of early transformational grammar, following work in the unification tradition in generative grammar. Cf J. Bresnan, 'Control and complementation', Linguistic Inquiry 13 (1982), 343-434, particularly 372. Control can be thought of as 'indirect predication'. Moreover, these patterns co-vary with patterns where gml/m33 takes an object clause. A good example is provided by the following variants of CT I 391a: FN9) SIC 481: m33.sn prr.l m k3r FN10) S2C 267: m33.sn wl pr.kwl m k3r When they see me leaving the shrine. This particular form of control is often termed 'raising to object', following early transformational grammar (cf J.P. Allen, in Crossroad, 14-5). In work within the unification tradition, however, gmilm33 lexically select the two observed patterns, whose synonymy is accounted for by control. Cf Bresnan, op. cit. 9.6 with fig. 21 !05 Contra Polotsky's heading II "Verb-forms in clauses of circumstance', "Egyptian tenses', 2. 106 Thj j stated in the most informal of terms. The ability to control an incorporated pronominal is to be analysed in terms of small clauses and government (MS in prep). 107 Similarly for the prt pw ir.n.f construction which exhibits adjunct control. However, this pattern also, of course, exhibits simple adjuncts without control and hence a wider paradigm of forms is to be found, cf section 5. s u s a
s
s
Mark Collier
50
Conclusion It has been argued that the verbal analysis of the circumstantial sg/n( f)/sd/n.n(.f) and the framework in which it is couched can provide a satisfactory analysis of the distribution of the circumstantial sg/n(.f)lsdjn.n(f) in Middle Egyptian. Indeed the framework leads to novel interpretations of grammatical phenomena which have proved problematic for the Polotskyan tradition. In this analysis, the syntax of the circumstantial sgjn(.f)/sgjn.n(.f) is to be analysed in terms of their nature as unconverted inherently non-nominal verb-forms, i.e. as simple verbal verb-forms. This contrasts markedly with the Polotskyan approach where the syntax of the circumstantial sdjn(.f)/sd/n.n(.f) is reduced to the distribution of simple adverbial parts of speech (principally prepositional phrases and pseudo-verbal predicates) based on substitutional paradigms. From the perspective of the present framework, Polotsky was misled by an unfortunate historical accident: he utilized the control environment of the oblique complement position after gmi/m33, discussed in section 7, as a strictly delimited environment in which to identify the circumstantial sdm(.f)/sdjn.n('./). This environment does indeed present the circumstantial sdjn(.f)lsçbn.n(.f) in a strict substitutional relationship with prepositional phrases and pseudo-verbal predicates. However, this distribution cannot be treated simply in terms of paradigmatic substitution, since it exhibits the syntagmatic dependency of control (indirect predication) which severely restricts the range of expressions which can appear as the oblique complements of gmi/m33. In particular, sg/n noxmjsdjn.n noun and noun + prepositional phrase predicate/pseudo-verbal predicate patterns (i.e. the bare sentence with adverbial predicate and the bare pseudo-verbal construction) have no place in this paradigm. It is thus unsatisfactory to present a generalized adverbial analysis of the circumstantial sgjn(.f)/sQjn.n(.f) on the basis of this environment without first taking control into account. However, when we investigate environments where the circumstantial sdm(-f)/sdjn.n(.f) occur with both pronominal and nominal subjects (i.e. exhibiting direct predication), then these patterns behave as clauses which share a close substitutional relationship with the bare sentence with adverbial predicate and the bare pseudo-verbal construction (even after auxiliaries). Although all three constructions can appear in paradigmatic relationships with prepositional phrases (e.g. in adjunct/focus positions), they can also appear in other environments where prepositional phrases are not found (e.g. after initial particles and after converters). In such cases, the adverbial analysis faces significant problems in accounting for this distribution. In contrast, the verbal analysis has the flexibility to deal with the overall observed distribution of the circumstantial sdjn(.f)/sdjn.n(.f) in Middle Egyptian and hence compares favourably. It would seem then that the circumstantial sdjn(.f)lsfjjn.n(-f) are indeed only circumstantially adverbial.
THE
CAREER
AND
THE
OF
KHNUMHOTEP
SO-CALLED
III;
"DECLINE
By Detlef
OF
OF
BENI
THE
HASAN
NOMARCHS"*
Franke
I. The so-called "Decline of the Nomarchs" In Egyptology, it is generally accepted that the title and office of Nomarch (provincial governor w i t h the title hrj-tp 3, literally "Great Overlord", of a n o m e ) disappears from our records in the time of Sesostris I I I . This feature is often described in egyptological studies with words like "elimination" or "breaking the power" , suggesting that the King has by a forceful, conscious action overcome the group of nomarchs, who had become too powerful and diminished the authority of the King. c
1
The idea seems to derive who wrote in his "Geschichte
from the work des Altertums":
of
EDUARD
MEYER,
"Nirgends [ist] in Aegypten ein Nomarchengrab oder auch nur ein Denkstein eines Gaufürsten erhalten, der jünger wäre als diese beiden Könige [Sesostris II. und III-1 > Das kann kaum Zufall sein; vielmehr drängt dieser Tatbestand zu der Annahme, daß unter Sesostris III. ... eine tiefgreifende Umwandlung durchgeführt oder wenigstens ... zum Abschluß gelangt und das Gaufürstentum beseitigt worden ist. ... So scheint die * This is an altered version of my paper for the Colloquium "The Residence and the Regions in the Middle Kingdom", held at Cambridge, 28-29 April 1988. Its second part on the interpretation of the finds from the sanctuary of Heqaib will be dealt with in my book "Studien zum Heiligtum des Heqaib auf Elephantine". My abbreviations are those of the Lexikon çter Ägyptologie (LÄ). 1
R.Delia, A Study of the Reign of Senwosret III. (PhD Columbia University 1980) 164ff.; I.Matzker, Die letzten Könige der 12. Dynastie (Europäische Hochschulschriften, Reihe III, Band 279, Frankfurt/Main-Bern-New York 1986) 11; W.C.Hayes, in: CAH , 1/2, 505f., cited by W.A.Ward, in: GM 71 (1984) 55f., W.C.Hayes, in: JNES 12 (1953) 31 ff.; H.G.Fischer, in: LÄ, II, 413ff. 3
51
Detlef Franke
52 Macht und Selbstherrlichkeit des Amenemhet III. vollständig gebrochen
Adels unter zu sein."
Sesostris
III.
und
2
The last scholar writing on this subject was EUGENE CRUZ-URIBE, in: Varia Aegyptiaca 3 (1987) 1 0 7 - 1 1 1 . I think, his description of the elimination of the nomarchs is basically right. His arguments are in favour of "a slow elimination of the office by not appointing successors to current holders of the title" (loc.cit., 1 0 8 f . ) . "As each nomarch died, the king would refuse to appoint a heir" (loc.cit., 1 1 1 ) . But he also maintains the common opinion, that "Sesostris III. ... began to replace all of the nomarchs with local mayors" {loc.cit., 108) . 3
But is this picture of a "revolution" from the King's correct ?
side
First of all, it seems to be wise to make a distinction between the disappearence of the title "nomarch" and the disappearence of great rock-tombs in Middle and Upper Egypt. Both traits can but need not be interrelated. The disappearence of a title could be due to changes in the mode of government that happen to occur in Egypt from time to time, earlier and later. The construction of a great tomb depends on wealth and access to royal and local resources and craftsmen. The w i s h and the possibility to build or not to build a great tomb are not dependent on a specific title but personal wealth and common élite funerary beliefs and custom. At the beginning of the 12th dynasty, Amenemhet I. and his son Sesostris I. appointed men in specific regions of Egypt, who are wearing the titles of "Mayor and Overseer of Priests (of the local tempel)" (ff3tjJmj-rS hmww-ntr, abbreviated as HA MHN) and "Great Overlord " (Hrj-tp 3 ~ HTA) of their nome. These nomarchs originate from local families, but their status and power was due to their apc
c
2
E. Meyer, Geschichte des Altertums, I ,2 (Stuttgart/Berlin 1909) §285, p.252 f. = I ,2 (1913) 276. I have to point out that there is much difference between a long-term socio-political process ("tiefgreifende Umwandlung") and a single, conscious action ("Beseitigung" "Macht brechen"). 2
3
3
I cannot agree with Cruz-Uribe's interpretation of the "Enseignement loyaliste" and the "Hymns in Honor of Sesostris III." in the Illahvm-papyri as literary works to justify the King's policy against the nomarchs. It is also impossible, to take the "Lamentations of Khakheperre-seneb" and the "Admonitions" as propaganda pieces of the nomarchs in response to their threatened situation, first of all, because none of these four literary works was originally written in the time of Sesostris III.
The career of Khnumhotep III
53
pointment by the King. Sesostris I. and h i s son A m e n e m h e t I I . supported the heredity of office and titles w i t h i n the same family, perhaps to ensure stability and c o n t i n u i t y in the g o v e r n m e n t . We have such nomarchs in Elephantine (nome 1 ) , H i e r a k o n polis ( 3 ) , Achmim ( 9 ) , Deir R i f e h ( 1 1 ) , Assiut ( 1 3 ) , M e i r (14), El-Bersheh (15) and Beni Hasan (16) ( f. H.G.Fischer, in: LA, II, 4 1 3 f . } . c
This is a very limited set of r e g i o n s . Actually, Egypt was no more divided into nomes, but the administrative r e g i onal areas were towns (n't) and the surrounding d i s t r i c t s (ww), governed by a HA MHN ("Mayor"). We have officials with the title "nomarch" from the b e g i n ning of the 12th dynasty and most of them seem to be recorded not later than the reign of Amenemhet I I . (nome 3 (Hierakonpolis), 9 (Achmim), 11 (Deir R i f e h ) , 13 ( A s s i u t ) , 14 (Meir), 16 (Beni H a s a n ) ) . C
The last officials with the title Hrj-tp S (+• nome) during the reigns of Sesostris I I . and I I I . :
lived
Elephantine : Sarenput II. (tomb 31) [Sesostris II./III.] Deir Rifeh: Nefer-Khnum (tomb I) and Nekht-anch (tomb VII) [Sesostris I./Amenemhet II./Sesostris II.?] Assiut: Djefai-Hapi I I . (tomb II) [Amenemhet II.] Meir: Ukh-hotep I I . (tomb B No.2) [Sesostris I.] El-Bersheh: Djehuti-hotep (tomb 2) [Sesostris III.] Beni Hasan: Amenemhet (tomb 2) [Sesostris I . ] . These are the only officials we can call " n o m a r c h s " . It is true that the last of them disappear from our records in the time of Sesostris I I I . But I think, no one seriously can call a group of about 6 men - though of h i g h rank and with local followers - a "powerful political group" that could be a threat to the might of the King. From other regions or nomes we do not have any comparable sources about nomarchs. Instead we have p e o p l e with the titles HA MHN only, who were probably the highest local administrators. Some of them were owners of d e c o rated rock-tombs like the n o m a r c h s . But to classify them under the heading 'feudal nomarchy' is certainly w r o n g . They were powerful and wealthy local administrators but nevertheless officials of the King. Some of these HA MHN built great rock-tombs, but most of them not, while other members of the élite, who were not HA MHN, have built rock-tombs of considerable size, too (e.g. Qubbet el-Hawa N o . 2 8 ) . The custom to construct such tombs came to the reign of Amenemhet III. at the latest:
an end
during
Detlef Franke
54
Elephantine; HA MHN Heqaib, son of Sat-Hathor (tomb No.30) [Sesostris III./Amenemhet III.] Oaw el-Kebir: HA MHN Wachka II. (tomb 1 8 ) [Sesostris III./Amenemhet III.] Assiut: HA MHN Djefai-Hapi III. and IV. (tomb VI and VII) a [Amenemhet II./Sesostris II.] Meir: HA MHN Ukh-hotep IV. (tomb C No.l) and HA MHN Khakheperre-seneb [Sesostris III./Amenemhet III.] El-Bersheh: HA MHN HTA Djehuti-hotep (tomb 2) [Sesostris III.] Beni Hasan: HA MHN Khnumhotep I I . (tomb 3) and his son HA Khnumhotep IV. (tomb 4, unfinished) [Sesostris II./III.] 4
6
Interesting enough, some of these last great decorated tombs show common features, part of them were classified under the rubric "usurpation of royal p r e r o g a t i v e s " : a distinctive symbolism connected with papyrus-plants and the heraldic plants of Upper and Lower Egypt (Heqaib at Qubbet el-Hawa No.30, and Ukh-hotep IV. at Meir, tomb C No.l) ; women performing agricultural tasks which are normally done by men {Ukh-hotep IV. at Meir and Wachka I I . at Qaw el-Kebir); processions of fecundity figures {Wachka II. at Qaw e l - K e b i r ) . 7
8
4
There is no proof that any of the so-called nomarchs of Qaw elKebir had the title HTA, cf. C.Liliquist, Ancient Egyptian Mirrors from the Earliest Times through the Middle Kingdom (MAS 27, München 1979) 139 n.1588. H.Steckeweh/G.Steindorf, Die Fürstengräber von Qaw (Leipzig 1936) 7(3), only mentioned that Wachka (I.) had the title hrj-tp S nj WSdjt without giving any further evidence. His tomb (No.7) dates to the middle part of dynasty XII, while his statue and stela (Torino Cat.No.Suppl.4265 and Cat.No.1547) are clearly posthumous, perhaps from the beginning of dynasty XIII. C
3
Djefai-Hapi III. more often has the title Jmj-rS htmt, and he is also Jmj-rS mä (P.Montet, in: Kemi 6 (1936) 131* f f . ) ; for tomb VII see: R.Moss, in: JEA 19 (1933) 33 = P.Montet, a.a.O., 134 f. c
6
Meir, VI; A.Kamal, in: ASAE 14 (1914) 74 ff. (not decorated (?)).
7
H.W.Müller, Die Felsengräber der Fürsten von Elephantine (AT 9, Glückstadt 1940) 93ff., Abb.46, Taf.XXXIX; Meir, VI, pl.XIII, XVI, and on his statues CCG 459 and Boston MFA 1973.87 (W.K.Simpson, in: Boston Museum Bulletin 72, no.368 (1974) 100 ff.; LÄ,VI, 821f.).
8
W.M.F.Petrie, Antaeopolis. The Tombs of Qau (BSAE 51, London 1930) pl.XXV, XXVIII; H.G.Fischer, in: Cl.Vandersleyen (ed.), Das Alte Ägypten (Propyläen Kunstgeschichte Band 17, Berlin 1985) 295, 302; J.Baines, Fecundity Figures (Warminster 1985) 238f.
The career of Khnumhotep III
55
Djehuti-hotep d e p i c t s in his tomb at E l - B e r s h e h the t r a n s port of his colossal statue, a scene of p u r i f i c a t i o n , and he and his father are wearing 5at-Amulets {El Bersheh, I, pis.XII, XV, X, X X X I I I ) . At three {or four) distinct places of Egypt we h a p p e n to find at about the same period a special and explicit symbolism of r e g e n e r a t i o n . This is a new trait in the decorum of tombs, which represent at the same time the end of the development of great rock-tombs of the 12th dynasty. I do not think that any of these features h a v e been a threat to royal p o w e r . I would like to connect t h e m with the changes in funerary beliefs during the reigns of Sesostris I I I . and Amenemhet I I I . that had i n f l u e n c e on the elite-custom of constructing great decorated tombs. I think, the existence of powerful administrators - called "nomarchs" - was limited to certain regions due to the specific history of these r e g i o n s . We happen to find them in the region called "the middle n o m e s " in M i d d l e E g y p t (from nome 9 to 16, cf. H.G.Fischer, Dendera, 6 5 f f . ) . In these regions there is a tradition of this office since the end of the Old Kingdom. At Elephantine, there w e r e n o HTA in the Old Kindom and First Intermediate Period, but there is a tradition of r o c k - t o m b s . Elephantine was the southern door-post of Egypt and important for the relations of the Egyptian state to N u b i a . Nomarchs like Djehuti-hotep of Bersheh were living fossils even during their life-time. And - as we know the dinosaurs died out, because they lived in an u n f a v o u r a b l e surrounding. The need for the King "Father's
'elimination' of the nomarchs - if there was any it! - perhaps was not that difficult as it s e e m s : just had to refuse to appoint the heirs to their place".
One way to facilitate this without force was the t r a d i tional practise to educate the n o m a r c h s ' male children at the King's residence. These children became loyal to the King and perhaps w e r e co-educated with the future King. They grew up to b e c o m e specialists in administration. Some of them were then a p pointed to local offices, others to jobs for the central administration.
56
Detlef Franke II. The case of Khnumhotep III, of Beni
Hasan
Perhaps the precedent and proof for this hypothesis is the case of Khnumhotep I I I . , the son of Khnumhotep II. of Beni Hasan, tomb N o . 3 . In the famous and much debated "scene of the A s i a t i c s " in Khnum-hotep 's tomb at Beni Hasan a group of so-called Asiatics ( 3mw, this means inhabitants of the eastern desert) are bringing gifts, especially black eye-paint, to Khnumhotep I I . They are led by two officials. c
9
On the writing-board presented by one of them towards Khnumhotep II. we read that they were brought by the "Sonof-the-Count" Khnumhotep in the 6th year of Sesostris I I . {BH, I, pl.XXXVIII,2) . This son Khnumhotep (III.) was the leader of the expedition, and he is depicted behind his father overlooking the whole scenery. He is labelled with the epitheton "whom the God (King) made (to someone)", pointing out, that he was educated at the King's court. This is stressed also by his mention in the long autobiographical inscription of Khnumhotep II. ( BH, I, p l . X X V I , lines 1 4 9 - 1 6 0 ) . 10
11
9
D.Redford, in: JARCE 23 (1987) 125ff., maintained that there is no proof for the assumption, that the people called "Asiatics" were living in the eastern desert. I think that the dwellers of the eastern desert were counted by the Egyptians in their mode of mythographical thinking under the rubric "Asiatic", people of the north-eastern regions outside Egypt proper. They do not come from the Sinai, but from the only region, where there was procured galena, the material for black eye-paint: Gebel Zeit. For Asiatics and the eastern desert see: R.Gundlach, in: SAK 8 (1980) 111; J.Osing, in: LA, II, 815f.; A.Saleh, in: BIFAO 81, Suppl. (1981) 107ff., B.G.Trigger et al., Ancient Egypt. A social history (Cambridge 1983) 142 n.l, p. 121; B.Couroyer, in: Revue Biblique 78 (1971) 558ff.; 80 (1973) 53ff.; and 81 (1974) 321ff.
1 0
P.E.Newberry, Beni Hasan I (London 1893) (cited here as BH, I ) , pl.XXX; B.Couroyer, in: Revue Biblique 81 (1974) 329f. See the stimulating article with some improbable readings and conclusions by H.Goedicke, in: JARCE 21 (1984) 203ff., and the interesting contribution by D.Kessler, in: SAK 14 (1987) 147ff. For the reading of the pertinent inscriptions cf. P.Vernus, in: BSEG 13 (1989) 173ff.
11
Urk.VII, 32. See also the remarks with a different interpretation on kjj wr and z3 wr shpr n t r by H.G.Fischer, Egyptian Studies I: Varia (New York 1976) 86.
The career of Khnumhotep III There his father wrote about him
57 (FIGURE 1 A ) :
9 2 "A great favour The other great
on the part of the King: one (i.e. Khnumhotep IIIJ was appointed to Sole Companion' (Smr w tj); 2 A big one among the courtiers, numerous of gifts (jnw) of the palace, a 'Sole Companion', there is no one of his kind: 2 Bearing to him are those, who are heard (judges: sdmjw), the sole mouth, closing (other) mouths. 3 One who brings useful things to their owner (i.e. the King), Door-post of the foreign countries: Neheris son Khnumhoteps son Khnumhotep, born of the Lady Khety." c
12
13
The role and status of Khnumhotep I I I . is described h e r e in some d e t a i l . He was a courtier with the rank of "Sole Companion" and has to deal with judges. But more i m p o r tant, he has to do with gifts and is bringing useful, p r e cious products for his Lord from foreign c o u n t r i e s . He is called "Door-post of the foreign lands". All other holders of this rare title have to do with the frontiers of Egypt and foreign people and products, which they had to deliver to the p a l a c e . 14
1 2
BH, I, pl.XXVI, 1.151 = Urk.VII, 32,9: ndw is nowhere attested as a title 'counselor' (as J.H.Breasted, Ancient Records I, §633, thought), perhaps the determinative of the "sitting man" is wrong (it should be: A.H.Gardiner, EG, Sign-List A2) . I would like to compare this phrase with stela London BM 574, 1.6 (HTBM, II, pi.8 = K.Sethe, Lesestücke, 75,7/8, also p.74,17); Urk.I, 216,1/2 (E. Edel, Altäg.Gramm., S1030 b)), Deir Rifeh, Tomb VII, 1.15. - For the last phrase see stela Boston MFA 29.1130, 1.3/4: "whose products inundated the two Lands, whose presents made the palace festive, who equipped the mansions of the one, who had advanced him..." (D.Dunham/J.M.A. Janssen, Second Cataract Forts I (Boston 1960) 59f., pis.9, 90).
1 3
Compare E.Blumenthal, Königtum, 380 (G 7.20 = CCG 20254, 1.1/2), and CCG 20485, 1.3; as role of the King: Pyr., §263 [248].
14
3 "door", meaning here "fortress, barrier". See;. H.G.Fischer, in: JNES 19 (1960) 261f. n.(a), and p.265 n.(o). Sarenput I. and Djehuti-hotep from Bersheh (El Bersheh, I, 16) were wearing this title, perhaps also Vachka II. from Qaw el-Kebir (H.Steckeweh/ G.Steindorff, Die Fürstengräber von Qaw, 49 No.7). On stela Boston MFA 29.1130 from Semna, we have "whom the King trusted in strengthening his monuments and in being a door-post for Egypt" (1.5: D.Dunham, op.cit.). The graffito from Kumma (D.Dunham, op.cit., 156) No.87, is perhaps the latest example for this title, datable to Dynasty XIII.
c
Detlef Franke
58
No other member of the Khnumhotep-tamily has the same titles and is described in this way. From this description of his role at the King's court follows that Khnumhotep III. was not a local administrator like his father or his brother Nekht, who was appointed "Ruler" (hqS) in nome 17 (Urk.VII, 31,6) by Sesostris I I . (his tomb is u n k n o w n ) . Perhaps he never got the possibility to become heir of his father in Menat-Chufu; he is never shown with his father's main title "Overseer of the eastern desert". * 1
A tomb of Khnumhotep I I I . is not known at Beni Hasan. Not he, but his half-brother, the son of Khnumhotep II. by his second wife, the female Treasurer ' Tjat, also named Khnumhotep (the fourth in the l i n e ) , seems to be the heir of Khnumhotep II. as administrator in the O r y x -nome. This Khnumhotep IV. owned the big unfinished tomb next to Khnumhotep I I . (tomb N o . 4 ) , w h e r e he is called "Noble" and "Count" (or "Mayor": H A ) . With him the line of administrators of the Oryx-nome and the eastern desert came to an end. 16
17
But if I am right in identifying of the Count Khnumhotep I I . and courtier and expedition-leader, traces of his life and career father.
1 3
" 17
Khnumhotep III., the son his first wife Kheti, as we can perhaps find some outside the tomb of his
The only inscription in the whole tomb, which seems to contradict this view, is that accompanying the bird-hunting scene (BH, I, pi. XXXII, left). There is said about a Khnumhotep, who is Khnumhotep III. or IV.: "Son of the Count of his body, to whom is given the heritage of the rulership, Khnumhotep, justified, in Menat-Chufu, as one gives his son as ruler of the 'Goddess-with-a-shell (? or basket ?)'-town." For the name of the town - combined with Hathor or Pachet? - see BH, III, pl.XXIX fig.91,' F.Gomaà, Die Besiedlung Ägyptens während des Mittleren Reiches. I. Oberägypten und das Fayyum (Beihefte TAV0 66/1, Wiesbaden 1986) 327f., is certainly wrong in his translation. For a different translation see Blumenthal, Königtum, 367 (G.6.68).- I am not sure if this really means that Khnumhotep III. was the heir of his father in full sense. W.A.Ward, in: GM 71 (1984) 56; BH, II, 12; LA,
I, 956.
BH, I, p.7; O.D.Berlev, Obscestvennye otnosenija v Egipte epochi srednego carstva (Moscow 1978) 205f.; W.A.Ward, in: GM 71 (1984) 51ff.
The career of Khnumhotep III
59
First I would like to m e n t i o n a stela, found by J. G.WILKINSON in the Wadi Gasus near the Red Sea, kept n o w in the G u l b e n k i a n Museum of Oriental Art at Durham, N o r t h umberland, with No.1935 (FIGURE I B ) . The text r e a d s : 1 8
14
"Year 1 (of King Sesostris
II.), establishing
of his monuments in the God's Land. (Rh-Njswt),
God's Sealer (and) true King's Confidant whom he (the King) loves in his heart, (and) who is in the heart of his Lord; Who knows the laws, (and) who is clever in acting, who follows the path of the one, who advanced him; Who not transgresses the order of the palace ( h) (and) the declarations of the court (stp-zS). The sole one of the (Lower Egyptian) King (Bjt), who was made in the palace ( h), (and) educated of Horus, Lord of the two Lands (the Who causes the Courtiers to ascend to the King, one truly precise like (the God) Thot: the Chamberlain Khnumhotep." c
c
King);
19
This stela represents an expedition-leader in the reign of Sesostris II., who worked in that remote a r e a . The inscription of this stela shows some similarities to the description of the career of Khnumhotep I I I . by h i s father in Beni Hasan tomb N o . 3 . The owner of the stela bears the normal titles of a leader of an expedition "God's sealer" and "Chamberlain", with the ranking title of "true King's confidant", which a c t u a l l y replaced the court-rank "Sole Companion" (Smr w tj) in the Middle K i n g d o m . c
20
1 8
A.Nibbi, in: JEA 62 (1976) 45ff., pl.X; A.M.Sayed, in: RdE (1977) 140ff., pi.8a (the text is written retrograde!).
29
1 9
s r here probably has the same meaning as st3w "to usher in" in other inscriptions, cf. Janssen, Autobiografie, II, Dh7; Blumenthal, Königtum, 382 (G 7.25/26); El Bersheh, II, pl.XXI, 1.9; Sinuhe B264.
c
2 0
Remember that Khnumhotep III. is called in the autobiographical inscription of his father's tomb only "Sole Companion". This was the lowest court rank and replaced in the MK by Rh-Njswt "King's Confidant", but there is no difference in rank between them (LA, V, 146f.; D.Franke, in: GM 53 (1982) 20). The title-sequence Rh-Njswt + Jmj-r3 hnwtj is found in several expedition-inscriptions. It should be mentioned that Khnumhotep II. too was Rh-Njswt and perhaps educated at the King's court, because his father Neheri was a Courtier, too (see 1.1, 68-71, 99ff. of his Autobiography). c
Detlef Franke
60
Further on he is described as one "who knows the laws" and "one truly precise like Thot". These epitheta could be compared with the mention of the judges above. Then there is the description "The sole one of the King, who was made in the palace, educated of Horus, Lord of the two Lands". So I would assume that this Khnumhotep is the very same Khnumhotep III. 21
22
He was educated as courtier and - perhaps due to his origin from the Oryx-nome and his father's office as an "Overseer of the eastern desert" - bestowed with the office of an expedition-leader. In Year 1 of Sesostris II. he conducted an expedition - presumably to the Sinai - via the port at Mersa Gawasis, in search for copper and malachite ; in year 6 of Sesostris II. he participated in an expedition to the Gebel Zeit, where the Egyptians extracted galena for black e y e - p a i n t . Then he led the expedition - perhaps accompanied by a group of native inhabitants of the deserts - through the Wadi Hammamat and passing Coptos back to the Nile valley, via his home-town Menat-Chufu and the local residence of his father, down the Nile to the king's residence. 23
24
Perhaps we can follow his line of career a step
further.
In 1894 J.DE MORGAN and his co-workers found north of the pyramid of Sesostris III. at Dahshur a brick-mastaba once belonging to a Khnumhotep (Mastaba No.2 in De Morgans 2 1
Janssen, Autobiografie,
I, Acl2,
2 2
Compare with the epithet of Khnumhotep III. shpr ntr (BH, I, pi. XXX, left); Blumenthal, Königtum, 286 (G.2.12~), 288 (G.2.19) and stela London BM 1213 (HTBM, III, pi.12, 1.7). The concept of education and training of the children of the élite was of course important for the continuity and stability of the political system.
2 3
A.M.Sayed, in: RdE 29 (1977) 147ff.; id., in: CdE 58 (1983) 28, 31f. On the stela the God Sopdu "Lord of the East" and "Lord of the Ssmt-land" is depicted, Ssmt is the mineral Malachite (green copper ore), found on the Sinai.
2 4
The Asiatics at Beni Hasan are presenting msdmt "black eye-paint" (galena = grey ore of lead), which is not found in the Sinai, but at Gebel Zeit, see: A.Lucas, Materials, 80ff.; G.Castel/G.Soukiassian, in: ASAE 70 (1984-85) 99ff.; id., in: BIFAO 85 (1985) 285 ff., 297f.; LÄ, I, 567ff., 826f.; Ill, 1166ff. - For "near-eastern" influence in the eastern desert compare the objects from a tomb of a hunter(?) found near Gebel Zeit (BIFAO 86 (1986) 382, pl.LXXI,C). The roads to the Red Sea will be surveyed in future and of course there were more than the Wadi Hammamat (cf. L.Bell et al., in: JNES 43 (1984) 27ff.).
The career of Khnumhotep III
61
n u m b e r i n g ) , and cased with limestone. T o m b - and s t o n e r o b bers had destroyed the tomb to a great e x t e n t , but part of the inscriptions from the limestone walls survived in the debris of the p i t . Some scattered fragments of i n s c r i p t i o n were p u b l i s h e d . But I think, enough is preserved to confirm GUSTAVE JÊQUIER's old idea, that the owner of the tomb b e l o n g e d to the Khnumhotep-iamily of Beni H a s a n and that he w a s in fact Khnumhotep III. 25
26
The owner of Mastaba N o . 2 had the h i g h e s t ranking titles, his office was that of "Chief Steward of the King's property" (Jmj-rS p r wr). He also bears the titles of ° a Vizier and a juridical e p i t h e t o n . On the front of the Mastaba there was mentioned the God Chnum, Lord of Her-wer, the only mention of this God outside Beni Hasan in the Middle K i n g d o m . Further on, Khnumhotep is described as one "to whom was given the gold of favour (by the King) in front of the nobles", a favour achieved sometimes by expedition-leaders as w e l l . 27
28
2 9
Unfortunately the fragments of the autobiographical inscription cannot be restored to a continous n a r r a t i o n . 30
2 5
26
J.De Morgan, Fouilles à Dahchour I (Vienna 1895) 18ff., general plan: p.16 fig.18; plan of the tomb: p.19 fig.20/21; Inscriptions: p.19 fig.23, p.20 fig.24, p.21 fig.25/26, p.22 fig.28. Compare also PM, III, 896. op.cit.,
23; W.K.Simpson, in: JEA 43 (1957) 27 n.2.
2 7
De Morgan, op.cit., 20 fig.24. For the same unusual sequence of the vizieral titles: W.M.F.Petrie, Memphis I (London 1909) pl.V, 1.10; De Morgan, op.cit., 31ff., fig.66/67 (Mastaba No.17); Wadi Hammamat M 113; Hatnub Graffiti 16 and 24. The juridical epitheton is "Master of the secrets of hearing alone".
2 8
This town was located on the westbank of the Nile opposite Beni Hasan/Speos Artemidos, cf. D.Kessler, Historische Topographie der Region zwischen Mallawi und Samalut (Beihefte TAVO 30, Wiesbaden 1981) 120 ff., 1 8 « . ; F.Gomaà, Die Besiedlung, I, 312ff.
»
e.g. Wadi Hammamat M 43, 1.6 [Amenemhet III., 2 ] ; cf. also the inscription of Inpy at El-Lahun, Fragment 62: H.G.Fischer, in: JEA 68 (1982) 49 n.(q) [Sesostris III./Amenemhet III.].
3 0
De Morgan, op.cit., 21 fig.26. The most important fragments are combined here, with some emendations, in FIGURE 2. I think we can identify two themes: some phrases belonging more to the religious sphere, others belonging to the (ideal-)biographical sphere.
Detlef Franke
62
We can identify the mention of an "order of the King to Ruler the title "Chief of troops of ..." and the mention "ships", a "captain" and the "steering-oar" (see FIGURE 2) .
the of
3 1
The phrase "overthrowing of the rebel, when he raises normal phrase of e x p e d i t i o n - i n s c r i p t i o n s .
himself"
is a
32
Some hints about the way of the troops are g i v e n : "...moving on a brook (or canal), seeking for the Land of the Asiatics , opening (?) (the Land of...?) (or: "...while [I] opened...") some of the activities took place at "the frontiers of the Bitter Lakes" perhaps a hint that they took the land way to the Sinai via the Wadi Tumilat and the Bitter L a k e s . 33
34
3 1
c
wd-Njswt n hqS ...(FIGURE 2, fragment 1 ) , Jmj-rS mä (fragment 11), dpwt (fragment 1 ) , nfw, hmw (fragment 2 ) .
n
...
3 2
shrt sbj hpr.f (fragment 5 ) , compare Graffito De Morgan, Cat. des Mon., I," 24 No. 165 = LD, II, 123d. In El Bersheh, II, pi. XIII, 1.8 = E.Brovarski, in: Fs Dunham, 18 n.(c), we have the exact parallel, but here in a ritual context. The determinative of the Hippopotamus or Pig is unusual, perhaps due to the "typhonic" quality of the enemy.
3 3
Fragment 7. The term "Land of the Asiatics" is also occuring in the just mentioned (n.29) Graffito from Wadi Hammamat M 43, 1.7: "...who overthrows the Nubians, opens the Land of the Asiatics, (1.8) and traverses all hill-countries ..." [Amenemhet III.,2], see: B.Couroyer, in: Revue Biblique 81 (1974) 332f. Another possibility would be to emendate the term into tS-ntr "the God's Land", which is written in nearly the same way on the above mentioned Stela Durham N.1935. But normally "God's Land" is written with the "God's sign(s)" in honorific transposition. For this term see: B.Couroyer, in: Revue Biblique 78 (1971) 59ff.; 80 (1973) 53ff., 274ff.; C.Kuentz, in: BIFAO 17 (1920) 121ff.
34
[ h3. ]n.f tSS.hr Km-wr (fragment 6 ) . For the emendation cf. Blumenthal, Königtum, 186f. (E 2.6ff.). For Km-wr see: Gauthier, DG, V, 201f.; Montet, Géographie, I, 216; P.Vernus, Athribis (BE 74, Le Caire 1978) 329, 344ff. The only locality near the eastern frontier of Egypt called with this term are the Bitter lakes (cf. Sinuhe B21; Merikare, 1.99; LA, I, 824f.). For a map of this region see G. Posener, in: CdE 26 (1938) 258; E.CM.Van den Brink, in: MDAIK 43 (1987) 7ff. A MK-graffito in this region: T.C.Townsend/ R.Engelbach, in: ASAE 33 (1933) 3, map pl. I; for a recent survey; C.A.Redmount, in: Newsletter ARCJ?No.l33 (Spring 1986) 19ff.
•\ ' j * I ; j ! ! !
:i I
< i
c
The career of Khnumhotep III
63
I think that Khnumhotep is describing here his curriculum vitae from an expedition leader to "Chief Steward" with the same rank and authority as a real V i z i e r . 35
To
summarize:
Khnumhotep III. was appointed Chamberlain by Sesostris II., after being educated at the r e s i d e n c e . He conducted expeditions to the mining a r e a s . Perhaps after the s u d d e n death of Sesostris II. after his 8th regnal year, Khnumhotep III. achieved the office of "Chief Steward" and built himself a tomb north of S e s o s t r i s ' I I I . pyramid at Dahshur. Because he was so strongly connected with the King and involved in residental matters, he - as the possible h e i r of his father's local office at Menat-Chufu - was p r a c t i cally excluded from his heritage. He was "'promoted' away to the residence" - to put it in the words of H . W i l l e m s . 36
Instead he got the highest court-ranks and t i t l e s . Khnumhotep III. did not get his father's office, but perhaps part of the wealth of his rich provincial f a m i l y . He did not invest it in his nome, w h e r e he was born, but as courtier he had to spent it at the residence, to build his tomb there for instance - not at Beni H a s a n . Perhaps the unfinished tomb of his half-brother Khnumhotep IV. at Beni Hasan is a hint that he simply had not e n o u g h money for that big tomb, because the wealth of the r e g i o n was accumulated in the residence. There are no signs for an elimination of Khnumhotep IV. by force.
If the career of Khnumhotep I I I . from a p r o v i n c i a l born boy to courtier and "Chief Steward" was not a single case, but was followed by other provincial families, this would make the elimination of the highest provincial families superfluous. The result would be their 'desiccation' not by throwing them out of their offices and rank, but by not appointing the heirs to their fathers offices - an action, which was fully legal on the King's side. The provincial élite would be transformed into members of the residence élite. 3 8
Khnumhotep presumably was not a leader of a military campaign - as R.Delia, The Reign of Senwosret III., 170-173, thought - but of an expedition to a mining country.
3 8
H.O.Willems, Chests of Life (Leiden 1988) 61f. When we discussed the matter, we realized that we had come to the same conslusions; it was the evening in Leiden I met Dr .M. Pollock, as far as I remember .
Detlef Franke
64
Of course this is only a description of the way it could have been. And I have to stress that Khnumhotep I I . never was a 'nomarch'. We do not have any idea of the m o t i vations and the way of decision-making that led to these results. Was it "business as u s u a l " or something extraordinary that the King did not appoint the heirs to their fathers office in the provinces? Was it some form of i n n o vation in egyptian policy? Was there any opposition against these administrative acts ? We do not know. 37
The potential heirs were fully integrated in the rows of courtiers and rewarded with the highest ranks at the residence. The King thus won power and wealth through the funds accumulated in the residence, not scattered in the provinces. And the loyal o f f i c i a l s , rewarded with rank, office and income by the King, would have had no reason for rebellion. The power and w e a l t h was concentrated in this way in one center: the m e m p h i t e area, where also the Kings and highest officials were buried. And it w a s concentrated in a few hands, too: the King and a s m a l l , selected group of officials. This b a l a n c e of power remained in force until the family of the Kings of the 12th Dynasty came to an end and in the first p a r t of the 13th D y n a s t y .
Of matches regions and the centres
course this is an risky h y p o t h e s i s , but I think it with the p i c t u r e of the relatively poor provincial with people who left not much to our knowledge, rich finds from the memphite a r e a or religious like Abydos, Thebes and Elephantine from the time 38
3 7
The main title of Khnumhotep II. was "Mayor in Menat-Khufu", but perhaps more important for his relation to the central administration was the title "Overseer of the Eastern desert". He restored the tombs of two of his predecessors in office at Beni Hasan (tomb No. 21 of his uncle, and No.23, cf. BH, II, pl.XXIIA, XXIV). There is only one other official with this title: The "Chief Treasurer" Meru in the time of Mentuhotep II. (Petrie, Season, pl.XV No.459). Apparently Khnumhotep II. was the last holder of this office. But there are some instances for the title "Overseer of (all/the Western) desert(s)": cf. W.A.Ward, Index of egyptian administrative and religious titles of the Middle Kingdom (Beirut 1982) Nos.290a-291a, 339-341 (all llth/early 12th Dynasty), the Chief Treasurer Ikher-Nofret is the latest example (CCG 20683, 1.6) [Sesostris III., year 19].
3 8
Near the royal pyramids at Lisht, Dahshur, Illahun and Hawara the highest officials of administration were buried: The Viziers, Overseers of the King's property, the Chief treasurers, and others. Most of the tombs of lower-rank people, which must have been buried there also, were destroyed long ago or were not excavated.
The career of Khnumhotep III
65
of Sesostris I I I . o n . Perhaps this accumulation of w e a l t h and power in the residental region was also one r e a s o n for the fact that from the time of Sesostris III. on the donations at these places from people of m i d d l e and l o w e r ranks (the numerous nhw nj n't, other military m e n , and employees of the different administrative d e p a r t m e n t s and workshops) from the residental r e g i o n are i n c r e a s i n g . c
But there is an obvious and clear gap or d e c l i n e in the quantity of finds of stelae or statues from p r o v i n c i a l towns like Edfu and even Elephantine from the t i m e of Sesostris I I I . down to the end of the 12th D y n a s t y . T h i s is an indication for the fact that something - we s t i l l do not know exactly, what - happened in egyptian s o c i e t y in that period. It was the result of a shift in the d i s t r i b u t i o n of wealth and reputation between residence and r e g i o n s : the higher social strata in the provinces were d e s c e n d i n g , the people of the residence and central administration (high and lower) were ascending. In the following 13th Dynasty, when there were w e a k e r (and probably poorer) K i n g s , the officials of the c o u r t could gain perhaps even more power and w e a l t h , as w e l l as a new growing stratum of provincial d i g n i t a r i e s , w h o o r i ginate from the town- and temple-administration or the army. From the beginning of Dynasty XIII on, w e have a r e markable increase in the number of donated stelae and statues from people of different social and regional origin.
Abstract: There was neither a powerful political group of nomarchs nor a conscious, forceful single action by King Sesostris III. to eliminate a group of nomarchs - as far as we can deduce from the sources. The so-called "Decline of the Nomarchs" was a socio-political process which resulted in the centralization of power and wealth at the King's residence and in the hands of the court élite. This is exemplified by following up the career of Khnumhotep III. of Beni Hasan, whose curriculum vitae from a provincial-born boy to courtier perhaps gives us an idea of how the King dealt with the provincial nobility in the middle part of the 12th Dynasty.
66
Detlef Franke 1*3 , •
•iSS f
IÏ
T
ü
i l
"if?
mm LT -*
A 4 k
I I I
K 111
• 11 +1
111
i
&
1 II
i?
JL. •
-
Xi LLJ
IF
M.
•A
• «
T •
FIGURE 1 A: Beni Hasan, tomb No.3 Beni Hasan, I, pl.XXVI, 1. 145 - 160
*ii r A i l : rr'H • r
FIGURE 1 B: Stela Durham N. 1935
3i• —
I8
+
The career of Khnumhotep III
67
FIGURE 2: Fragments from the Biography of Khnumhotep III. (adapted from: J. De Morgan, Fouilles à Dahchour, I, p.21, fig.26)
The Coffins of the Middle Kingdom: The Residence and the Regions
James K. Hoffmeier Wheaton College, Archaeology Department Wheaton, Illinois 60187-5593
The wooden coffins of the Middle Kingdom are one of the most celebrated artifacts of the period for their artistic value as well as for the significant corpus of funerary texts found inscribed or painted on them. Thus there has been considerable scholarly investigation of Middle Kingdom coffins over the years. Recently there have been a number of studies concerned with the distinguishing and dating of coffins of the 1st Intermediate Period and the Middle Kingdom.^ Before we examine relationship between the coffins of the residence and the regions for the Middle Kingdom, a short history of the development of the coffin and its decoration is necessary. Wooden coffins are first attested in Dynasty 2 at Saqqara.2 Similar coffins are found in Dynasty 3.^ These show the panelled pattern characteristic of Archaic period architecture. These earliest coffins have convex or arched lids with flat parapets on the ends. These two features of the lid must have been
regularly
included on early coffins and sarcophagi to have been reproduced so faithfully in the coffin determinative
(Gardiner's sign-list Q-6) in the word krs from the Old
through the Middle Kingdoms.
4
(We shall return later to the history of the arched lids
as they relate to the Middle Kingdom). The coffins of Dynaties 2-3 are anepigraphic. Decorative panels and/or the palace façade motif, with false door, are also found on sarcophagi and coffins of the Old Kingdom, both for royalty (i.e. that of Men5
kau-re ) and nobility.6 These motifs are regularly included on coffins of the Middle Kingdom. Funerary inscriptions appear in and on sarcophagi and coffins in Dynasties 567 The frieze d'objet scenes and offering lists that had adorned the walls of tombs in earlier centuries are found on the inner walls of coffins and on the underside of lids early in Dynasty 6.^ Dynasties 5-6 also witnesses the appearance in coffins of the polychrome 69
70
James K. Hoffmeier
matwork that had been painted in early tombs (sometimes associated with the palace façade/false door). The magical eyes that are a regular feature on the outsides coffins of 9
1st Intermediate Period and Middle Kindom are introduced at this time too. Thus we can see that before the Old Kingdom ends most of the standard features common to Middle Kingdom coffins are known, in some cases with prototypes that reach back to the Archaic period. The fact that these artistic features are continued into succeeding centuries makes the task of determining the date of coffins of the 1st Intermediate Period and Middle Kingdom, and distinguishing coffins from the centers of politcal power and the regions difficult indeed. The 1st Intermediate period is a formative one in the evolution of the coffin, with some innovations, but mostly experimentation with earlier motifs. The most important innovation of this period is the introduction of Pyramid and Coffin Texts to the coffins of non-royal individuals which is recognized to be a part of the "democratization" of the funerary cult in the aftermath of the Old Kingdom.
10
Perhaps owing to the limited
amount wall decoration and inscriptions in tombs of the 1st Intermediate Period, the interior of the coffin becomes more significant in the funerary cult, and thus is more elaborately decorated, including agricultural and domestic scenes, the false door, as well as portrayals of the decedent seated before the offering table.
11
Coffins of this type are
1
frequently called "Heracleopolitan" or "Middle Egyptian" -^ and originated in the 9th10th Dynasties. The exterior of coffins of the 1st Intermediate Period through the early part of Dynasty 12, by and large, contain little decoration (Figure 1). The magical eyes on the side of coffins are faithfully included throughout the period and beyond. At Bersheh a single wide band containing the htp ii nsw formula is painted or inscised around the upper part of the coffin, with another band running the entire length of the l i d .
13
This
pattern prevails at Gebelein (Figure 1), Gournah (CG 28022), Beni Hasan, Sedment, Abu 1
Sir and Saqqara. ^ At Assiut, there is more variation in the number and arrangement of 1
the columns containing inscriptions, as Donald Spanel's dissertation has shown. *' Added to the horizontal band are three or four vertical ones on a side and one to three may be found on the ends. A coffin of one Nbt.it(s) at Emory University Museum in Atlanta contains single bands (Figure 2), as does Mallawy Museum #569 belonging to a lady named hnnyt.
The coffin of it.f ib or tf.ib (FM 88917), on display at the Field Museum of
Natural History in Chicago, and a virtually carbon copy of Nakht (FM 881917) are decorated with a double band of inscriptions on the coffin while a triple band occurs on 1
the lid. ** Triple and quadrupal bands are also attested. An example of the latter is Mallawy #567, belonging to Hor-Hotep. In the Assiut group one can find combinations in
The coffins of the Middle Kingdom
71
the numbers of bands on a coffin; e.g. double vertical bands on the walls and triple on lid (Spanel's IIF - e.g. the Field Museum coffins), double on the sides and triple on lid and ends (Spanel's EH), triple columns on the coffin and quadrupal on the lid (Spanel's II I), etc. The introduction of the vertical collumns, which is known elsewhere during the latter part of the 1st Intermediate Period, (e.g. Bersheh), is an important development that laid the groundwork for placing the the niched panelling, palace façade and false doors on the sides and ends of coffins that began towards the end of the 11th Dynasty and becomes a regular feature of the Dynasty 12. E.L.B. Terrace^ described the coffins of Upper Egypt from the 1st Intermediate Period as being the product of "untutored provincial artists of those unsettled times" who "were experimenting with the forms of decoration in an effort to determine the the most suitable one (i.e. type)." These "unturored artists" produced cruder quality work and experimented with a variety of scenes on the outside of the coffin, and Gebelein, located just south of Thebes, is cited as an area where such experimentation took p l a c e d Cairo 28033 contains a scene depicting men doing laundry, while another example from Gebelein in the Cairo Museum shows a man leading his dog by a leash. Akhmim, situated between Thebes and Assiut, has yielded a number of coffins with peculiar idiosyncracies. A number of these coffins have the offering list on the outside^ (Figure 3). Other scenes, including the mummy of the deceased lying on a bier (CG 28015, 28016), are painted on the exterior. A development found at Akhmim, and attested elsewhere towards the close of the 1st Intermediate Period, is the addition of the false door under the magical eyes (CG 28015 & CG 28033). The combination of these features is a part of the canon of the Middle Kingdom. Diana Magee,20 in her investigation of an Akhmim coffin in the Ashmolean, draws attention to the inclusion of the palace-façade type false door at the north end of this coffin. A pair of wdßt eyes surmount the lintel, which is often included of Middle Kingdom coffins.
According to Henry Fischerei this detail occurs first during the
Herakleopolitan period at Saqqara. M a g e e ^ points to epigraphic and paléographie characteristics that she considers to be Theban as evidence for dating the Ashmolean coffin to just after the reunification under Montuhotep Nebhepetre.
In view of the
difficulty of dating the Akhmim coffins, her observations are significant. However, we must bear in mind H.O. Willems' recent caution that Theban influence north of the Diospolite nome could have begun prior to uniiication.23 The simplicity of the external decoration of Theban coffins from the latter part of the 1st Intermediate Period or the first half of Dynasty 11 can be seen in CG 28022 from Gurnah and that of Aashyt, wife or concubine of Montuhotep Nebhepetre (JE 47355). It
72
James K. Hoffmeier 2
has been observed that this coffin stands within the Theban provincial tradition, "* with some Memphite modifications which suggest that this burial equipment dates to just 2
after the reunification. "* The Middle Kingdom witnesses the transferal of some of the decoration that had, by and large, adorned the interior of coffins of the 1st Intermediate Period to the 2
exterior. E. Brovarski ^ has traced the evolution of external decorations of the Bersheh coffins from the DynastylO through the mid point of Dynasty 12 which proceed from 2
simple to complex. Although the addition of the vertical bands, Spanel ^ has shown, is 2
attested in the Assiut coffins earlier than at Bersheh. Hayes ** discerned a similar development, from simple to complex exterior decorations, in the coffins from Meir and Assiut. It is probably fair to say that this pattern is true throughout Egypt for the Middle Kingdom. Let us now turn to consider the coffin types for the Middle Kingdom. Mace and Winlock, in their publication of Senebtisi's funerary equipment, spoke of "Court and Provincial types" and three geographic / chronological classifications for the Middle Kingdom: 1) Heracleopolitan, 2) Upper Egyptian and 3) the Court style, the latter appearing with kingship of Amenemhet I which "dominated the royal cemetaries 2
throughout his dynasty." ^ For Mace and Winlock, and subsequently W.C. Hayes,30 the simply decorated coffins of Senebtisi, which have vaulted lids, were classic examples of the court type. (Figures 4-5) The palace façade that had appeared occasionally in coffins of the Herakleopolitan period becomes a regular feature of 12th Dynasty, "polychrome" coffins-^ Those coffins which had a single false door, surmounted by the w3d_t eyes, might have blank space between the remaining vertical bands on the front and back (these generally are from earlier in Dynasty 12), or they might be filled with panels or additional false doors.^
2
These latter two patterns continued throughout the 12th
Dynasty even after the false door ceased to be painted on the interior. The earlier decorative traditions continued in the provinces with some borrowing of the court type by nomarchs, which then filtered down to lesser officials in that region.^3 The foregoing classification scheme, advanceded by Mace and Winlock has 4
received considerable support over the years, especially by Hayes.^ Recently, however, Bruce Williams-^ has questioned the dating of the Senebtisi funerary materials which would effectively dismantle the classification scheme of Mace and Winlock. He notes that beyond the burial shaft of Senebtisi being in the Amenemhet II mortuary complex, there is no other evidence for a synchronism.^^ The Senebtisi coffins, according to Williams,^ show a remarkable similarity to the 13th Dynasty coffins of King Hor, Princess Ita, and others discovered by de Morgan at Dashur. However, Awibre Hor's
T h e coffins of the Middle K i n g d o m
73
coffin lid is flat. While recognizing that vaulted sarcophagi date back to the Williams argues that the "arched" lid coffin is a development of the 13th Dynasty and thus cannot 38
be the so called "court type" of the 12th D y n a s t y .
Williams reminds us that the
juxtaposition of Senebtisi's burial with that of Amenemhet I is not a sure criterion for dating since there are 13th Dynasty burials intruding upon, or being located adjacent to 12th Dynasty royal funerary structures.
39
If Williams is correct, then some of the major differences in coffin types between 4
the residence and the regions during the Middle Kingdom are eliminated. ^ The tomb of Imhotep at Lisht is situated within the pyramid complex of Senusert I, and has been dated to the 13th Dynasty.'*
1
However, the recent ceramic analysis of materials from
the Mastaba of Imhotep by Dorothea Arnold suggests an earlier date during the reign of Senusert I I .
42
While Williams gave careful consideration to the construction of the
coffins and canopic chest, as well as the staves and scepters, he failed to investigate the the pottery. Janine Bourriau's study of the pottery from Senebtisi's tomb supports a 12th 43
Dynasty date for the tomb. . The ceramic evidence, coupled with the evidence on the vaulted coffin lid presented here suggests that Senebtisi's burial should be placed in the 12th Dynasty. The popularity of the arched lids in the 13th Dynasty and beyond does not preclude the use of such lids in the Middle Kingdom. It is well recognized that the coffin is a microcosm of the archaic period mastabas, which in turn developed from domestic architecture.
44
4
The wooden coffins of the 2nd and 3rd Dynasties have curved lids. ^ For
Emery, the coffin evidence led him to reconstruct the arched tops with flat parapets on 4
the ends of certain mastabas of the archaic period. **
Cairo 67567 continues in the
tradition of the Archaic period into the 3rd Dynasty. The convex sarcophagus lid is also 4
attested in Dynasty 4 at Meydum, ^ as well as many examples from Giza now on display at the Cairo museum (e.g. 48078, 51950, 18220 with parapets, and 53149 where they are absent). The tradition continues in the 5th Dynasty at Abu Sir with the two sarcophagi 48
4
in the Mastaba of Ptahshepses, and at Saqqara in the 6th Dynasty. ^ It is difficult to believe that the arched coffin or sarcophagus lid, which had such an early and continuous history, from Dynasty 2-6, should suddenly drop out of use, only to reappear in the 13th Dynasty. After coming to these conclusions, I discovered C. Lilyquist's^O short article on the subject in which she raises similar questions. She points to stone sarcophagi and canopic chests from Lisht that have arched lids and asks, "Do we dismiss these because 1
they were made of stone rather than wood? I'm not sure we can do that."^
74
James K. Hoffmeier If the arched lid is found on the stone sarchophagi of the 12th Dynasty, then one
might to expect that there should be wooden coffins of the same variety for that period. The difference between a wooden outer coffin and a stone sarcophagus is only a matter of material, not function. The distinction between these terms, I believe is artificial. Some clarification in terminology is necessary. Perhaps the term "outer coffin" should be replaced by sarcophagus. 2
Hayes^ points to a gray granite sarcophagus from el Lisht (CG 28102), which has the four transverse support battens typical of wooden coffins, as evidence that this 12th Dynasty sarcaphagus was fashioned like the coffins of the period (e.g. BM 30482). A number of sarcophagi from Dashur discovered by de Morgan also were covered with the arched lids/>3 but whether these can be securely dated to Dynasty 12 is not certain. Aidon Dodson recently has made the ingenius suggestion that two of the alabaster sarcophagi found in the Senusert Ill's pyramid complex were in fact removed from Saqqara, likely from Djoser's pyramid galleries, and reused at Dashur.^
4
In the Teti
pyramid cemetary at Saqqara, a number sarcophagi with curved lids were found that appear to date to Dynasty 12.^5 There seems to be ample evidence for convex lids on sarcophagi of Dynasty 12 to believe that some wooden coffins of this design should date to the same period. In fact there are such wooden coffins that can be dated to the 12th Dynasty. 6
Lilyquist-* presents to the inner coffin and canopic chest of Hapi-ankhtifi at Meir as examples. A 12th Dynasty coffin from Beni Hassan, discovered by Garstang,^^ has a slightly curved lid, with parapets on the ends (Plate 1). At the International Congress of Egyptology in Cairo (November 1988), Ladislav Bares of the Czechoslovak Institute C
announced recent the discovery of the tomb priest Hrvy- nh at Abu Sir.58
He reports
that the lid was vaulted with "raised ends", and is quite convinced that this coffin dates to the 12th Dynasty when there was an attempt to revive the funerary cult of King C
Raneferef of Dynasty 5. Hwy- nh
appears to have been a priest assoicated with that
cult. In his extremely thorough study of Middle Kingdom coffins, Harco Willems lists four coffins with vaulted lids of Theban provenance which he dates to the late Middle Kingdom and Second Intermediate period.59 Presently, I am not aware of any coffins of the Heracleopolitan period that exhibit this type of lid. However, the words like krs/krst (Wb. V, 63-65), drw(t) (Wb. c
V, 601) and nb nh (Wb_. II, 228.14) in texts of the 1st Intermediate p e r i o d ,
60
still are
determined with the sarcophagus sign of the Old Kingdom that shows the vaulted lid
The coffins of the Middle Kingdom with parapets,
t—J
75
, although some variations exist. So indeed there are examples of
wooden coffins which exhibit the curved lid, with or without parapets in Dynasty 12, and I suspect that a thorough investigation of the coffin corpora, published and unpublished, would reveal more examples. The arched-lid coffins cited above from the Archaic Period through Old Kingdom are found in Saqqara and nearby necropoli which suggests that the origin of this design might be a northern one.
In the 1st Intermediate Period, the more simple
rectangular coffin predominated. Could it be that with the establishment of Lisht as the capital, by Amenemhet I, followed by the building of other adrrdnistrative centers in the north during the 12th Dynasty, that the older, northern tradition of the arched-lid sarcophagus and coffin made a come-back? Amenemhet I and his successors regularly reached back to the Old Kingdom for inspiration in art and architecture. Smith
62
61
W. Stevenson
noted that
Amenemhet brought the court within the range of old Memphite influences which still survived in the form of ancient monuments, and the effect is strongly evident in the art of the Twelth Dynasty. Further, if Dodson's suggestion concerning the reuse of sarcophagi from Djoser's burial complex is correct, and it does seem compelling, then we have direct evidence of the 12th Dynasty monarch's looking at earlier prototypes for inspiration. Then too, despite our limited knowledge of the 13th Dynasty, it seems, based on some of the surviving royal names, the continuity of the 12th Dynasty cities in the north, and retaining practices of the previous century, that the 13th dynasty rulers considered themselves to be legitimate successors of the Irj-towy Dynasty.
63
Thus we might expect
the arched coffin lid to be an earlier convention that they continued just as they followed other 12th Dynasty artistic traditions.
64
From the 11th to 12th Dynasties a clear evolution in the exterior decoration of 6
the coffin is evident, developing from simple to complex. ^ It is with this development in mind that one should examine Senebtisi's coffins. The outer coffin (sarcophagus) has a single horizontal band around the top and four vetical bands on each side and the rectangular panel with zv3dt eyes, while the inner coffin also displays the eyes, but has only a single band on the lid with a spell from the Pyramid Texts.
66
If the arched lid
was not considered, the outer coffin might pass for one Dynasties 9-11. The classical "black coffin" with vaulted lid of the 2nd Intermediate Period generally has decorations between the panels or more panels, with or without decorations (CG 28106, 3103). This development is similar to what we saw in the 12th Dynasty, viz. moving from simple to
76
James K. Hoffmeier
complex exterior decoration.
67
It is true that King Hor's queen, Neb-hotep's coffin is 6
similar in decoration to that of Senebtisi. ** But this might mean that the "court style" of Dynasty 12 was continued into the 13th Dynasty. Bruce Williams' study has certainly forced us to re-examine the so called "court style" of Dynasty 12. If he is correct in his assesstment of Senebtisi's date, then we need to redefine what is the "court style" of Dynasty 12. However, there is evidence for a near continuous history for the vaulted coffin lid from the Archaic Period through the Middle Kingdom. Admittedly, the evidence for the First Intermediate is scant (further research is needed), and, if there was a hiatus, it might be suggested that the 12th Dynasty, not the 13th marks its return. Willems sees the the appearance of this type of lid in the latter half of the Middle Kingdom, although, he suggests that the slightly vaulted 6
underside of some earlier flat lids might be a precursor to the arched lid. ^ A question that must be raised is what if any religious significance there may be to the vaulted lid, with or without the parapets. The arched shape lid might have been so designed symbolize the celestial vault which would be closely associated with the sky goddess, Nut. Just as Nut or the sky is over the earth, the micrcosm of the coffin illustrates this theology. Nut's association with the coffin is well established and can be traced back to the Old Kingdom Pyramid Text's. PT §616 reads: Nephthys has collected all your members for you in this her name of 'Seshat, Lady of Builders'. <She> has made them hale for you, you having been given to your mother Nut in her name of 'Sarcophagus', she has embraced you in her name of 'Coffin', and you have been brought to her in her name of 'Tomb'. 7 0
Commenting on this line, Faulkner observed that the lid is being a "simulacrum of the 7
sky" which was identified with Nut. *
There is ample evidence for connecting Nut
with the coffin lid, and this is widely recognized/
2
Some flat lids are decorated with
stars and constellations on the underside (as in not a few New Kingdom tombs); e.g. Mallawy 567, Horhotep; Aashyt, wife of Montuhotep Nebhepetre.^ The figure of Nut is found on the underside of sarcophagi of the New Kingdom onwards, and found as late as 74
Roman times.
Beginning in the New Kingdom, one can find illustrations on the inside
bottom of the coffin. The association of the coffin lid with Nut may be in mind when the PTs speak of the sky-goddess being "spread over" the deceased (PT §§580, 637, 777, 825, andl607). This 7
expression is found in CT VI, 118 and 264. *> The idea of Nut protecting the dead may have developed from this expression (PT §§825, 827-28,1629 and 1896; and CT VII, 3, VII, 9), and thus she could be called the "Great Protectress" (PT§§ 827, 834,838, 842, and 1607). The evidence from the PTs and CTs make it clear that Nut is associated with the coffin
T h e coffins of the Middle K i n g d o m
77
lid which may explain the vaulted shape of coffins and sarcophagi that enjoy a long history from the Archaic period all the way down to Roman times. Since there is an obvious connection between the sky-goddess and the coffin lid, one might wonder why more lids were not arched to give the appearance of the celestial vault. The writing of Nut's name is regularly determined by the sky sign,
1
J
. And
this factor may well be behind the more common flat lid. In fact, the construction of some Middle Kingdom coffin lids reveal that the outline of the pf -sign when viewed from the end. This phenomenon is found in some of the coffin lids found by Garstang at Beni Hasan; 6
e.g. E70 1903 in the Fitzwilliam Museum, Cambridge/ and another Beni Hasan coffin from tomb 3 (Plate 2). This type of lid is considerably more difficult for a carpenter to make than the simple flat lid, which suggests that there is something significant in the form itself. Also at the Fitzwilliam, the coffin of Lady Nakht from tomb 23 at Beni Hasan (E68 1903), has the pf-like lid, and inscription on the lid of the coffin which, as if to confirm suggestion offered here, says, "Oh Osiris, Lady Nakht, Nut is spread over you." This same expression is also found on the anthropoid cover of Userhet in the 7
Fitzwilliam/' as well as on the lid of the outer coffin of Userhet, now in L i v e r p o o l /
8
The connection between the sky and the coffin lid may be in view in the "Tale of Sinuhe" when the prospect of being buried in Egypt is dangled before the Egyptian exile to lure him to return from Retenu. Sinuhe B 192-93 states: A funeral procession is made for you on the day of interment, a mummy case of gold, a head of lapis, and the sky (pt ) above you as you are placed on the sledge, oxen drag you..7 9
While only few examples of the pf-like coffin lid, can be cited now, it be that with further investigation additional examples will come to light. In conclusion,
it
appears, then, that the vaulted coffin and sarcophagus lid, as well as the flat variety, have iconographie features which tie them directly to the sky-goddess Nut. There are a few other new features of 12th Dynasty coffins that should be mentioned. The "Book of Two Ways," with its guide map on the bottom of some coffins, appears early in Dynasty 12. In 1972 L e s k o
80
noted that all of the 18 published and 4
unpublished examples with the "Book of Two Ways" were from El-Bersheh.
81
Even if a
few occurrences of "The Book" do come to light from other areas, the concentration at Bersheh suggests that this indeed is a peculiarity of the Hare nome where it likely originated. Another artistic feature that occurs on some 12th Dynasty coffins is a polychrome 8
cavetto cornice bordering the top or actually on the side of the lid. Brovarski ^ has noted that this cornice occurs on some coffins at Bersheh (e.g. CG 28099), originating around the
78
James K. Hoffmeier
time of Senusert III. Similar decorations are painted on coffins from Meir, Assiut and Rifeh
Might this cornice be another Upper Egyptian innovation? At the outset of this paper, it was noted that much of the recent scholarly
investigation of coffins has concentrated on distiguishing those of the 1st Intermediate 4
Period from those of the Middle Kingdom, especially Dynasty 11.** The same rigorous analysis of 12th Dynasty coffins is now needed, especially in light of Williams' questions about the so called "court style." Brovarski's study is a good start, but it utilizes far to few examples to draw firm conclusions. The foregoing survey, has raised more questions that it has answered. In conclusion some questions need to be posed for our discussion and further investigation. 1)
What is the "court style" of Dynasty 12?
By when is this standardization
established? 2) Should we continue to use the terms sarcophagus and coffin since they are so similar in form, decoration, and since outer wooden outer coffins function as a sarcophagus? 3) Why does the Hare nome, even in Dynasties 9-11, retain such a high quality of workmanship in coffin decoration when other areas lapse into cruder forms? And related to this, how do we explain the apparent origin of the "Book of the Two Ways" at Bersheh? 4)
Can we detect any Theban influence on coffins of the memphite region after
Amenemhet's move to Ijt-towy?
E.g. Elke Blumenthal, Die Datierung des Nhrj Graffiti von Hatnub," Altoreintalistiche Forschungen 4 (1976) 35-62. W. Schenkel, Frühmittelä gyp tische Studien (Bonn 1962). "Repères Chronologiques de l'histoire Redactionalle des Coffin Texts," (Referat, vorgetragen auf dem 29. Internationalen Orientalistenkongreß, Paris, 1973) 29-36. "Repères chronologiques de l'histoire rédactionelle des Coffin Textes, 98103," in Actes XXIXe Congres des Oreintalistes. Egyptologie, ed. G. Posener, (Paris: L'Asiatheque, 1975) 98-103. "Zur Redaktions- und Überlieferungsgeschichte des Spruchs 335 a der Sargtexte," Götttinger Totenbuchstudien Beiträge zum 17. Kapitel. ed. W. Westendorf (Wisbaden: Otto Harrassowitz, 1975) 37-79. E. Brovarski, "The Bersheh Nomarchs of the Intermediate Period and the Middle Kingdom," in W.K. Simpson and W.M. Davies (eds.) Studies in Ancient Egypt, the Aegean, and the Sudan. (Boston: Museum of Fine Arts, 1981) 14-30. Donald Spanel, Beni Hasan in the Herakleopolitan Period, (Toronto: Ph.D. Dissertation, 1985). H.O. Willems, The Nomarchs of the Hare Nome and Early Middle Kingdom History," Taarbericht Ex Oriente Lux 28 (1985) 80-102. W.B. Emery, Archaic Egypt (Baltimore: Pelican, 1961) plates 24-25, Figure 77. W.C. Hayes, The Scepter of Egypt I, (Cambridge, Mass.: Harvard University Press, 1953) Figure 30. Wb V, 63. For the Middle Kingdom see Beni Hasan I (London: Kegan Paul, 1893), plates VIL XII & XV.
The coffins of the Middle Kingdom
79
^ Cyril Aldred, Egypt to the End of the Old Kingdom (LondonrThames &: Hudson, 1965) 96. S. Hassan, The Mastaba's of Hemet-Rc_ and Others [Excavations at Saqqara 1937-38. Vol. Ill], (Cairo: Government Printing Offices, 1975) plates XLII-XLIII. W.M.F. Petrie, Deshesheh (London: EEF, 1897) plate XXVII. The Mastabas of Hemet-Rg- and Others. Plates IV, pp. 7-8; Figure 31, pp. 57-58. S. Hassan, The Mastaba of Nv-cankh-Pepy and Others [Excavations at Saqqara 193738, Vol. II] (Cairo: Government Printing Press, 1975), Figure 9. 9 J. Garstang, The Burial Customs of Ancient Egypt (London: Constable & Co., 1907) 166. Deshesheh, plate XXVII. E.L.B. Terrace, Egyptian Paintings of the Middle Kingdom (London: George Allen & Unwin Ltd., 1968) 32. ^ P. Barguet, Textes Des Sarcophages Egyptiens Du Moyen Empire (Paris: Editions Du Cerf, 1986) 9-10. L. Lesko, The Ancient Egyptian Book of the Two Ways (Berkeley: University of California Press, 1972) 5-6. Ü Terrace, Egyptian Paintings..., 32, 34. 12 Senebtisi 50-54. Brovarski, "The Bersheh Nomarchs...," 23, Scepter I, 315. 1 Terrace, 156; Brovarski, 23 & Figure 13; cf. the coffin of Aha-Nakht in Philadelphia in S. Fleming, et. al.. The Egyptian Mummy. Secrets and Science (Philadelphia, The University of Pennsylvanian, 1980) #13. Spanel, 181. Spanel, 181-182, illustrations IA-E and IIA-M. 16 A word of thanks to Mr. Frank Yurco of the staff of the Field Museum for helping me with my inquiries and assisting me with records and photography. 1 Terrace 32. Terrace 33. 19 p. Lacau, Sarcophages Antérieurs au Nouvel Empire [Catalogue Général des Antiquités de L'Égyptiennes du Musée du Caire, Nos. 28001-28086, Vol. 11 & 14], Cairo, 1903) 28001, 28004, 28005, 28006, 28008 & 28013, plates MIL Diana Magee, "An Early Middle Kingdom Coffin from Akhmim in the Ashmolean Museum (No.l911.477)," TSSEA XIII-No.4 (1983) 241-248. 21 H.G. Fischer, "Inscriptions from the Coptite Nome," Analecta Orientalia 40 (1964) 40. Magee, 248. Willems 93-101. H.G. Fischer, "An Example of Memphite Influence in a Stela of the Eleventh Dynasty," Artibus Asiae 22 (1959) 240-252. Spanel 188. Spanel 188-189. 26 Brovarski, p. 23; Fig. 13 on p. 29. Spanel 194. Scepter 1.318. ^ A.C. Mace & H.E. Winlock, The Tomb of Senebtisi at Lisht (New York: Metropolitan Museum of Art, 1916) 54-55. Scepter 1.317. Senebtisi 55. Scepter I, 315. The coffins of Nakht and Userhet in the Fitzwilliam exhibit illustrate the pattern of the latter; cf. J. Bourriau Pharaohs and Mortals (Cambridge: The University Press, 1988) Nos. 70-71, pp. 85 & 91). Senebtisi 55. 6
7
8
3
1 4
1 5
7
1 8
2 0
2 2
2 3
2 4
2 5
2 7
2 8
2
3 0
3 1
3 2
3 3
80 3 4
James K. Hoffmeier
Scepter 1,303-320. 3^ Bruce Wiliams, "The Date of Senebtisi at Lisht and the Chronology of Major Groups and Deposites of the Middle Kingdom," Serapis 3 (1975-1976) 41-55. Williams 41-44. ^ Wiliams 45. Williams 47-49. Williams 43-44. Williams 49. Sally B. Johnson, "Two Wooden Statues From List: Do They Represent Sesostris I?" TARCE XVII (1980) 11-20. Dr. Dorothea Arnold, in her presentation at the colloquium, entitled "The Mastaba of Imhotep at Lisht South" argued for at date during Senusert II's reign. Her dating by ceramic measurement is most compelling, and would consquently push back the date of wooden statues, which Sally Johnson has dated to Dynasty 13, squarely in Dynasty 12. Dr. Dorothea Arnold, in her presentation at the colloquium, entitled "The Mastaba of Imhotep at Lisht South" argued for at date during Senusert II's reign. Dr. Dorothea Arnold, in the comments that followed my presentation, pointed out that Janine Bourriau is studying of the ceramics and other materials in Senebtisi's tomb for publication and has concluded that the tomb dates to Dynasty 12. She was kind enough to confirm in writing (22 August, 1990) her view which she mentioned at the colloquiium at Cambridge. We look forward to seeing the published results of this study, since the coffins of Senebtisi have become the lynch-pin in the debate over the "court style" of Dynasty 12. W. Emery, Archaic Egypt (Baltimore: Pelican, 1961) 130-131. E.L.B. Terrace, Egyptian Paintings of the Middle Kingdom (London: George Allen & Unwin Ltd., 1968) 32. W.C. Hayes, The Scepter of Egypt I. (Cambridge, Mass.: Harvard University Press, 1953) 312-313. Archaic Egypt. Plate 24-25. Figure 77. Scepter of Egypt I, Figure 30. Archaic Egypt 131. ^ W.M.F. Pétrie, E. Mackay & G. Wainwright, Meydum and Memphis III. (London: British School of Archaeology in Egypt, 1910), pl. X, 4; p. 4. Zdenëk Uherek & Miroslav Verner, Preliminary Report on the Czechoslovak Excavations in the Mastaba of Ptahshepses at Abu Sir (Prague: Charles University Press, 1976, does not show a picture of these sarcophagi. But one was able to view these granite sarcophagi in November 1988. S. Hassan, The Mastabas of Hemet-Rc. and Others [Excavations at Saqqara 1937-1938, Vol. Ill], (Cairo: Government Printing Office, 1975). 50 Christine Lilyquist, "A Note on the Date of Senebtisi' and Other Middle Kingdom Groups," Serapis V (1979) 27-28. Lilyquist, 27. The Scepter of Egypt 1.319-320. J. de Morgan, Fouilles A Dahchour 1894-95 (Vienna: Adolphe Holzhausen, 1903, 74-76 = Cairo Cat. 28102. Aidon Dodson, "Egypt's First Antiquarians?" Antiquity 62, Nu. 236 (September 1988) 513-517. 55 C.M. Firth & B. Gunn, Excavations at Saqqara: Teti Pyramid Cemetaries I. (Cairo: IFAO, 1926). 56 Lilyquist, 27. The Canopic box of Hap-Ankhtifi is depicted in Scepter I, fig. 209. 3 6
3
3 8
3 9
4 0
4 1
4 2
4 3
4 4
4 5
4 6
4
4 8
4 9
5 1
5 2
5 3
5 4
The coffins of the Middle Kingdom 5 7
58
59 6 0
61 6 2
6 3
6 4
65
6 6
6 7
68 69 7 0
7 1
7 2
7 3
7 4
7
81
J. Garstang, Burial Customs of Ancient Egypt. Fig. 189. In May 1988 I had the opportunity to examine the Gartang archives at the School of Archaeology and Oriental Studies at Liverpool. The photographs (B. 209-210) make it very clear that the lid is arched and the parapets are very obvious. My thanks to Professor A.F. Shore and his staff for assistance in my research, and for providing me with photographs of this coffin, and for the permission to use this photograph (B 210) and that in Plate 2 (B 213) in this publication. I am indebted to L. Bares for sending me a copy of the paper he presented, along with slides of the coffin in situ, which will be published in a forthcoming issue of ZÄS. See also the Abstracts of the Fifth Internatinal Congress of Egyptology, p. 13. Harco Willems, Chests of Life (Leiden: Mededelingen en Verhandelingen van het Vooraziatisch-Egyptisch "Ex Oriente Lux" XXV, 1988), 172, n. 178, and 116-117 An example of this writing is found on the Stela of Pepi Seneb (Dynasties 8-9) from Nag ed Deir which was recetnly displayed in an exhibition at Berkeley Univeristy. W.K. Simpson, The Ancient Neart East: A History. (New York: Harcourt Brace Jovanovisch, Inc., 1971) 245-246. W.S. Smith, The Art and Architecture of Ancient Egypt (Revised edition with W.K. Simpson - New York: Penguin, 1981 ) 168. A.H. Gardiner, Egypt of the Pharaohs (Oxford: The University Press, 1961) 147-154. W.C. Hayes, "Egypt: From the Death of Amenemes HI to Seqenenre II," in Cambridge Ancient History II, part 1, (Cambridge: The University Press), 44-48. Stevenson Smith. The Art and Architecture of Ancient Egypt. 216-217. E . Brovarski, "The Bersheh Nomarchs of the Intermediate Period and the Middle Kingdom", in Studies in Ancient Egypt, The Aegean, and the Sudan, eds. W.K. Simpson & W-M. Davis, (Boston: Museum of Fine Arts, 1981) p. 29. Senebtisi, outer - Fig.l & 9, inner plate XVII a-b & Fig. 15 Brovarski, op.. cü-, 29. cf. Dahchour I, plate XXXVI. Willems, Chests of Life, 171-172. R. O. Faulkner, The Ancient Egyptian Pyramid Texts (Oxford: Clarendon, 1969) 119. Loc. Cit. n. 7. W. C Hayes, Royal Sarcophagi of the XVIII Dynasty (Princeton: The University Press, 1935) 67. A.J. Spencer, Death in Ancient Egypt (Middlesex: Penguin, 1982), 16567. H.E. Winlock, "Egyptian Expedition for MCMXX-MCMXXI, III. Excavations at Thebes", BMMA H, November 1921,49 fig. 24. e.g. that of Thutmose III, cf. Hayes, Sarcophagi, plate viii; Spencer, op_. cit., plate 30. In the 3rd Intermediate period, her figure is found on the bottom floor of the coffin, a development which is puzzling.
5 A variation of garbled version of this spell occurs in CT I, 312b-c. Willems, op. cit. 45, n.l, ties this Nut-spell with the coffin lid. My thanks to Dr. Janine Bourriou for allowing me to study and photograph these coffins during my visit in April 1988. The outline of the pf-sign can be seen in the phtograph of this coffin in Janine Bourriau, Pharaohs and Mortals. #68, p. 88. Garstang, Burial Customs, 175 and 191,1.4; J. Bourriou, Pharaohs and Mortals 92. Garstang. Burial Customs. 191,1.3.
7 6
7 7
7 8
82
James K. Hoffmeier
79 Lichtheim, Ancient Egyptian Literature I (Berkeley: University of California Press, 1975) 229, and W.K. Simpsom, The Literature of Ancient Egypt (New Haven: Yale University Press, 1972) 68 render pt as "sky". I am not sure if they take this to mean the canopy of the sledge as Wilson does in Ancient Near Eastern Texts Relating to the Old Testament (Princeton: The University Press, 1969 - 3rd edition), 21 n. 29. Willems, Chests of Life, p. 45 understands pt as the coffin lid. L. Lesko The Ancient Egyptian Book of the Two Ways (Berkeley: University of California Press, 1972) 3. 81 Prof. Jac. J. Janssen, in the dialogue that followed the presentation of this paper stated that the de Buck archive at Leiden contains no examples of the "the Book of the Two Ways" from anywhere but Bersheh. 82 Brovarski, "The Bersheh Nomarch...", 25. Scepter I, Fig. 204 & 207; W.M.F. Petrie, Gizeh and Rifeh (London: BSAW, 1907) Plate X. idem . note #1. 8 0
8 3
8 4
The coffins of the Middle Kingdom
James K. Hoff meier's "The Coffins of the Middle Kingdom: The Residence and the Regions" (Facsimile drawings, not to scale)
(Figure 1) Turin Supplement 13.268, after Brovarski, Hughes Festschrift 9th Dynasty (?), Gebelein.
(Figure 2) Emory University Museum 1921.2, coffin of Nbt it .(s) Assiut, late 11th Dynasty.
i
James K. Hoffmeier
(Figure 3) Ashmolean No. 1911.477, Akhmim, mid-llth Dynasty
•///////////////////////////////////////////////
(Figure 4) Senebtisi's outer coffin, after Mace & Winclock
/ / / / / / / / / / /
/ / /
!
The coffins of the Middle Kingdom
(Figure 5) Senebtisi inner coffin, after Mace and Winlock
86
James K. Hoffmeier
PLATE 1 (Courtesy of the School of Archaeology & Oriental Studies, Liverpool, B210)
PLATE 2 (Courtesy of the School of Archaeology & Oriental Studies, Liverpool, B213)
Non-Egyptians
recorded on Middle-Kingdom in
Steiae
Rio de Janeiro K.A. Kitchen
-
Liverpool
Introduction In 1826, the National Museum in Rio de Janeiro acquired an entire mixed
collection
of
Egyptian
antiquities,
reputedly
from
the
prior
excavations of Belzoni. These included 51 inscribed stone stelae and one wooden one, according to the 1844 inventory (the earliest known), plus three blanks.
All of these, with one query, can still be accounted-for
today, a century and a half later. Out of that series, 22 or nearly half are from the Middle Kingdom. One [Inv. 627 = 2419] is dated by cartouche to Sesostris III, and a second [Inv. 645 = 2435] also by cartouche to Amenemhat IV (not
III, as cited in some
earlier
monument
private c
literature). hands
Ha ankhef,
Of the rest,
(and
seemingly
the famous
one 'contentious' so
father-in-law
since
about
1907)
of Neferhotep
still
in
belongs
to
I and Sobekhotep c
"IV" (really III) of the 13th Dynasty; it alone names H a a n k h e f s as well as his mother.
1
father
Until the present writer was granted access to it
in September 1987, no Egyptologist had seen it for 60 years.
Most of the
stelae are probably of the 13th Dynasty, perhaps occasionally of the late 12th.
Possible
genealogical
links (for Inv. 646 = 2436)
and stylistic
peculiarities (Inv. 640 = 2430) date some clearly to the 13th Dynasty, and to the period of Sobekhotep "III" and "IV" (II and III); for the rest, only a general date can be surmised.
c
Excluding the Ha ankhef
stela
(from
Heliopolis?), most appear to come from Abydos, and especially from the North Cemetery area behind the old 'metropolitan' temple of Osiris, on the 'Terrace of the Great God', to borrow the title of W. K. Simpson's wellknown
monograph. As for publication, the entire corpus (plus the rest of the Collection)
is included in a comprehensive new Catalogue.
2
iSee M. Dewachter, RdÉ 35(1984), 198-199, figs. 3,4. M . Belträo, K.A. Kitchen, Catâlogo da Coleçâo do Antigo Egito existente
2
no Museu Nacional, Rio de Janeiro - Catalogue of the Egyptian Collection in the National Museum, Rio de Janeiro, 1989. 87
88
K.A. Kitchen
Foreigners
on the MK Stelae
3
c
On Inv. 627 [2419), bottom register, is the 3m Asiatic Gebgeb, bora of 'Ummi'. perhaps meaning 'the lame', of gb','tall',
4
Gbgb,
ir.n Imi, 'the
His name is almost certainly Egyptian,
and not Semitic (only a reduplicated form
might be compared), but his mother's name may very
well
be 'Ummi, 'mother' (hypocoristicon for 'Ummi-DN, 'the goddess X is my mother'?), although it should be noted that 'Imi personal
name.
c
5
nfr, 3m,
ir.n
mwt.f,
born
his
mother'.
of
is also a common MK
In the bottom lines of this same stela occurs s3.f 'his [stela owner's] son, Heru-nefer the With
a
purely
Egyptian
name,
Hrw-
Asiatic, Herunefer
('Happyday') seems to be the lesser son of the stela-owner Iunefer by an 'Asiatic' secondary
wife
or concubine,
to deserve
On the selfsame stela (end, at bottom), Skr
his epithet
'Asiatic'.
the parent of Reri is not
V
likely to have been a West-Semitic Sukru, quite different Sfcr-names k).
from the
clear,
family
indicating
unambiguous
examples
of
West-Semitic
in Middle Egyptian (using the vertical sjk-sign, not just s +
Also a Reri son of Skr js
same
(i) because the orthography is
group),
a possible
where
otherwise attested (Cairo CGC 20296, of this Skr
nisbe-typ&
has
the
Sokar-barque
determinative,
name Sokari ('he of Sokar'), cf.
later
Sety - and certainly not a Semitic name, therefore. Stela Rio Inv. 680 (No.21 in the new Catalogue,
pl.45), belonging to a
foreman(?) Kara, has in the bottom register two people of interest (both c
are named, but only one depicted). Most explicit is 3m, Twti,
'the Asiatic and chief of craftsmen , Twti'.
the West-Semitic root DWD, the
name
Dodi-(h)uatu
(Eg.
imy-r
hmwwt,
The name comes from
attested already in our period (c.1730 BC) in twt-(h)w3t),
No. 18 of the foreigners
Papyrus Brooklyn 35.1446, edited by W.C. Hayes long since. personal name is to be taken as Dawdi, Hebrew
Dodi,
Dodo,
etc.,
and
directly
comparable
early-2nd-millennium
in
Our Rio with
Amorite
later
Dawdi,
Dawdaya, etc., as attested in the Mari cuneiform archives for example 3
See my paper, "Early Canaanites in Rio de Janeiro and a 'Corrupt' Ramesside Land Sale", in S. Groll (ed.), Studies in honour of Dr. Miriam Lichtheim, Jerusalem, 1990. Cf. Wb., V, 165:9-10. Cf. Ranke, Aeg. Personennamen,!, 25:17. 4
5
Non-Egyptians on Stelae in Rio de Janeiro
89
our man is a "David" of patriarchal times, nearly the psalmist-king. ir.n
c
Ibi, 'the chief of craftsmen,
fruitful
a millennum
Also named (with figure) is the imy-r
source of West-Semitic
before c
hmwwt,
pr
c
Epir bom of 'Abiya'.
Again, pr
personal names in the Middle
,
is a
and
Late
Bronze ages (i.e., MK and NK); so, in the Brooklyn Papyrus and in the Execration
Texts.
been one also. Ibi
6
c
If Epir was a Semite,
then his mother may well have
can be a good Egyptian name,
7
but the
West-Semitic
name 'Abiya (hypocoristic for 'Abi-DN) is well attested in Amorite (as at Mari), later Hebrew, etc. To the onomastic data proper two other items may be briefly added. First, on Rio Inv. 643 reverse
process:
an
[2433], Cat. No. 17, we have fresh data on the
Egyptian
name
that
Semitic (and, via Greek, in modem usage). shortened to ssn,
in
either ssn
West sssn
'Susan'.
occurs, that would give an even closer
equivalent for the borrowed Semitic form.8 become
currency
and entered Hebrew to give us Shoshan(na),
On this Rio stela, the variant ssn
could
gained
It is well known that Eg.
(progressive
Evidently, in Egyptian,
assimilation)
or ssn
sssn
(regressive
assimilation). Second, the stela Rio Inv. 652 [2440], Cat. No.35, attests (in the NK) a foreign loanword that goes straight back to the Middle Kingdom era, as is evidenced by its frequent presence in the Mari archives, and helps to clinch
its meaning: the word hrr,
in Semitic haruru(m),
hururu(m).
This term is used for a bedspread of a particular type, corded or striped cf. our candlewick bedspreads in more recent times. the owner Haunefer 'Chief
maker
of
Pharaoh, L.P.H.'
is entitled
bedspreads
was evidently
favour, and was then made in Egypt itself. Final 6
c
hry irw hrr n t hnkt Pr- 3,
corded/candlewick The type
On his Rio stela,
c
for
imported
the into
c
.w.
s. ,
bedroom Egypt,
of
found
9
Perspective
Cf. W.F. Albright, Journ. Amer. Oriental Socty 74(1954), 225, No.9, references. ?Cf. Ranke, op.cit. (n.5), 1,19:16, 20:7-10. Cf. Kitchen,Varia Aegyptiaca 3(1987), 29-31. S e e Kitchen,"Of Bedspreads and Hibernation: from Rio de Janeiro to the Middle Euphrates", in Ascribe to the Lord: Biblical and other studies in memory of Peter C. Craigie, ed. L. Eslinger, G. Taylor, (JSOT Supplement Series, 67), Sheffield, 1988, 21-27. 8
9
90
K.A. Kitchen After the Middle Kingdom itself, the Rio monuments continue to show links with Western Asia. Suffice it to recall Raia of stela Inv. 654 [2442], Cat. No.33, who bore the the Canaanite title qasin,
'leader, boss', a term
found in both Ugaritic and Hebrew.* ^ Thus, the continuing publication of such museum collections should have a high priority, so that our work - on the Middle Kingdom as on all else - may have available the maximum amount of
first-hand
data upon
which to draw for its findings; rich are the gains to be made thereby!
up
M
rj—i
.11
o HI III £U hörnern man
2 mulheres 2 women
oferendas offerings
T t
A
« 15
mulher woman
.. PL
I0
man
U
f«*/i ^?
13 h o m e m
t
oferendas offerings
oferendas offerings
n
°
m
a
n
Kitchen, JEA 73(1987), 218-220, fig. 1.
m
a
If
n
oferendas offerings
TEACHINGS, DISCOURSES AND TALES FROM THE MIDDLE KINGDOM
R. B. P a r k i n s o n
INTRODUCTION One of the landmarks in the study of Egyptian literature is Posener's pioneering article 'Les richesses inconnues de la littérature égyptienne' (RdE 6 [1951], 27-48; supplemented in RdE 9 [1952], 117-20). Although he was careful to explain why certain texts were omitted fi Dm his catalogue of 'oeuvres purement littéraires', he offered no definition of 'literary' qualities. The meaning of this term can, however, no longer be taken for granted, and much discussion has attended the problem of defining 'literature' in any context. Critical theory has seen a movement from semantic and structural definitions of what is literary to historical definitions: literature is analysed as a cultural construct. Semantic theories have identified certain linguistic features of a text as the distinguishing marks of literary composition: thus 'semantic density' is one such mark, and yet this can be lacking in a clearly literary text and present in a non-literary one. Structuralism has 'recognized that the literary work, like any other product of language, is a construct', that it is not an essentially unique form of discourse. Literature is an institution, and should be approached accordingly. This is not the place to attempt a definition of literature; what I wish to do is to provide a revised version of the Middle Kingdom sections of Posener's catalogue - which is itself a good Egyptian genre. This, however, requires a few preliminary remarks on 1
2
3
4
1
Egyptological discussions include: J. Assmann, 'Der literarische Text im alten Ägypten: Versuch einer Begriffbestimmung', OLZ 69 (1974), 117-26; P. Kaplony, 'Die Definition der schönen Literatur im alten Ägypten', in J. Assmann et al. (ed.), Fragen an die altägyptische Literatur: Studien zum Gedenken an Eberhard Otto (Wiesbaden: Reichert, 1977), 289-314; H. Brunner, LÄ III, 1067-8; A. Loprieno, Topos und Mimesis: zum Ausländer in der Ägyptischen Literatur ( Ä A 48, 1988). A n accessible guide to literary theory is: A. Jefferson & D . Robey, Modem Comparative Introduction (London: Batsford, 1986).
Literary
Theory:
a
2 See for example S. T. Olsen, The End of Literary Theory (Cambridge: Cambridge University Press, 1987), 73-87; T. Eagleton, Literary Theory: an Introduction 3
4
(Oxford: Blackwell, 1983), 1-16, passim.
Eagleton (n. 2 above), 106. J. Baines, 'An Abydos List of Gods and an Old Kingdom U s e of Texts', in Baines et al. (ed.)
Studies and Other Essays Presented
to I. E. S. Edwards
(EES Occasional Publications 7), 130-33.
91
Pyramid
92
R.B. Parkinson
principles of selection and dating. THE DEFINITION OF THE CORPUS
In the light of modern theory, Posener's corpus cannot be upheld as the totality of Egyptian literature, which might be defined, as it has been for 18th century England, as a recorded body of statements of cultural value, beyond the immediate and practical conveyance of information. His catalogue consists of two groups of texts, which have often been correlated intuitively with modern impressions of what constitutes 'literature': wisdom texts and tales. A general definition of genre is problematic, and this is particularly true of Egyptian genres, which have no direct posterity in western tradition and are not characterised by a consistent terminology. No specific designation for a tale is known, and the terms sb3jt 'teaching' and mdt 'discourse' can describe a wide range of texts, although they have a more specific reference to genre in certain contexts. This absence does not imply that no concept or system of genre existed; 'the processes of generic recognition are . . . fundamental to the reading process'. The genres of Posener's corpus can be described as follows: the tales are non-commemorative narratives; the wisdom texts are a group of various genres, some explicitly defined and others less so. While the wisdom tradition is expressed in other genres, such as the funerary autobiography, the wisdom texts themselves are predominantly sapiential discourses, all concerned with certain ultimate values, such as comprehending the nature of the ideal life. The coherence of the generic group of wisdom texts is established by form, theme and style. What I discuss here is the association of these two generic groups in the institution of Egyptian literature. No explicit ancient definition of literature as a cultural construct is preserved, nor is there any Egyptian critical discourse or analysis of individual texts. The place of Posener's corpus within the wider entity of Egyptian literature must be inferred. Evidence is provided from within the texts by their treatment of genre and topoi, and more external evidence can be found in the context of the texts, namely the audience and the occasions of reading, the textual tradition, and the status of the author (including the fictionality of the protagonist). I discuss the context first. 5
6
7
8
9
10
The context of literature is a deeply problematic criterion for a historical definition: the
See Eagleton (n. 2 above), 17. ^A. Fowler, Kinds of Literature: an Introduction to the Theory of Genres and Modes (Oxford: Clarendon Press, 1982); H. Dubrow, Genre (The Critical Idiom 42; London: Methuen, 1982). 7 H. Brunner, 'Die "Weisen", ihre "Lehren" und "Prophezeiungen" in ägyptischer Sicht", ZAS 93 (1966), 32-33. G. Posener, RdE 6 (1951), 46-7. On the general phenomenon s e e the remarks of J. Baines, Fecundity Figures: Egyptian Personification and the Iconology Fowler (n. 6 above), 130-48. o
of a Genre (Warminster: Aris & Phillips, 1985), 36.
8
Fowler (n. 6 above), 259; cf. 256-72. See also E. D . Hirsch, Validity in Interpretation
( N e w Haven:
Yale University Press, 1967). Compare the approach o f G. Buccellati, 'Wisdom and Not: the Case of Mesopotamia', (1981), 35-47. 1 0
JAOS
101
Teachings, discourses and tales
93
social Sitz im Leben is unknowable, but the 'implied readers' of a text must have been members of the élite. Although it is not necessarily any single private individual who is addressed, the mode of address is itself personal. " The discursive texts often address an audience directly, but they are self-referential - that is, an audience is defined within the setting of the text itself (for example, the children for whom the Teachings are made). This description of an audience often includes a portrayal of the audience's response. A favourable reaction is given to teachings and discourses, in which they are 'perfect (nfr) to the heart'(e.g. P. Prisse 2.6-7, P. Berlin 3025, 131). The same reaction is evoked by other genres, such as eulogies and hymns (e.g. Urk. IV, 685.10-2), and in the Tale of the Court of Cheops a similar phrase describes king Snefru's reaction to the sight of maidens rowing (P. Westcar 5.14-15). In autobiographies there are the epithets 'one who speaks well (nfr) before his lord' and 'one who says what is good (nfrt)', and the 'perfection' which is mentioned in wisdom texts can be an ethical quality (as is seen in Ptahhotep's evocation of his son's response as obedience and wisdom resulting in ethical behaviour rather than pleasure: ed. Zâba, 507-87). These instances suggest that 'perfection' of words is not necessarily or exclusively an aesthetic quality. However, aesthetic pleasure is a relevant concept, as is shown by the 'perfect words and choice phrases' of Neferti which are for royal 'entertainment (dßjj-hr)' (ed. Helck, 2j-k). These features accord with Loprieno's definition of literary texts as 'die von den zwei Variablen eines selbstständigen Autors auf der einen Seite und gewisser ästhetischer und gesellschaftlicher Erwartungen des Rezipienten auf der anderen ausgehen'. They do seem to be particularly, if not exclusively, associated with Posener's corpus. These expectations and responses are not inherent in the literariness of the discourse, but are themselves defined by an awareness of the literary institution. More indirect evidence for the nature of this institution is supplied by the groupings of texts in libraries, one of which is known from the Middle Kingdom. The wide range of texts in this (the 'Ramesseum library') has suggested the broad definition of literature already advanced: texts from Posener's corpus were present together with an onomasticon, hymns, rituals, and medical and magical texts. The physical context of the copies of texts is not a unique or infallible criterion for the isolation of a genre or a group of genres within the literary institution, and the existence of a text on papyri or ostraca does not correspond to a distinction between it and other texts preserved on other media. To give a practical example: the 'Loyalist' Teaching was first known only from a copy on a stela; might not then didactic texts on other funerary monuments be 11
1
13
14
15
J. Baines (under xxi) JEA 76 (1990), 57, n. 9. On the compositional background to literature s e e C. J. Eyre, 'The Semna Stelae: Quotation, Genre and Functions of Literature' in S. Israelit-Groll (ed.), Studies in Egyptology Presented to Miriam Lichtheim (Jerusalem: Hebrew University, 1990), I, 160-5. 12 see Loprieno (n. 1 above), 84-97. 1 3
J. M. A . Janssen, De traditioneele
egyptische Autobiografie
voor het Nieuwe Rijk (Leiden: Brill, 1946),
69 (Bb.4), 122-5 (He, passim). 1 4
1 5
Loprieno (n. 1 above), 14. See Eagleton (n. 2 above), 80.
Another, the 'Berlin library', can be posited, but its exact extent is unknown. It seems, however, t o have contained only tales and a wisdom text. See R. B. Parkinson, The Tale of the Eloquent Peasant (Oxford: Griffith Institute, in press), 1.2.
R.B. Parkinson
94 16
copies of otherwise lost wisdom texts? A text could also be transferred from a monumental context to a manuscript, such as the building inscription of Senwosret I 