Among the private statuary of the Old Kingdom is a group called reserve heads or portrait heads (Ger., Ersatzköpfe or Porträtköpfe Fr., têtes de remplacement or de réserve). The label “magical heads” seems more accurate, because it expresses the contradictory functions of these objects, as revealed by recent inquiry.

At this writing, thirty-three magical heads have been discovered, as well as eight isolated ears separately carved and initially stuck with plaster. The Giza head presented to the media in 1992 is not to be counted as a magical head: the beginning of the shoulder and the irregular break of the neck clearly demonstrates that this head was part of a statue. However, a small unfinished head (unpublished), found in 1989 by the Lisht expedition of The Metropolitan Museum of Art, in a twelfth dynasty stratum, may tentatively be considered a late resurgence of this type of object.

Typological Features.

Reserve heads are metonymical objects: a head and its neck, without the shoulders. The neck section is flat-bottomed so that the head does not need any further support. When the head is posed on the flat section, its gaze is much higher than the horizon line and the gaze of complete statues (except in very rare exceptions). The famous Nefertiti bust and the other composite statues of Tell el-Amarna have nothing in common with these heads, nor do the busts emerging from false doors in some Old Kingdom mastabas (e.g., Neferseshem-ptah and Idw, Giza, sixth dynasty). The bust of the vizier Ankh-haf (reign of Khafre, fourth dynasty) is however an exception, since it shares some features with the magical heads, such as the ablation of the ears.

For the private statuary of the Old Kingdom, a smaller scale than life-size was usually preferred, but, except for one late occurrence, the magical heads are all life-size. They are always monochrome, unlike other statuary of the Old Kingdom. The heads usually have the color of their material, a white limestone; two examples in grey siltstone from the Nile have been found. In three cases, tiny traces of paint have been discovered (two red and one black).

The most characteristic feature of the magical heads is that they present at least one, and frequently more, of the four following types of marks, produced by a clumsy hand certainly after the finishing touch of the sculptor: a circle around the neck, a little higher than its base; a groove running from the skull to the nape; the ears destroyed, unstuck, or omitted; and a clumsy outline of the edge of the close-shaven hair. None of these features appears on any private statues except the magical heads.

Geographical Situation.

The origin of these heads (and ears) is the area around Memphis in the broadest sense of the term (except for a head from a private collection in Belgium and one at University College, London, of which the provenance is not certain). The vast majority of them have been found in Giza. A few isolated examples come from Abusir, Dahshur, and Abu Roash.

Chronological Situation.

The most ancient magical head, found in Dahshur (Cairo CG 519), can be dated to Sneferu's reign (beginning of the fourth dynasty). It is during this period that the magical heads ritual seems to have appeared. The ritual came into particular favor under Khufu (the majority of the heads come from the G4000 cemetery of Giza). The heads are also well attested in Khafre's cemetery, G7000. The latest examples (fifth and sixth dynasties), isolated in time as in space, are characterized by poor stylistic quality. This stylistic decay seems to be parallel to the progressive extinction of the specific ritual practice.

Situation in the Tomb.

H. Junker, the Austrian archaeologist who led excavations in the Giza necropolis at the beginning of this century, concluded from his observations that the heads were fixed into the masonry of the wall separating the funerary chamber from the bottom of the shaft, behind a monolithic portcullis. Since the chamber invariably opened to the north, the heads then looked in the direction of the circumpolar stars. While destroying the blockage, tomb-robbers let the heads roll to the bottom of the shaft or inside the chamber. It can be concluded from these observations that the magical heads belonged to the underworld, a context that differentiates them from other private sculpture of the Old Kingdom, which was always placed in close contact with the living (in serdabs or in offering chapels). This specific situation must be related to the marks described above.


When first discovered, the heads were believed to be prosthesis heads to replace magically the head of the dead, if it were damaged. This theory was largely developed by H. Junker (1914) who came to suggest, after his excavations in the fourth dynasty necropolis of Giza, an identity of function between the realization of the magical heads, the practice of mummification, the creation of plaster masks, and even private statuary in general. This interpretation does not attribute to these heads any specific meaning inside the semantic field of the image as a substitute of the dead. The mutilating marks, for instance, are not taken into account; their meaning could be of some importance, however, because this kind of mark is never found on any other type of substitute, whether mummies, masks, or normal statues. Incomplete though it might prove to be, Junker's theory became a dogma accepted by almost all Egyptologists.

In contrast to the idea of the heads being substitutes come A. L. Kelley's (1974) and N. B. Millet's (1981) theories. In their two short articles, they suggest that these heads were models for sculptors, used in the making of the funerary masks or of plaster casts. The breaking of the ears and the damage to the nape would have been caused during the unmolding of the head. This theory has major flaws. First, it does not take into account the particular situation of the head within the tomb. Second, we may wonder why a head should have been so carefully carved just to produce one funerary mask. Third, the idea of sculptors making a funerary mask using such a technique is irrelevant: the heads should have been bigger than lifesize, so that the masks would fit over the thick linen wrapping the head of the deceased. Fourth, it is unlikely that an Egyptian artist would be unable to unmold a mask without damaging the original. Finally, if these heads were sculptor's models, how can we explain the fact that no normal private statue has been found belonging to the important individuals who had a magical head set among their funerary furniture?

The most recent study on the subject (Tefnin 1991) is based on a thorough reading of the archaeological documents, on the historiographical analysis of the previous theories, and on connecting the heads to the magical and religious literature. There is, however, a serious methodological problem linked to the late date of the texts (eighteenth dynasty at the earliest for the Book of Going Forth by Day [Book of the Dead] and the ritual of the Opening of the Mouth; Greco-Roman period for the rituals of Apophis' destruction). There does not exist any text contemporary with the heads. Because of the scarcity of texts from the fourth, fifth, and sixth dynasties, two choices appear: either we decide to ignore any explanation, or we agree to extrapolate, knowing the traditional character of Egyptian thought and myth.

Thus, the circle around the neck, which could not have been made by the hand of a sculptor, evokes the vast semantic field of decapitation, attested in some prehistoric cemeteries, well known in pharaonic law and used to punish Apophis (enemy of Osiris) in the late magical texts (Papyrus Bremner-Rhind, Papyrus Jumilhac). It can also be connected with execration plaques buried in the Giza necropolis during the eighteenth dynasty, studied by G. Posener. Dangerous dead persons were stamped on these plaques in the guise of prisoners; their throats were then cut by a line of red paint, certainly during the performance of a ritual. As in the case of the magical heads, we thus discern two distinctive moments: first, the fashioning of the figure by a trained artist or craftsman; subsequently, a violent and destructive gesture performed by a nonartist, certainly a ritual specialist. The deep groove at the back of the head represents a fracture of the skull, with blood gushing as in the A14 hieroglyph of a wounded prisoner, meaning “the enemy” or “the dead.” The magical plaques of Giza, mentioned above, show the same bleeding wound, painted in red on top of the stamped figure during the performance of a violent ritual. This ritual action can be connected with a rather enigmatic moment of the well-known ritual of Opening of the Mouth, a moment of extreme tension between the sem-priest (eldest son or priest representing the family), the sculptor who is asked to create a likeness of the father, and a craftsman called “polisher” (?) who has to take the father's head and hit it despite the priest's opposition. This one says: “I am Horus-Seth, I shall not allow you to make my father's head become white (sḥd).” Is not this head “made white,” bleeding white, in the text of the ritual, exactly equivalent to the white limestone magical head, deeply grooved at the back? And does not the text suggest admirably the double nature of these heads: first a naturalistic sculpture made by a master sculptor, and then violently hit during the ritual?

The texts are not so clear about the mutilation of the ears, sculpted as a whole or stuck to the head with plaster, and then cut off with a wooden blade or a metallic chisel, or even omitted in late occurrences. Those marks are so evident that we cannot avoid the idea of a ritual deafening of the dead, imprisoning him in the ἰgrt, the world of silence, and reducing him to sensorial impotence, like Apophis's punishment. The clumsy outlining of the hair is also evident and specific for this group of heads. No text explains clearly that ritual practice, so that we may only appeal to anthropology, which provides many examples of an association between hair and life.

From the contrast between the perfection of the carving and the clumsiness of the ritual marks there arises a semantic ambiguity: these heads were meant to be beautifully carved, even bearing resemblance to the deceased— as may be suggested by the words of the Opening of the Mouth—but they were also magical objects that could become dangerous when placed close to the dead, and that had therefore to be ritually “canceled.” The same superstitious fear led, as is well known, to mutilation of the hieroglyphs eventually written in the underground part of the tomb (but never the signs written in the funerary chapel, which belonged to the world of the living). However, there remains a major mystery: Why did the eminent individuals of that short period in the Old Kingdom not all have such magical heads in their tombs? The idea of royal favor must be rejected. It remains the hypothesis that certain dead persons were potentially more dangerous than others, owing perhaps to the circumstances of the death (e.g., drowning, violent death in general, or an ill-fated day of death). Nevertheless, it seems that the link between these magical heads and execration rituals cannot be denied.


  • Junker, Herman. Giza. Vols. 1, 7, 8, 12. Vienna, 1929–1955. These volumes of the monumental excavation reports of the Austrian team at Giza contain all information on the conditions of discovery of the heads in the cemeteries near the pyramids of Khufu and Khafre.
  • Junker, Herman. “The Austrian Excavations, 1914.” Journal of Egyptian Archaeology 1 (1914), 250–253. Brief account of the excavations mentioned above, already containing the major elements of Junker's interpretation.
  • Kelley, A. L. “‘Reserve Heads’: A Review of the Evidence for their Placement and Function in the Old Kingdom Tombs.” Journal of the Society for the Study of Egyptian Antiquities 5 (1974), 6–12.
  • Millet, N. B. “The Reserve Heads of the Old Kingdom.” In Studies in Ancient Egypt: Essays in Honor of Dows Dunham. Boston, 1981. These two studies propose an alternative to the theory of the reserve heads: they are described as models of sculptor and/or models for plaster masks.
  • Smith, W. S. A History of the Egyptian Sculpture and Painting in the Old Kingdom. 2d ed. Boston, 1949. In this monumental book, the author reviews Junker's theories and adds precious information extracted from the excavations of G. A. Reisner (Harvard-Boston Expedition) in the same necropolis. Excellent illustrations.
  • Tefnin, Roland. Art et magie au temps des Pyramides: Lʾénigme des têtes dites “de remplacement.” Monumenta Aegyptiaca, 5. Brussels, 1991. Not opposed to Junker's conclusions but completing them, emphasizing the mutilation inflicted on the heads, not taken into account in the earlier theory. Also offers a detailed, illustrated catalog of all known heads and fragments, with bibliography.

Roland Tefnin; Translated from French by Angélique Corthals