Navigation – Plan du site

AccueilNuméros121Polythéismes antiquesOsiris and the Deceased in Ancien...

Polythéismes antiques

Osiris and the Deceased in Ancient Egypt :
Perspectives from Four Millennia

Conférences de l’année 2012-2013
Mark Smith
p. 87-102

Résumé

The four lectures in this series took place between 17 May and 27 May 2013. Their subject was how the ancient Egyptians conceived of the relationship between the god Osiris and those who died. The lectures did not provide a continuous or comprehensive account of Egyptian beliefs about this relationship, but rather focused upon four distinct periods of their development, spread over four millennia. The periods selected for attention were ones in which significant changes in the Egyptian conception of the Osirian afterlife are known to have occurred, or where it has been argued that they did. An important aim of the lectures was to investigate when and why such changes happened, and how they can be recognised in the historical and archaeological record.

Haut de page

Entrées d’index

Haut de page

Texte intégral

  • 1   The material presented in them is a condensed version of selected chapters from the speaker’s for (...)

1The four lectures in this series took place between 17 May and 27 May 2013. Their subject was how the ancient Egyptians conceived of the relationship between the god Osiris and those who died. The lectures did not provide a continuous or comprehensive account of Egyptian beliefs about this relationship, but rather focused upon four distinct periods of their development, spread over four millennia. The periods selected for attention were ones in which significant changes in the Egyptian conception of the Osirian afterlife are known to have occurred, or where it has been argued that they did. An important aim of the lectures was to investigate when and why such changes happened, and how they can be recognised in the historical and archaeological record1.

I. Unreading the Pyramid Texts : So who is Osiris ?

2The title of the first lecture was ‘Unreading the Pyramid Texts : So who is Osiris ?’ This dealt with the problem of the origin of Osiris and his relationship with the deceased in the Old Kingdom. Despite numerous efforts, we still do not know how or why the god entered the Egyptian pantheon. According to a view espoused by Racheli Shalomi-Hen, Osiris was originally a foreign deity, since the determinative used in the earliest writings of his name is a seated figure with a beard, which she thinks is characteristic of foreigners. However, the beard in question is not the same as those worn by foreigners, and it is seen in representations of other Egyptian deities well before Osiris makes his first appearance. Thus there is no evidence to support the idea that he was of foreign origin.

3Bernard Mathieu has argued that the creation of Osiris and the diffusion of his cult throughout Egypt were decreed by the state. The problem with this theory is that there is no historical evidence to support it. Instead, Mathieu uses passages selected from the Pyramid Texts as proof of his view. Thus, according to him, the messengers mentioned in one spell must be the royal messengers who spread the Osirian gospel throughout the country. Threats made against various beings in another spell must reflect the threats made against adherents of other deities who were reluctant to accept the new god. This is what is known as an adaptive reading, one in which texts are read in a certain way to support a previously determined interpretation. The argument is essentially a circular one and therefore not compelling.

4In the earliest sources that refer to Osiris, inscriptions in private tombs dating to the fourth or the fifth dynasty, the god and the deceased are regarded as distinct beings. There is nevertheless a connection between them, since the dead are dependent upon the deity for their needs in the afterlife. He ensures that they receive sustenance and enjoy other benefits. Osiris is not the only god who does this, however, but one of a number of deities who are said to care for the dead and perform services for them. Some, like Anubis, are attested in this role before Osiris is.

5In the Pyramid Texts, the earliest specimens of which date to the end of the fifth dynasty, a clear distinction is made between Osiris and the deceased as well. But in that corpus the situation is more complex, because we also find the dead identified with the deity. They can even be distinguished from Osiris and identified with him in the same spell. Moreover, Osiris is not the only god with whom the deceased share this complicated, apparently contradictory, relationship. Some of the same spells that identify the deceased with Osiris identify them with other divinities as well. But they can also be distinguished from those divinities. How can we explain this paradox ?

6In the lecture it was argued that the key to understanding the Pyramid Texts is to recognise them for what they are, ritual utterances, not expositions of theology or descriptions of what the Egyptians thought the afterlife would be like. Their constituent sentences are functional, designed to make things happen. This is true of all Egyptian ritual utterances, so it is important for our reading of the Pyramid Texts that we situate them within a continuum of works designed to be recited in ritual contexts rather than viewing them in isolation.

7One important technique used in ritual texts is identification. The ritualist identifies himself with a deity or some other being, or attributes such identity to a third party. But the purpose of spells employing this technique is not to transform the ritualist or the third party permanently into a deity or another being. Often their goal is something completely different. Claims of identity with another being, or attributions of that identity to someone else, are a means to this end, whatever it may be, not an end in themselves. Such claims are valid within the framework of the ritual, inasmuch as they help to achieve its effect, but they do not describe an objective reality. That is why we sometimes find two seemingly contradictory identifications, or identifications with more than one being, asserted in the same spell. If we read the Pyramid Texts in this way, then many of their apparent inconsistencies are resolved. In particular, those spells that both distinguish the deceased from Osiris and identify them with the god become more comprehensible.

8But this approach to the interpretation of the Pyramid Texts raises a serious problem. If what a spell says is not always a reliable guide to what it is supposed to do, then how can we determine its function ? If spells that identify the deceased with Osiris or say that they will become Osiris were not intended to transform them into that deity, what were they supposed to do ? What were the expectations of those who used them or those for whose benefit they were recited ? What did they hope would happen to them as a result ? Beyond the words of the spell and the ritual context in which they were recited, the Egyptians believed there was a wider reality, a world beyond the spell. How can we determine what impact the recitation of the spell was supposed to have in this world ?

9One important indicator is provided by paratextual evidence : titles, rubrics, colophons, and so on. Although relatively rare in the Pyramid Texts, such notices are valuable since they are less ambiguous than the spells themselves. They comment upon the function of these without being embedded in their ritual context. Of particular interest are the colophons of Pyramid Text Spells 456 and 561B. According to the first, whoever knows and recites that spell will be an intimate of the sun god and join his following. According to the second, whoever worships Osiris and recites the spell for him will live for ever. Thus the colophons state explicitly what the spells to which they are attached are supposed to achieve. The desired result in both cases is that the deceased be subordinate to a deity, a member of his following or a worshipper, not identified with or transformed into him.

10The divine-human relationships envisaged in these colophons are actually reflected in some Pyramid Text spells. Thus those spells can provide us with reliable information about Egyptian aspirations for the afterlife, provided that they are used judiciously. In the lecture, the following approach to analysing them was advocated. If a specific statement about the fate of the deceased in a Pyramid Text spell is paralleled either in a remark like those in the colophons of Spells 456 and 561B, or in another less ambiguous Old Kingdom source or sources outside the corpus of the Pyramid Texts, then we are justified in accepting this statement as evidence of something that the Egyptians of that time actually hoped or expected would happen to them after they died. But if that statement is contradicted by such paratextual evidence, then it was probably valid within the context of the ritual, but had no reality beyond the world of the spell. In other words, paratextual evidence is a more reliable guide to Egyptian aspirations for the afterlife than the Pyramid Text spells themselves are and should be given more weight accordingly.

11What less ambiguous sources outside the corpus of the Pyramid Texts do we have to help us understand Egyptian hopes for the afterlife in the Old Kingdom ? The most important are the wishes in offering formulas in private tombs, since they give us a good idea of the sort of things to which the deceased aspired during that period. We can find direct parallels for all of these in both the paratextual evidence we have identified within the Pyramid Text corpus and in actual Pyramid Text spells. Not only are the same hopes and aspirations found in private offering formulas reflected in those spells as well, they cluster together in the same groups in the Pyramid Texts as they do in the private offering formulas, indicating that both drew upon a common source. This suggests that the aspirations of kings for the afterlife were fundamentally the same as those of their subjects. Moreover, the evidence of contemporary private tomb inscriptions shows that the latter sought to fulfil these aspirations in the same way as their rulers, by means of spells and ritual utterances. The spells in question were called sakhu or glorifications, and in fact these are attested for non-royal individuals well before they are known to have been used by kings.

12As far as we can judge, the relationship between Osiris and the deceased envisaged in the Pyramid Texts is the same as that found in contemporary private tomb inscriptions. The deceased, whether royal or non-royal, are distinct from the god and subordinate to him. Statements in some Pyramid Text spells that the deceased is Osiris or will become Osiris, like those identifying them with other deities, are valid in the specific context of the ritual during which they are uttered, but not beyond this. Likewise, the locution Wsỉr NN, in which the name of the deceased is preceded by that of Osiris, does not indicate identity between them. The use of this locution was restricted to certain contexts ; it served initially to identify the dead as recipients of sustenance under the auspices of Osiris, although it soon came to be a more generic marker of the deceased’s association with the god.

13In ‘unreading’ a text or corpus of texts one attempts to show that a previous reading is incorrect. At its most basic, a reading is simply a translation, in other words, it is concerned with what a text says. On another level, it involves interpretation, what a text is about. Finally, a reading can be concerned with purpose or function : what a text is for and what it is supposed to do. ‘Unreading’ can work at any or all of these three levels. The unreading of the Pyramid Texts offered in this lecture identified certain problems with previous readings of that corpus. In particular, it focused upon the type of reading that assumes Pyramid Text spells are a sort of menu or blueprint setting out what the Egyptians of the Old Kingdom wanted to happen to them after they died, where every statement can be taken at face value. But in fact, the Pyramid Texts are composed of ritual utterances and should be read and interpreted as such. So who is Osiris ? We cannot say much about his origins, but from his earliest appearance in the historical record he is already a deity closely associated with the dead. Who can be Osiris ? In the ritual moment, everyone can be Osiris, but in the world beyond the ritual there is only one god, with whom every deceased person hoped to enjoy the same relationship.

II. Re resting in Osiris, Osiris resting in Re :
Osiris, sun god and the deceased in the New Kingdom

14The title of the second lecture was ‘Re resting in Osiris, Osiris resting in Re : Osiris, sun god and the deceased in the New Kingdom’. This examined whether the relationship between the deceased and Osiris was influenced by developments in solar religion during the New Kingdom and, if so, how. The lecture was divided into two parts. In the first part the status of Osiris as god of the dead during the Amarna Period was investigated. According to one widely held view, there was no place for Osiris or the Osirian afterlife in the religion of Akhenaten. As a result, the elaborate belief system that had grown up around the deity was replaced by a much simpler one. The deceased were buried in their tombs, they slept in them at night, and awakened at dawn. Each day their bas would leave these resting places in order to participate in the cult performed in the temples in Akhetaten. The one who provided the deceased with the means to do this was Akhenaten himself, who was the only guarantor of the afterlife.

15However, unlike Amun, there is little evidence for proscription of Osiris during the Amarna Period. We even know of cases where his name and image were left intact but those of Amun erased. Moreover, we do find references to Osiris, as well as some other deities associated with the afterlife like Anubis, in texts inscribed during Akhenaten’s reign. The lecture surveyed a number of examples. So evidently belief in the god and his relationship with the dead was preserved. Some think that this was only a marginal phenomenon surviving alongside the dominant official cult of the Aten or solar disk. But according to one theory, proposed by Alfred Grimm and Hermann Schlögl, Osiris was actually incorporated in the new theology of Akhenaten. They maintain that the king neither proscribed nor ignored Osiris, as some have thought, but rather deliberately took over his functions and characteristics, even retaining the god’s traditional iconography, so that in effect he became a new Osiris for his subjects, removing any need for the old one. Specifically, Akhenaten identified himself with Osiris as the son of the Aten.

16The evidence for this view is chiefly iconographic, consisting of what Grimm and Schlögl claim are two and three dimensional depictions of Akhenaten in the form of Osiris. Several of these were presented and analysed in the lecture. It was argued that in cases where an unambiguous representation of Akhenaten is involved, it is not actually in the form of Osiris, while in cases where we have an unambiguous representation of Osiris, there is no obvious connection with Akhenaten. In particular, there are no convincing examples where the god is depicted with the facial features of the king, contrary to what Grimm and Schlögl maintain. Other problems with their theory were identified as well, not least the fact that the dating of many of the objects that they cite in support of it to the Amarna Period is questionable, and it was concluded that there is no basis for thinking Akhenaten ever identified himself with Osiris.

17Despite this, the evidence examined in the lecture suggests that we probably need to reconsider some of our fundamental assumptions about the afterlife in Amarna religion, for example, the idea that Akhenaten thought only in terms of a single world, this one, inhabited by both the living and the dead. References in contemporary texts, some in inscriptions from tombs of high-ranking officials at Amarna itself, show that the underworld as a distinct realm of the dead remained an important concept. The evidence for continued belief in Osiris as a god of the dead during this time is more abundant than one might have expected. Nor is it confined to a few specific geographical areas or years within Akhenaten’s reign. Was he aware of this survival and, if so, was it a matter of concern to him ? If it was, then the king’s ability to influence developments in the religious sphere and, in particular, the beliefs of his subjects may have been more circumscribed than is generally supposed.

18The second part of the lecture investigated the phenomenon of the ‘solar-Osirian unity’ and how this affected ideas concerning the relationship between Osiris and the deceased. The Egyptians imagined that the sun god Re entered the western horizon and passed through the underworld each night. At a certain point during this nocturnal journey, the solar deity encountered Osiris and the two experienced a temporary union, which can be described in Egyptian sources as ‘Re resting in Osiris and Osiris resting in Re’. This union had a positive effect on both participants. As a result, Re emerged newly born from the eastern horizon while Osiris, who remained behind in the underworld, was revivified. The conception of this nightly union becomes especially prominent in the New Kingdom, when it figures in both guides to the underworld and the Book of the Dead, although some would trace its existence as far back as the Old Kingdom.

19One view maintains that a change in the conception of the solar-Osirian unity took place in the latter part of the New Kingdom. According to this view, the nocturnal union of Re and Osiris ceased to be regarded as a temporary merger of the two gods, and was seen instead as something more substantial and permanent, resulting in a completely new type of composite deity, the giant, cosmos-spanning figure of Re-Osiris, described in texts as the great god. It was this new divinity who emerged from the eastern horizon at dawn, and it was with this figure that the dead were now associated. Thus the traditional relationship between Osiris and the deceased changed dramatically as a result of this new conception.

20The lecture examined the evidence in support of this view and found it wanting. In particular, the idea that a group of compositions preserved in the tombs of Tutankhamun, Ramesses VI, and Ramesses IX illustrates the new doctrine of the solar-Osirian unity did not stand up to close scrutiny. Joachim Quack has demonstrated convincingly that the main theme of these compositions is not the union of Re and Osiris. Rather, they are concerned with the beings the solar deity encounters in the underworld and how he interacts with them. The sun god is all-important, and Osiris much less so. The texts never speak of the union of the two divinities. A number of passages that have been interpreted as references to this union are simply references to the sun god.

21Another source that has been cited as evidence for the new doctrine of the solar-Osirian unity is the Book of Caverns. It has been claimed that the end of this text describes how Osiris leaves the underworld merged with Re in the form of the great god. In fact, this does not happen. Re and Osiris are clearly distinguished throughout the Book of Caverns. The former is said to perform various services for the latter. Osiris is in the following of Re and adores him. At the conclusion of the text, Osiris requests and is given a place in the bark of the sun god, just as the deceased hoped they would be given one. But his position is clearly that of a dependent. Those who tow the bark acknowledge only Re as their passenger. They make no reference to Osiris or to any composite form involving him and the solar deity.

22In the Book of Caverns and in the other underworld guides as well, the pre-eminent position is always given to the sun god. He is very much the dominant figure, the main actor, and features in every scene. Osiris is important but he is only one of a number of beings that the sun god encounters in the underworld. Osiris continues to exercise his traditional functions in relation to the dead in these texts, but the solar deity seems to act in a supervisory capacity, with Osiris subordinate to him. It is this emphasis upon the primacy of Re, rather than any new conception of the nightly union between him and Osiris, that characterises the underworld guides and other sources of the New Kingdom.

23These texts assign Osiris a crucial part in an eternally recurring cycle of rebirth and regeneration. As a result, he is not just the ruler of the underworld and judge of the dead, but a deity on whose continued well-being and daily rejuvenation the entire cosmos is dependent. Is this concept an innovation of the New Kingdom ? We are unable to say, mainly owing to uncertainty over the original date of composition of the underworld guides. What does seem innovatory is the fact that they begin to be inscribed in tombs at this time, and references to the concept start to appear in other sources like the Book of the Dead. So perhaps what is new is that these texts begin to be linked with the deceased more regularly.

24What effect does all this have on the relationship between Osiris and the dead ? How do they benefit from the link between Osiris and the sun god ? In the lecture, four benefits were identified. First, the fact that the cosmos continues to function is obviously a precondition for the continuity of the afterlife. Second, the relationship between Re and Osiris provides a model for that between the ba and the body of the deceased. The ba alights upon the body and regenerates it each night just as Re meets with Osiris. Third, the dead can participate in the solar cycle, for example, by travelling in the sun god’s bark, following the precedent of Osiris. Fourth and finally, knowledge of what happens in the underworld is important. What is beneficial for the deceased is not just that this cycle takes place, but that they know about it as well, since such knowledge is a means of gaining admittance to the entourage of Osiris. But the connection between Osiris and the sun god does not affect the basic relationship between Osiris and the deceased. What happens instead is that this relationship, like every other aspect of life in the underworld, is placed under the supervision of the sun god.

III. New rulers, new beliefs ? Osiris and the dead during the transition from the Late Period to the Ptolemaic Period

25The title of the third lecture was ‘New rulers, new beliefs ? Osiris and the dead during the transition from the Late Period to the Ptolemaic Period’. This focused upon the consequences of the establishment of the Ptolemaic dynasty at the end of the fourth century BC. Two questions were investigated : did the change from native Egyptian rule to Greek rule have an impact on Egyptian ideas about the afterlife, and did it have an impact on the way in which the Egyptians conceived of Osiris and his relationship with the deceased ? In connection with the second question, particular attention was devoted to the introduction of a new god, Sarapis, who was sometimes identified with Osiris. Did this influence ideas about Osiris and the dead and, if so, how ?

26A survey of the written and material evidence for Egyptian beliefs about the afterlife in the period of transition between the Late Period and the Ptolemaic Period revealed that no significant changes occurred in this sphere. Osiris was still regarded as the ruler of the underworld. The judgement of the dead remained an essential rite of passage. Mummification and its attendant rituals continued to be important for the posthumous well-being of all. The deceased’s transition from this life to the next was supposed to be effected by means of sakhu, or glorification spells, just as before. Egyptian conceptions of the various aspects in which the dead were supposed to continue their existence in the next world, akh, ba, and ka, stayed the same as well. There was also a significant amount of continuity in terms of the texts used to ensure a happy afterlife for the dead. Several works composed prior to the beginning of the Ptolemaic Period remained in use. Although a large number of new texts were introduced at this time, many of these were clearly based on earlier models.

27Broadly speaking, such continuity characterises the material record as well. In terms of how tombs were provisioned to supply the needs of the dead in the next world, there was little change, if any, between the Late Period and the Ptolemaic Period, so much so that it is sometimes difficult or impossible to determine whether a particular burial belongs to the end of the former or the beginning of the latter. In the sphere of funerary art, some innovations do begin to appear, for example, the dead are sometimes depicted in a more naturalistic way, or with non-Egyptian modes of dress and adornment. But these do not represent a new way of conceptualising life after death so much as the adoption of new artistic conventions to express an older symbolism. How an individual chose to be portrayed on items of burial equipment (or how relatives chose to have that individual portrayed) had nothing to do with the nature of his or her hopes and expectations for the next world.

28The lecture went on to examine three specific things that some have interpreted as new developments in the conception of the relationship between Osiris and the deceased in the Ptolemaic Period. The first is the practice whereby deceased individuals are referred to as Wsỉr n NN, ‘Osiris of NN’, with a genitival adjective inserted between the name of the god and that of the dead person, instead of the form Wsỉr NN found in earlier periods. Some have argued that this reflects belief in a closer connection between the god and the deceased than before, which arose under Greek influence. But this is incorrect, since the designation Wsỉr n NN was in use long before the Greeks began arriving in Egypt. In the lecture, evidence was presented to show that even in cases where the name of the deceased follows immediately after that of the deity, the meaning is still ‘Osiris of NN’, expressed with the direct rather than the indirect genitive. As noted in the first lecture, this designation identifies the deceased as a follower of Osiris in the afterlife. The ‘Osiris’ of a person is the Osirian form that he or she acquires as a result of the rites of mummification and justification. The Osirian form of each individual is unique, just as each individual is unique. Thus we could say that the element ‘Osiris’ in the locution ‘Osiris of NN’, whether written with or without the n, is not so much a name as it is a signifier of a particular class or category of things. As such, it can be preceded by the definite article in periods when that article was in use.

29Another phenomenon identified by some as an innovation of the Ptolemaic Period is the practice of writing the name of the goddess Hathor instead of that of Osiris in front of the name of the deceased. It has been argued that this may have been introduced as a result of Hellenistic influence. However, the designation ‘Hathor of NN’ was already in use during the Third Intermediate Period, centuries before the arrival of the first Greeks in Egypt, so this cannot be the case. According to one view, the use of this designation reflected a desire to create a closer relationship between deceased and divinity. Linking a deceased woman with a goddess rather than a god supposedly made it possible to avoid the barrier that gender imposed between her and the male deity. Another view is that this was done to maintain in the afterlife the gendered role, identity, and body that individuals assumed in Egyptian society. However, neither explanation accounts for the fact that ‘Hathor of NN’ never totally replaced ‘Osiris of NN’ as a designation for deceased women. Both continued to be used side by side, in some instances even interchangeably within the same manuscript to refer to the same person. The real function of the former was to identify the deceased as followers of Hathor, just as the latter identified them as followers of Osiris. One could be a follower or devotee of more than one divinity, which explains why the two statuses were not mutually exclusive.

30The third new development examined in the lecture was the introduction of the god Sarapis by Ptolemy I. This deity was sometimes identified with Osiris. Some think that, as a result of this identification, Osiris acquired new attributes and characteristics which had an impact on how his relationship with the deceased was envisaged. In the lecture it was noted that most of the Ptolemaic evidence for the identification of Sarapis with Osiris comes from the middle or late Ptolemaic Period, not the early part. In fact, the bulk of the evidence actually comes from the Roman Period, and only then do we find for the first time texts that record what the dead wished for or expected from Osiris in this form. In such texts, there is an emphasis upon Osiris/Sarapis as lord and ruler, an arbiter of conflicts who punishes enemies and provides aid in time of need. We no longer see a passive Osiris, one who needs to be protected from his foes, but an active deity, a king who exerts himself in defence of others. Since Sarapis is himself a royal god, some think this image of Osiris and what he was expected to do for the deceased was influenced by his association with the Greek divinity.

31In the lecture, this view was considered and rejected in favour of a different explanation. As noted by scholars like Laurent Coulon and Didier Devauchelle, the concept of Osiris as a royal god became increasingly prominent during the first millennium BC, even before the Ptolemaic Period, and there was a growing tendency at this time to portray him not just as the ruler of the underworld but the ruler of this one as well. Connected with this was an increased tendency to see in Osiris as much a god of the living as of the dead. This period also witnessed the rise of popular cults of Osiris in which he was venerated with epithets like ‘the saviour’, ‘the one who rescues the miserable’, and ‘he who saves his servant in the underworld’. Thus, it was not so much that in the Ptolemaic Period Osiris was identified with Sarapis and acquired new powers and attributes as a result. It was rather that the new image of Osiris which arose in the first millennium BC, itself building upon earlier ideas, and the powers and roles that were attributed to him as a consequence, facilitated and led to his identification with Sarapis. To be sure, in fostering the association between Osiris and Sarapis as symbols of kingship, the Ptolemies were exploiting this for political purposes, but as Coulon has pointed out, many earlier rulers of the first millennium did the same.

32The conclusion of the lecture emphasised four points. The first was the need to look at all the evidence pertaining to a particular phenomenon. When we fail to do so, we may overlook precedents for what otherwise appears to be an innovation. The designation Wsỉr n NN, ‘Osiris of NN’, is a good illustration of this. Far from being an invention of the Ptolemaic Period, it can be traced back as far as the Coffin Texts of the Middle Kingdom. The second point was that political change does not invariably bring about change in other spheres. As we saw, the imposition of Ptolemaic rule over Egypt does not seem to have resulted in fundamental changes in Egyptian conceptions of the afterlife. The third point was that we should not think of the Graeco-Roman Period as a cultural unity. There were significant differences between its two halves, not least in the sphere of religion. Thus it was only in the Roman Period that the popularity of Sarapis reached its peak, even though the cult of the god had been established centuries earlier when the Ptolemies first took control of Egypt. This serves nicely to illustrate the fourth and final point, which is that the ability of the state to influence developments in the religious sphere was limited. Personal names derived from or incorporating that of Sarapis enjoyed their greatest vogue in the second century AD, to judge from the onomastic record. Fewer people bearing such names are attested from the third century BC than they are from the seventh century AD, when Christianity was the dominant religion in Egypt. From this we may conclude that introducing the cult of Sarapis was one thing ; persuading the populace to venerate him was quite another.

IV. Where is the king of the two lands ?
The end of belief in the Osirian afterlife

33The title of the fourth lecture was ‘Where is the king of the two lands ? The end of belief in the Osirian afterlife’. This traced the disappearance of Egyptian belief in Osiris’s role as god of the dead and guarantor of their posthumous existence. The lecture began with some general remarks on method and approach. It was argued that the end of belief in the Osirian afterlife should be investigated in the context of the end of traditional Egyptian religion as a whole, but we should remember that the two are not identical. Not all features of Egyptian religion came to an end at the same time. Some survived longer than others. Equally, we should be prepared to separate belief in the Osirian afterlife from belief in Osiris himself. It is possible that the latter could have persisted after the former had died out.

34Two models for the end of traditional Egyptian religion were discussed and compared. In one, the conflict and triumph model, temples and other venues for traditional cultic activity continued in operation until destroyed or turned into churches. In the other, the gradual end model, temples fell out of use over a period of time ; some may have been abandoned and stood empty for a while before Christianity became dominant. Re-use or destruction was not necessarily motivated by religious feelings. Some temples were converted for secular purposes or exploited as sources of building material. Proponents of the first model rely heavily upon literary sources, most of them later than the period they purport to describe, as evidence to support their view. Proponents of the second model give more weight to contemporary documentary and archaeological evidence. Thus, in assessing which model is superior, one first needs to establish which type of evidence is more reliable.

35The lecture focused upon the disappearance of belief in Osiris in three places : Akhmim, Philae, and Abydos. This was done partly for practical reasons, since there was insufficient time to investigate the phenomenon in every part of Egypt. But it was also done to emphasise the importance of looking closely at local and regional developments. We should not assume that belief in the Osirian afterlife disappeared at the same time in every part of Egypt. It may have survived longer in some places than in others. Each city or town had its own distinctive local pattern of religious development that could have had a bearing upon this. The three places selected for investigation in the lecture are all sites from which abundant evidence pertaining to the cult of Osiris has been recovered, and this evidence has been the subject of considerable discussion. Thus they are well-suited for use as case studies.

36It was acknowledged that an investigation of this nature is subject to certain limitations. It is impossible to determine with certainty when belief in the Osirian afterlife actually came to an end in a particular place. The most that we can do is to establish when the evidence for it comes to an end. It remains possible that the belief itself persisted even longer, but we cannot track it in the historical record. Furthermore, evidence for belief in the Osirian afterlife is just that. It tells us that the belief in question existed in a particular place at a particular time, but it does not tell us how widespread that belief was or what portion of the population in a given area held it.

37The first place to be investigated was Akhmim and its environs. The most important local cults were identified and described. Their history was briefly traced. Particular attention was given to the evidence for religious life and the forms it took in the Graeco-Roman Period. Osiris was a very important deity at this time. Numerous texts attest to belief in him and the benefits that the deceased hoped to enjoy in the afterlife as members of his following. Such texts are still abundant in the second and third centuries AD, but not thereafter. The latest precisely dated text that tells us anything about its beneficiary’s hopes for the afterlife was inscribed on 24 February 268 AD. This records the wish that the woman for whom it was written live before Osiris. Thus evidence for belief in the Osirian afterlife at Akhmim ceases after the third quarter of the third century AD.

38This does not mean that all polytheistic belief in the Akhmim region died out. Greek documentary texts of the fourth century AD mention temples, priests, and priestesses at Akhmim, but most of these served Greek cults rather than Egyptian. The writings of the archimandrite Shenute (c. 347-465 AD), an implacable foe of all non-Christian forms of religion, contain many references to the beliefs and practices of his religious opponents in the Akhmim region. But there is no trace in his accounts of the very distinctive features of Egyptian cult and worship that characterised this area in earlier periods. In fact, there is very little in his descriptions that could be called Egyptian at all. When he cites the myths of his opponents, they are invariably Greek ones. The deities to whom he refers are always given Greek names, the Egyptian equivalents being added as glosses if at all. One has the impression that by Shenute’s time, all that survived of the old indigenous religion was a few ideas lingering on in the debased form of popular superstition and more or less subsumed within the broader framework of the dominant Hellenism. Whatever the case, there is no evidence that Shenute knew anything of Osiris or his cult, which supports the idea that belief in that god had already ceased in the Akhmim region by the time his career began.

39The second place to be investigated in the lecture was Philae. The religious history of the island was summarised and the most important cults described. As with Akhmim, particular attention was devoted to local forms of cultic activity during the Graeco-Roman Period. Since Isis was the main deity worshipped at Philae, it is not surprising that Osiris, as her brother and husband, played an important part in religious life there as well. The most notable features of his cult were described. Important information about this is provided by hymns and other texts inscribed on the walls of the temples on the island, including graffiti left by priests who were employed there and pilgrims and other visitors to the site.

40The graffiti, written in demotic or Greek, are especially valuable for what they tell us about the cult of Osiris at Philae during its final centuries of existence. The picture they present is one of gradual contraction. The latest precisely dated graffito from Philae that mentions Osiris is a Greek one written on 20 December 452. In it the writer and his brother give thanks to their mistress Isis and their master Osiris. Many of the writers of the latest graffiti from the island were members of the same family, in whom responsibility for maintaining the cult was invested, and one wonders how far outside their circle worship of deities like Isis and Osiris extended. Some think that they functioned mainly as priests for tribes living to the south of the Egyptian border, who retained the old faith longer than the Egyptians themselves did.

41The graffiti at Philae suggest that worship of Osiris and other traditional Egyptian deities died out there sometime around the middle of the fifth century AD. Literary sources, all of them composed long after this time, present a different picture. The Coptic Life of Aaron dates the triumph of Christianity on the island to the fourth century AD, which is patently untrue. An account written by the Byzantine historian Procopius claims that the temples of Philae continued to function until sometime between 535 and 537 AD, at which time they were destroyed by order of the emperor Justinian. This account is equally untrustworthy, since the temples on the island remain standing to this day. Finally, a petition written by Dioscorus of Aphrodite in 567 AD accuses someone of committing a long series of crimes. These include consecrating pagan shrines and statues, and encouraging Nubian tribes who had converted to Christianity to apostasise. Most of these accusations are probably rhetorical. Some think the shrines and statues mentioned in the petition were those on the island of Philae, but this is only surmise. Nor is there any suggestion that the activities of which the plaintiff is accused included the veneration of Osiris. It may well be that belief in traditional Egyptian deities survived longer in Nubia than it did in Egypt itself, but there is no compelling evidence to suggest that it persisted as late as the sixth century AD.

42The final place to be investigated in the lecture was Abydos. Osiris was already associated with that city in the Old Kingdom, but it is not until later that we find clear evidence of a connection between the god and the deceased there. Osiris is supposed to have buried at a place called Poker, which is thought to be the predynastic royal cemetery of Umm el-Qaab, about 1.7 km away from the cultivated area at Abydos. The tomb that was identified as his there was originally that of the first dynasty ruler Djer. An important feature of the rites celebrated at Abydos was a procession from the temple of Osiris to the god’s tomb at Poker, attested from the twelfth dynasty onwards, in which the deceased hoped to participate. The cult of Osiris continued to be of major significance into the Roman Period, although there is evidence that by this time cultic activity at Umm el-Qaab had ceased. The god’s tomb was now believed to lie elsewhere, probably in the Osirieon behind the temple of Seti I, and this supplanted Umm el-Qaab as the destination of the processions held in his honour.

43The latest evidence for belief in Osiris as god of the dead at Abydos is provided by funerary stelae of the first or second century AD. The last precisely dated stela attesting to his veneration was inscribed in 75 AD. Some graffiti left by visitors to the temple of Seti I invoke Osiris, or more often Sarapis, but the dates of these are uncertain. No precisely dated graffito from the Roman Period mentions either god. Thus solid evidence for belief in the relationship between Osiris and the deceased at Abydos ceases in the middle of the second century AD, if not before. A particularly interesting development is that Osiris and, after him, Sarapis, are eventually replaced by Bes as the most important divinity at Abydos. The popularity of that deity is abundantly attested by later Greek graffiti from the temple of Seti I. He was especially important as an oracular god, in which capacity he became internationally famous. The oracle of Bes at Abydos was closed down by imperial order in 359 AD. As far as we can judge, however, the cult of Osiris at Abydos had come to an end long before this.

44Some texts postdating the second century AD, although not from Abydos, do make reference to Osiris in connection with that city. These include mummy labels of the late third century AD from Akhmim and magical papyri from Thebes dating to the third, fourth and possibly the fifth centuries. Some, for example, Ute Effland, Julia Budka, and Andreas Effland, have maintained that the magical papyri attest to the continued existence of the cult of Osiris at Abydos at the time when they were written. In the lecture, the evidence that has been cited in support of this view was scrutinised and judged to be unconvincing.

45A separate question is when traditional Egyptian religion as a whole came to an end in the region of Abydos. Some have claimed that it survived there as late as the sixth century AD, based on two passages from a later source, the biography of Moses of Abydos, the head of a local monastery who lived at that time. In the first passage, Moses and his followers go to the temple of Seti I to exorcise a demon called Bes who has taken up residence there and is afflicting those who pass by with various illnesses. Although the passage preserves a folk memory of a once-prominent local deity, it gives no indication that Bes was still the object of worship or veneration. Rather the populace regard him as a type of pest or vermin, and call in the monks to eradicate him.

46The second passage describes the destruction of a temple of Apollo. This is caused by a miraculous earthquake brought on by the prayers of Moses and his fellow monks. Several features of this passage discredit it as a historical source. One is the supernatural elements it contains, the miraculous earthquake and the intervention of an angel who appears and encourages Moses. Another is the stereotyped literary topos it presents, that of a monk who duplicates a feat attributed to a famous figure from the Bible, in this case causing the destruction of a pagan sanctuary through prayer. Yet another is the vagueness of the information the passage provides about the location of the temple. It is said to be somewhere in the toou, or ‘mountain’, of Abydos. Contrary to Effland, Budka, and Effland, who think that this term designates Umm el-Qaab, it actually refers to the entire stretch of arid land between the cultivation to the east and the high desert to the west. Furthermore, the temple is not that of any of the major gods of Abydos ; rather it is dedicated to a Greek deity, Apollo. Whoever wrote the description of the temple and its priesthood had no knowledge of how a real Egyptian temple functioned. Evidently, he confused it with a church, which was probably the only kind of place of worship with which he was familiar. So this passage does not provide trustworthy evidence for the continuation of a pagan cult at Abydos in the time of Moses.

47The three case studies presented in the lecture make it clear that belief in Osiris did not end at the same time in all parts of Egypt. At Akhmim, evidence for it stops in the third century AD, at Philae in the fifth century AD, and at Abydos in the first or second century AD. In the case of Philae, although the cult of Osiris continued to function until the middle of the fifth century, we have no evidence that belief in his relationship with the deceased persisted until that date. We know nothing about the afterlife expectations of the last servants of Osiris on the island because, unlike the temples whose walls preserve their graffiti, their tombs have not survived.

48These case studies serve to illustrate how one should approach the problem of trying to discover when the cult of Osiris or any other Egyptian deity came to an end. The lecture concluded by proposing the following nine guidelines for such an undertaking.

491. We should proceed on a regional or local basis.

502. We should start by familiarising ourselves with the evidence relating to the specific region we are investigating. What forms did local religion take, what was religious life like there, who were the most important deities and how were they worshipped, especially in the Graeco-Roman Period ?

513. We should be sensitive to the chronological distribution of the evidence for such religious activities, noting in particular when it runs out or comes to an end.

524. In some cases, when there is no other evidence, this may be as much as we can do. In other cases, we may also need to take into account later sources, for example, Coptic hagiographies and Greek literary or documentary texts.

535. We need to evaluate the latter critically. We should be cautious about using later material, especially Coptic sources like the writings of Shenute and the Life of Moses of Abydos as evidence, bearing in mind that these are literary, often highly rhetorical sources, and taking into account the specific purposes for which they were written.

546. We should ask whether the writers of such sources record accurate information about the cults and monuments in the places they describe. Do they portray local traditions correctly or not ?

557. We should consider whether or not these accounts incorporate supernatural elements, for example, angelic interventions and other miracles.

568. We should note whether there is a significant chronological gap separating sources of this type relating to a particular locality from the latest contemporary evidence for traditional cultic activity there. If so, how can one account for such a gap ?

579. If a source displays no awareness of local traditions, contains demonstrable inaccuracies, incorporates supernatural elements, or is set in a period long after all first-hand evidence for the sort of religious activities it purports to describe has ceased and one cannot account for the chronological gap between it and them, then it is not a reliable source of historical information.

Haut de page

Notes

1   The material presented in them is a condensed version of selected chapters from the speaker’s forthcoming book on the relationship between Osiris and the deceased in ancient Egypt, to be published by Oxford University Press.

Haut de page

Pour citer cet article

Référence papier

Mark Smith, « Osiris and the Deceased in Ancient Egypt :
Perspectives from Four Millennia
 »
Annuaire de l'École pratique des hautes études (EPHE), Section des sciences religieuses, 121 | 2014, 87-102.

Référence électronique

Mark Smith, « Osiris and the Deceased in Ancient Egypt :
Perspectives from Four Millennia
 »
Annuaire de l'École pratique des hautes études (EPHE), Section des sciences religieuses [En ligne], 121 | 2014, mis en ligne le 19 novembre 2014, consulté le 18 avril 2024. URL : http://journals.openedition.org/asr/1224 ; DOI : https://doi.org/10.4000/asr.1224

Haut de page

Auteur

Mark Smith

Directeur d’études invité, École pratique des hautes études – Section des sciences religieuses, Université d’Oxford, Grande-Bretagne

Haut de page

Droits d’auteur

Le texte et les autres éléments (illustrations, fichiers annexes importés), sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search