David M. Rohl: Pharaohs and Kings (Crown (New York), 1996) $37.50
Articles in the Journal of the Ancient Chronology Forum (JACF):1
– The Early Third Intermediate Period, JACF 3 (1989/90) 45
– A Test of Time, JACF 5 (1991/92) 30
(with M. Ibrahim) Apis and the Serapeum, JACF 2 (1988) 6
The major features of Egyptian chronology as it is generally accepted today were rapidly established after Champollion’s decipherment of hieroglyphics. The surviving fragments of the Egyptian historian Manetho supplied a kinglist, comprised of 30 dynasties, which was supplemented and corrected from several similar ancient Egyptian lists, such as the Royal Turin Papyrus, and the kinglists of Abydos and Saqqara. The monuments of the individual kings could be placed within this framework to provide much fine detail of individual reign-lengths, periods of coregency etc. The relative chronology so established was far from complete, but could apparently be tied to absolute dates at several key points, from several astronomical synchronisms, and from synchronisms between the Egyptian sequence of kings and similarly constructed sequences for other countries of the ancient Near East. These included the kings of Israel, through the identification of the biblical Shishak as Shoshenq I of the 22nd dynasty, and the kings of ancient Assyria and Babylon.
For most of this century, work in the field has consisted of refining this model and trying to fill in the missing pieces. For example, there was a debate in the 1980s on the location for the sightings of the heliacal rising of Sothis (Sirius), which is significant because the answer affects the dates of certain recorded observations of this event. This debate resulted partly in proposals for lowering both the end of the Middle Kingdom and the start of the New Kingdom by three decades, and partly in a rejection by many of the validity of the claimed astronomical synchronism which had previously governed the date of the start of the New Kingdom. As a result of these fine tunings, the date for the start of the Egyptian New Kingdom is now set between 1570 BC and 1539 BC, with the lower date being preferred; it may yet drift down a few decades. Similar analyses have moved the accession of Ramses II from 1304 first to 1290 and then to 1279 BC. However, while such debates are intensely fought by the specialists involved, the underlying sequence of events has never been revised: the 18th, 19th, 20th, 21st and 22nd dynasties reigned in succession, while the 23rd, 24th and early 25th dynasties overlapped the last part of the 22nd.
Recently, this basic model has come under challenge from two directions.
With the advent of C dating, refined by dendrochronological (tree-ring) sequences, it is now theoretically possible to provide highly accurate dates for archaeological samples which are not dependent on any historical data. By and large, these techniques have been held to confirm historically-based Egyptian dating, although dendrochronological dates have in fact tended to be older than the historical evidence suggests. The date assigned to the explosive eruption of the Aegean island of Thera (Santorini) by these techniques has created a particularly severe problem in this regard. According to the currently favoured scientific analysis, this eruption occurred in 1628 BC. A recently published Anatolian tree ring sequence supports this date, while also supporting the conventional dating of the Amarna period of the 18th dynasty in the latter half of the 14th century BC.14 However, there are clear traces of the Thera eruption in levels dated to the early 18th dynasty at the Hyksos capital of Avaris, Tell el-Daba, and elsewhere. According to conventional archaeological and historical measures, this places Thera in the late 16th century BC, a century after the date preferred by the dendrochronological evidence. If Thera occurred in 1628 then there is a fundamental error in our reconstruction of the 18th dynasty. Since this is one of the best documented periods in Egyptian history, there seems to be no possibility of finding an extra century in the historical record, as the 1628 date would require; but nor does the dendrochronological record admit a major volcanic event in the 16th century BC.
This controversy has also, so far, largely been confined to the academic establishment. A far deeper challenge to conventional Egyptian chronology has been mounted in the public arena. This school of thought has attacked the conventional chronology from the other direction: it argues for a radical collapse of later Egyptian chronology by up to 350 years.
The first salvo was fired in 1991, with the publication of Centuries of Darkness.2 The authors pointed out the existence of archaeological dark ages throughout the Old World in the early Iron Age, noted that Old World chronologies were ultimately derived from Egyptian chronology, and argued that Egyptian chronology must be fundamentally flawed. They proposed a compression of Egyptian chronology for the Third Intermediate Period (dynasties 21-25) of about 250 years. Such a compression would have the effect of eliminating the dark ages that were felt to be so troublesome. In a special review issue of the Cambridge Archaeological Journal these proposals were roundly rejected by experts in all disciplines in Old World archaeology, a result virtually assured by the failure of the authors to present more than an outline restructuring for Egyptian chronology.
Now, in this beautifully produced and lavishly illustrated book, and in the accompanying three-part television series, David Rohl has returned to the charge, this time providing many of the details of Egyptian chronology missing from the earlier book, and focussing on their implications for locating Biblical narratives in historical and archaeological contexts. While some of his proposals for a revised chronology are very similar to those of Centuries of Darkness (which is not surprising, since Rohl and James were initially collaborators), he also advances many chronological proposals not made by the earlier book. These revisions have some dramatic results for Biblical history. Rohl finds it possible to identify references to many Biblical characters in the El-Amarna letters, including David and Saul. Moreover, he presents proposals for identifying the stories of Joseph and Moses in the Egyptian record: Joseph becomes a vizier to Amenemhat III, with an Israelite settlement around his palace at Area F of Tell el-Daba; Moses spent his youth fighting for Sebekhotep IV; the catastrophe which overwhelmed Level G at Tell el-Daba is to be identified with the biblical Plagues of Egypt; and Exodus is also dated to this time, with the Israelites, fleeing Egypt, crossing paths with the invading Hyksos.
Sensational claims such as these drive professional Egyptologists crazy. And with good reason. The popular bookshops and the cable channels are filled every year with works of the likes of Robert Bauval and Graham Hancock, who jump to the most spectacular and far-reaching conclusions, frequently based on the slenderest of evidence, much of it misunderstood, and usually accompanied by swingeing rhetorical attacks on the hidebound Egyptological establishment who are held to be incapable of seeing the obvious wisdom of the new theories because of their purblind and musty academic vision. In such circumstances, it is no wonder that the automatic reaction of many professionals is to reject the latest popular radicalism out of hand. Simply by launching their theories through the popular media, Rohl and James have made themselves and their theories appear to be no different from any of these others.
Moreover, neither James nor Rohl have helped themselves by their associations with the disciples of that arch-heresiarch of yesteryear Immanual Velikovsky. Even though both deny Velikovsky's theories, both have found congenial intellectual homes in Velikovskian groups and Velikovskian publications, and Rohl has accepted some of Velikovsky's observations in formulating his theses on Moses and Joseph. This sin is compounded by the fact that the "New Chronology" is a direct attack on Kenneth Kitchen's The Third Intermediate Period in Egypt,3 which is widely regarded as one of the major intellectual achievements of modern Egyptology. It is not surprising, then, that Kitchen, in a savage review in the Times Literary Supplement,4 condemned the authors of Centuries of Darkness as "sons of Velikovsky" and consigned them to the same oblivion as their accursed master. No doubt the authorities at the British Museum, who have banned this book from the BM Bookstore, feel much the same way about David Rohl.
Such a reaction is misguided. The message may be sensational, and Rohl may even be completely wrong in his theories, but there is a world of difference between his intellectual standing and that of Velikovsky, or even Peter James. Rohl is a trained Egyptologist, with a degree in the subject and a PhD in course of preparation at University College London, one of the most prestigious Egyptological departments in Britain. Nor did this book appear out of the blue. He has for several years edited and published the Journal of the Ancient Chronology Forum, which has presented not only his own theories, but also articles of considerable merit by many mainstream archaeologists. It is evident, from the pages in this book, from the articles in JACF, and from personal exchanges, that Rohl has a considerable mastery of his material, and has thought long and deep about it. Even if the debate is ultimately not resolved in his favour, it is worthwhile because it forces a re-examination of long-held assumptions and of difficult problems. This can only be healthy for the discipline, and should be welcomed for this reason.
Having said that, there remains a significant obstacle in undertaking a full-scale critical review of the New Chronology. Although Rohl has gone considerably further than James in detailing his proposed chronological revisions, particularly for dynasties 21 and 22, there are many aspects of Egyptian chronology for which he has not yet published a new model, either in this book or in his JACF articles.
· The late Third Intermediate Period – dynasties 23 to 25 – is barely discussed. It appears that the Manethonian dynasty 23 is seen as a Delta dynasty succeeding the 22nd and lasting until quite late in the reign of Psamtek I. In terms of overall chronology, this is perhaps a moot issue, since dynasty 23 is seen by all parties as being contemporary with existing dynasties. Indeed a very similar model has been proposed by Anthony Leahy, without, however, requiring that earlier chronology be collapsed.5
· The keystone of Rohl's chronology is to identify the campaign of the biblical Shishak in year 5 of Rehoboam of Judah, not with the campaign undertaken by Shoshenq I late in his reign, but with one recorded in year 8 of Ramses II. This places the accession of Ramses II at c932 BC,6 while, on other grounds, Shoshenq I is placed at c823 BC. Under this scheme, it is necessary not only to make the 21st and 22nd dynasties contemporary but also to reduce the duration of the 19th and 20th dynasties after Seti I from about 200 years to, at most, a little over a century. Since Ramses II reigned 66 years, Ramses III reigned 31 years, and Ramses XI reigned 28 years, making 125 years before any of the other 13 Ramessid kings are even considered, this would be a remarkable feat. However, the only vague hint Rohl has published as to how it is to be accomplished is a suggestion that Ramses IX and Ramses XI might have been contemporaries,7 which is clearly insufficient. Perhaps in recognition of the difficulty of this problem, several other "New Chronology" solutions have been proposed for Shishak's campaign, including year 53 of Ramses II, Ramses III or even Ramses IX.8
· In order to lower the accession of Ahmes from 1539 to 1193 (346 years), while only lowering the death of Amenemhat III from the current lowest date of 1772 to 1645 (127 years) it is necessary to extend the conventional length of the Second Intermediate Period by 219 years. Part of this – some 50 years – is done by giving early 13th dynasty kings average reigns of 5 years, about twice the length normally assumed. An uncertain amount, perhaps 10-20 years, comes from extending the reign of Sebekhotep IV beyond the decade normally ascribed to him. However, the remaining 150 years or so can only come after the reign of Dudimose – that is, at the end of the 13th and in the 17th dynasties. Although Rohl is a close student of the excavations at Tell el-Daba, he has not discussed any of the Egyptian historical data on the chronological issues surrounding this period. From my own studies of this subject, I have recently proposed a chronology for exactly this period which fits well with conventional low model for the New Kingdom and the middle or high models (but not the low) for the Middle Kingdom (i.e. I would set the length of the Second Intermediate Period at about 250 years).9 The synchronisms supplied by the data are, in my opinion, loose enough that this chronology will support a reduction of perhaps a decade or two in the late Second Intermediate Period, and an extension of perhaps four decades. But an extension of a century and a half is simply out of the question.
In the book, Rohl excuses such omissions by the need to conform with the publication regulations surrounding London University PhDs. In a posting to the UseNet group sci.archaeology, he amplified this point, noting that he was not informed of these regulations until his book and television deals were well advanced. In effect, he argued that he was trapped in a situation where he was forced to publish his book before his academic arguments could be properly examined. Despite this experience, he has now taken another leave of absence from his PhD to write a second book and a third is also planned. It appears unlikely that these analyses will see the light of day for some time.
Perhaps the most far-reaching claim in the book, made also by Peter James, is that the current standard chronology leads to contradictions in interpreting the historical and archaeological evidence which are so severe that this chronology must be regarded as fatally flawed, and discarded. This is a very serious claim: it says that, regardless of whether Rohl's own proposals are correct, some alternative scheme must be found to the current chronology. These arguments require review.
Three chronological anomalies are claimed.
The first concerns the Apis Bulls in the Serapeum at Memphis. It is well established for the securely-dated Saite and Ptolemaic periods that the average lifetime of an Apis Bull is 18 years. However, for the period between the first Apis of Ramses II and that of Psamtek I (606 years in conventional chronology) the evidence only permits at most 23 Apis Bulls to be recognised – sufficient for 396 years. There are no signs of any bulls for the 21st dynasty, nor indeed for the early reigns of the 22nd.
One of Kitchen's major criticisms of Centuries of Darkness is the alleged aversion of the authors to gaps in the evidence; indeed he cites this particular argument as a prime example.10 If it were only a case of missing Apis Bulls, he would have a point. Mariette found evidence for one hidden Apis burial of at least three bulls, not yet fully excavated; there may be others. What is much more troubling is that there is no apparent sign of any activity at all in the Serapeum during this period, despite high levels of activity in both the preceding and succeeding periods.
This is certainly a phenomenon requiring explanation, and Kitchen is wrong simply to brush it aside. However, precisely because it is a gap in the record, it is not necessarily evidence of a fatal chronological flaw. There may once have been evidence of 21st dynasty activity. Of approximately 1200 stelae discovered in the Serapeum, only 268 have been published, and nearly 400 stelae, fully a third of those found, were destroyed by floods in the Bulaq museum in Cairo last century. Such evidence may yet be found, either in the basement of the Louvre, or in the Serapeum itself. The Apis burials may have been moved for a century or so, to a new set of chambers elsewhere in Saqqara, as David Aston has suggested,11 or even further. It may be relevant that the first surviving post-Ramessid Apis is dated to year 23 of Osorkon II, and that the first known royal tomb of the 22nd Dynasty is that which his mother queen Kapes built for him and his father Takeloth I.
The second "anomaly" concerns the famous cache of royal mummies found in tomb TT320. Dockets on the mummy cases of Seti I and Ramses II, which are known to have been found near the entrance of TT320, describe a movement of these coffins out of the tomb of Seti I on 17 Peret IV in year 10 of Siamun of the 21st dynasty, and a second movement involving the kai ("high place" or "high track") of queen Inhapi, on 20 Peret IV, which is the same day that an inscription at the entrance of TT320 records the burial of the High Priest of Amun Pinudjem II. However, the mummy linen of the Second Prophet of Amun Djedptahefankh A is dated to year 11 of Shoshenq I. Since his coffin rested deep in the tomb and it is physically impossible to get it past the coffin of Seti I, Rohl concludes that Shoshenq I reigned before Siamun – conventionally, their positions are reversed.
An alternative, and chronologically far less drastic, solution to this problem has been proposed by Nicholas Reeves.12 Reeves pointed out that TT320 cannot be the kai of Inhapi, because Inhapi herself was found near the entrance to TT320. Thus, the Djedptahefankh linen shows that the movement from the kai to TT320 must have occurred at some undocumented date after year 11 of Shoshenq I. Indeed, after initially denying it,13 Rohl now accepts that TT320 is not the kai, though he still places great emphasis on the physical difficulties of manoeuvring the coffins. The true crux of his case is that he now interprets the dockets to mean that the coffins rested in the kai for only three days while awaiting interment in TT320, on the day of Pinudjem II's burial. However, this interpretation is at variance with that accepted by all other Egyptologists since Cerny,14 namely that it took the authors of the dockets three days to move the mummies from the tomb of Seti I to the kai, and that the authors thought of the kai as the final resting place, not an interim staging post. Rohl has not justified his alternative interpretation (nor even acknowledged that it is one).
In the book, Rohl gives no recognition to Reeves' proposal. Elsewhere, he has objected to it on the grounds that it requires too many assumptions. In particular, he points to the lack of a docket for any movement after 20 Peret IV, the date of the burial of Pinudjem II. However, neither Inhapi nor Amenhotep I, who were moved to TT320 on the same day as the other kings under all hypotheses, have dockets for 20 Peret IV, indicating that they were not moved on that day. It is also to be noted that Reeves' assumptions, be they excessive or not (and, in this reviewer's opinion, they are not) are entirely local in scope. When the alternative proposal on the table is to jettison the existing chronology, one may wonder who is making the larger leap of faith!
The third "anomaly" relates primarily to the northern wall of the tomb of Osorkon II at Tanis. For part of its length, this wall is immediately adjacent to the southern wall of the tomb of Akheperre Pasebakhaenniut (conventionally, Psusennes I), to the north, which was conventionally built over a century earlier. At the points of contact, this wall is a single thickness; on both the east and west sides of these points, the wall is of double thickness, with the extra layer being placed on the outside. Thus, the southern wall of the tomb of Psusennes I appears to be inserted into part of the northern wall of the tomb of Osorkon II, with the outer thickness of the wall having being removed, for the length of the southern wall of Psusennes' tomb, in order to allow for the extension of this tomb to the south. Rohl's argument is that such a structure would ordinarily imply that the tomb of Psusennes was later than that of Osorkon, though in conventional chronology they are in reverse order. This observation caused the original excavators, Montet and Lézine, considerable difficulties. Their suggestion was that Osorkon's tomb was not original to him, and that its northern wall was rebuilt for him.
Of the three "anomalies" this is the one which has met with the most favourable response from professional archaeologists, many of whom apparently agree that no other clear explanation has been provided. Moreover, one has to agree with Rohl that the interpretation made by Montet and Lézine makes very little sense, since the inscriptions on the interior face of the northern wall of Osorkon's tomb, and other evidence, clearly show that the tomb was constructed for him by his mother.
Nevertheless, not all archaeologists agree with Rohl. Jean Yoyotte, who currently holds the concession for the site, argued, in an interview briefly excerpted in the television series which is not referred to in the book, that Rohl must wait for a new survey of the tomb interiors to be completed before drawing any final conclusions. Rohl himself has argued that the alterations made to the interior of Psusennes' tomb to accommodate the burial of his general Wendjebaendjed, resulting in the southward extension that abuts Osorkon's tomb, support his argument for a pre-existing tomb to the south. A detailed plan of the tomb interiors, based on Montet and Lézine's work, is reproduced as Figure 108 of Rohl's book, which allows this argument to be followed in detail. Yet a close examination of the interior plan of Osorkon's tomb as shown in this plan underscores Yoyotte's point.
It is clear that the tomb walls are indeed frequently of double thickness. This is so for every wall in Psusennes' tomb, and for the west wall and parts of the east and north walls of Osorkon's tomb. However, it is not true for the burial chamber of Takeloth I and other structures on the south side of Osorkon's tomb, nor is it true for any of the interior walls of this tomb. Moreover, except where the inner thickness is a lining of granite (notably, the north and south walls of Osorkon's own burial chamber), there is no reason to believe that the interface between inner and outer walls is aligned. If we take away the outer skin, as Rohl argues has been done here, then the new external face should follow an irregular line. But that is not what the Montet/Lézine plan shows. The northern wall of Osorkon's tomb is certainly of single thickness, where it abuts onto the southern face of Psusennes', but its outside face is shown to be aligned exactly as we would expect for a facing that was originally intended to be the exterior face of the tomb.
Now, it must be noted that the plan of Montet and Lézine is not fully accurate on this point: there are some signs of such dressing.15 In particular, the stones in Osorkon's tomb at the eastern end of the inset section show notching to accommodate the wall of Psusennes' tomb. But, was the pre-existing wall shaped to accommodate new stone or was the new stone shaped to fit a pre-existing wall? Common sense suggests the latter. In order to align an irregular intermediary interface once it has been exposed, one must first dress the stone along the length of that interface. This is an expensive activity, and not one likely to be undertaken if the result was immediately going to be hidden by another tomb with an immediately adjacent wall, unless absolutely necessary. In fact, no such adjustments are likely unless they impact the structure of the new wall. If it is necessary to make a fit between two such surfaces, it is at least as likely that the stone for the new structure, which is far more manoeuverable, will be dressed to fit to the existing shape. However, it will take very close examination to determine whether the dressing occurred during construction of Osorkon's tomb or at a later time.
When considering the hypothesis that an outer skin has been stripped away from the northern wall of Osorkon's tomb, it must also be noted that the double thickness at the western end of this wall is of a different character to most instances of this phenomenon in the tomb complex. The point at which Psusennes' tomb recedes to the north coincides with the midpoint on the interior north-south wall within Osorkon's tomb that marks the eastern boundary of his burial chamber (a coincidence, incidentally, which is difficult to explain if Psusennes' tomb were built last). The entire western section of the north wall is therefore the external face of the north wall of Osorkon's burial chamber. The interior thickness of this section is the granite lining of this chamber, and as noted above this is likely to be aligned on the interface to the exterior wall because it is a different type of stone. In a very real sense, virtually the entire exterior north wall is of single thickness stone. Only at the eastern end, near the entrance and past the end of the south wall of Psusennes' tomb, is there a projection from the north wall which breaks this pattern.
Thus, while I have no explanation for this projection, the part of the wall that actually adjoins Psusennes' tomb does not particularly support the theory that it is of earlier date.
There is another factor which Montet and Lézine were unable to consider. South of the main passageway in Osorkon's tomb lies the burial chamber of a king Takeloth (in which, incidentally, all walls are of a single thickness). Montet and Lézine considered this to be Takeloth II, who was later than Osorkon II, but it has since been established that this is the tomb of Osorkon's father, Takeloth I.16 Thus the joint tomb forms a single complex, built under the direction of Osorkon's mother queen Kapes, and of much more complex design than would be the case if each tomb had been built at different times. This being so, there may have been a change of plan necessitated by the presence of the more northern tomb.
Finally, one may note that the designs of the tombs of Osorkon II and Akheperre Psusennes are quite different from each other, both in their layout and their construction. Under Rohl's chronology, the two kings are contemporaries. Why is this not reflected in the architecture of their tombs?
I want to stress that the issue here is not whether Rohl's explanations or any particular counter-explanations for Rohl's three "anomalies" are correct. In my own view, only the burial of Djedptahefankh A in TT320 has received a satisfactory explanation under the "conventional" chronology. One can even concede, for the sake of argument, that Rohl's explanations may in fact be correct, reserving only the concession that they have not been proven. The point is that there exist "conventional" explanations which are reasonable and economical with their assumptions, while remaining consistent with the current chronology. The problems are all worthy of study and analysis, but neither alone nor together do they force us to solutions involving drastic chronological revisions. To this extent, the prejudice that much of the academic establishment has shown with regard to Rohl's theories is well-founded: just like other "revisionist" theorists, Rohl has leapt for a radical conclusion when it is far from clear that such solutions are necessary.
Nevertheless, the problems are real, and Rohl is right to point them out. A deeper study of them would certainly reveal valuable insights into the history of the time. Rohl's own activities indicate how this can happen. For example, the Serapeum creates problems for Rohl's own scheme: even if a reduction of 200 years were granted on the basis of the "missing" bulls, it would not meet the 350 required for his chronology. Rohl proposes to meet this problem by analysing some of the alleged Bulls out of existence on the grounds of insufficient evidence. Moreover, he must do so while simultaneously solving the problem of the Bulls of the early 22nd dynasty, which are missing under the New Chronology as well as under the conventional scheme (unless they be the three that Mariette saw but could not excavate). In view of the clear gap in the record, such a reanalysis has merit, regardless of whose chronology is correct, and regardless of the outcome for the sequence of Apis Bulls.
It is, of course, perfectly possible that the standard chronology is wrong, even though no fatal logical contradictions have as yet been found. Rohl correctly points out that the synchronisms on which the current chronology rests are few in number and are not without difficulties of interpretation. The dated Sothic sighting which was once held to fix New Kingdom chronology is, as we have seen, now widely discounted as not being a Sothic sighting at all.17 The lunar observations which date the reigns of Ramses II and Thutmosis III admit multiple solutions, repeated in a 25-year cycle. Assur-uballit of Assyria does have a different father in the Amarna letters (Assur-nadin-ahhe II) from that given to Assur-uballit I in the kinglists (Eriba-Adad I).18 The Palestinian campaign of Shoshenq I does not match well with the Judean campaign of Shishak described in the Book of Kings (though it must be said at once that Rohl's proposed identification of this as the Moabite campaign in year 8 of Ramses II is no better). And calibrated C dating, to the extent it can be applied in Egypt, does diverge from historically derived dates – though in the opposite direction to that preferred by Rohl.14 It is salutary to be reminded of these facts.
Having noted the limitations of the basic chronological synchronisms, Rohl has erected an alternative chronology based on a reinterpretation of much the same evidence as the standard chronology. In a brief review such as this it is not possible to critique all the arguments that are relevant to this thesis, ranging as they do from archaeoastronomy to Palestinian archaeology to dendrochronology to Biblical criticism to ancient linguistics to statistical analysis, even had I the necessary expertise, and even though some of these areas – such as the difficulties of applying the New Chronology to the Assyrian kinglist – present major, if not insuperable, objections to the theory. Since Rohl's arguments on Egyptian genealogy will be of particular interest to the readers of this journal, I will focus on this area for the remainder of this article.
Rohl makes use of genealogy for three purposes. First, he uses genealogies to estimate time, based on an assumed average inter-generational interval. Second, he argues that the genealogical data supports his case that Egyptian chronology must be compressed. And finally, his theories require that a number of genealogies held to support particular points of the conventional chronology must be reexamined to fit the New Chronology.
When using genealogies to estimate periods of time, Rohl uses an intergenerational average of 20 years. Two justifications are given for this number: the authority of its use by Kitchen and by Morris Bierbrier19; and an independent calculation supplied by Rohl based on the regnal statistics for a number of ancient dynasties. The first observation is somewhat misleading. While Kitchen and Bierbrier did indeed use this number for calculational purposes, Bierbrier concluded that an average generational interval of 25 years was much more appropriate. In view of this conclusion, it would be fair to use 25 years when estimating time from genealogies under the conventional chronology.
The second argument, presented in Rohl's Figure 407, is a remarkable piece of work. Essentially, he has taken 19 dynastic groups whose relative chronologies are held to be reasonably secure, totted up the total amount of time and the total number of kings, and divided one by the other to arrive at an average reign length of 16.75 years. This is then increased to 20 years "to allow for occasional usurpations and brother successions".
As an estimate of an average intergenerational interval, the number of methodological flaws in this analysis is almost too large to count. The dynasties listed include four (19th and 20th dynasties, early Assyrian kings, and Babylonian kings) whose chronologies are directly affected by Rohl's own theories, and hence must be removed as sub judice. The Babylonian kings between Sumu-abi and Sennacherib are treated as a single dynasty, as are the kings of Israel, even though it is well-documented that multiple dynasties are involved in each case, many of which consisted of only one or two kings. The late Arsacid kings of Iran are included even though their genealogy is almost completely unknown. Dynasties are included from completely different societies – Egyptian, Mesopotamian, Judean, Greek, Iranian – with no apparent recognition that the marriage customs of these societies were entirely different and therefore likely to result in different ages for marriage and childbirth. Not only brotherly succession may occur but all sorts of jumps within a family can occur; indeed, the very definition of what constitutes a generation is a thorny issue when there are such jumps, or when, as with the Ptolemies in this group, there is a significant amount of inbreeding. No recognition appears to exist for the biasing effects of dynastic termini: the founder of a dynasty may achieve the throne late in life; the last ruler may lose the throne early in life. And so on, and so on.
In an examination of the utility of oral genealogies for estimating intervals of time, the anthropologist David Hennige performed a generational analysis on 737 well-documented dynasties world wide, resulting in a distribution chart showing dynastic averages ranging from 15 to 49 years.20 Partly in reaction to Bierbrier's work, Henige provided a summary of this work to the Egyptological community in which he argued that the average could easily be even larger than Bierbrier's 25 years. This work, while not specifically focussed on Egypt, is much more extensive and methodologically far more sound than Rohl's effort, and it is curious Rohl has made no note of it.
Since Bierbrier's work was undertaken in the conventional chronological framework, and since similar studies have not, to the best of my knowledge, been carried out for the upper classes of the Egyptian populace in the chronologically secure Saite and Ptolemaic periods, the only reasonably valid cross-check one can easily make is to estimate the intergenerational interval of the ruling dynasties themselves, based on their relative reign-lengths. However, there are very few dynasties which can reasonably be used to establish such an intergenerational interval: the 12th, the 18th from Thutmosis I to Akhenaten, the 26th from the accession of Psamtek I to the death of Wahibre (Apries), and the Ptolemies. Others are either too little known, too short to establish a valid sample, or, as noted above, sub judice. Even the 6th must be discounted owing to the truly exceptional reign of Pepi II (94 years),21 and one could argue that the Ptolemies are more Greek than Egyptian, though their enthusiastic practice of dynastic incest suggests otherwise. To these we may add the Ptolemaic High Priests of Memphis, from Amenher II (died 217 BC) to Pedubastis IV (died 30 BC). Making allowances for uncertainties in chronology and the number of generations involved, one arrives at the following chart, which clearly supports Bierbrier's and Henige's arguments for generational averages significantly higher than 20 years.22