Extra Resources: B 29 (When) Did it Happen?

Further material referred to in note 8 of the accompanying Grove booklet B 29:

In The Lost Testament, Rohl raises the date for the end of the Late Bronze Age to c. 886 BC (p. 452), which may seem to resolve the Samaria problem. However, while it would (at a pinch) explain Samaria's Iron Age I pottery, it is inconsistent with Rohl's own Egyptian dates.

There is good evidence for placing the Late Bronze-Iron Age transition in the reign of Ramesses III, for whom Rohl gives dates of 863-832 BC (Lost Testament, p. 454). In Rohl's revised Egyptian chronology, the Late Bronze Age should therefore end around 850 BC. This is not early enough to solve the Samaria problem.

Rohl has argued (in personal communications) that some Iron Age I pottery forms may have come into use before the end of the Late Bronze Age. This is reasonable, and could explain the Iron Age I pottery at Samaria. However, it does not explain why Late Bronze Age forms are absent there.

Further material referred to in note 16:

Both James et al. and Rohl argue that the 21st and 22nd Dynasties should be overlapped because: (i) there is an unexplained gap in the burials of Apis bulls (an important cult at Memphis) from the end of the 20th Dynasty until the early 22nd Dynasty; (ii) a cache of mummies whose burial is dated to the reign of Siamun, penultimate king of the 21st Dynasty, included the mummy of an official who died early in the following (22nd) Dynasty; (iii) at Tanis the tomb of Osorkon II (22nd Dynasty) appears to have been built before that of Psussenes I (21st Dynasty).

How strong are these arguments? Kitchen dismisses the first on the grounds that the missing Apis-bull burials may yet turn up, or may have been destroyed (K. A. Kitchen, The Third Intermediate Period in Egypt (1100-650 BC), 2nd edition, reprinted with new material, Aris & Phillips, Warminster, 1996, p. xliii). As for the mummy cache, it has been argued that dockets dating from the reign of Siamun do not in fact record the movement of the mummies to their final resting place, which therefore may have occurred later, in the 22nd Dynasty (see C. Bennett, 'Temporal Fugues', Journal of Ancient and Medieval Studies 13, 1996, pp. 4-32, accessible online at www.christiananswers.net/abr/docs/temporalfugues.rtf). The evidence of the Tanis tombs has turned out to be much more complex in the light of new excavations, and the case seems less compelling than it once did (see P. Brissaud, 'Le monstre du Loch Ness est-il ne dans le Lac Sacre de Tanis?', Bulletin de la Societe; Francaise des Fouilles de Tanis 10, 1996, pp. 3-28, with Rohl's reply online at www.nunki.net/PerRenput/Reaction/ReplyBrissaud.html).

In short, the archaeological evidence from Egypt for a shorter chronology is far from watertight. This is particularly damaging to Rohl's case, as much of his argument for chronological revision rests on it. It is less of a problem to James et al., as their argument is much more broadly based.

Further material referred to in note 17:

Common objections to a shorter chronology examined

1. Synchronisms

Synchronisms between Egypt and other powers provide a crucial test for any proposed revision. For example, during the 'Amarna period' of Egyptian history (when the pharaoh Akhenaten relocated the capital from Thebes to a site now known as El-Amarna), diplomatic correspondence passed between Egypt, Assyria, the Hittite empire, the Kassite kings of Babylonia, and numerous city-states in Syria-Palestine (the famous Amarna letters). If any of the rulers who corresponded with Egypt during this period (conventionally c. 1360-1340 BC) could be shown independently to have reigned in the 14th century BC, any drastic revision of Egyptian chronology would be excluded.

So are these rulers witnesses for or against a shortened chronology? Only a few examples can be mentioned here (for all the relevant synchronisms see James et al., CD, pp. 304-308, 340-344).

The case against: Assyria

As noted in chapter 4, the Assyrian practice of keeping eponym lists means that Assyrian chronology is accurate to within a year back to 911 BC. Actually Assyriologists would say it has been reconstructed with reasonable accuracy back to at least 1400 BC. Although eponym lists are fragmentary before 911 BC, the Assyrian King List (extant in five copies) is held to provide a reliable framework from the 8th century BC back to the middle of the second millennium.

An Assyrian king called Assuruballit sent letters to Egypt during the Amarna period, and the King List does indeed place a king of this name in the 14th century BC. This is apparently strong evidence for the conventional scheme. However, proponents of a revised chronology point to a weakness in this synchronism. The Assuruballit known from the Amarna letters names his father as Assur-nadin-ahhe, but the father of the Assuruballit known from the Assyrian King List (and other inscriptions) is consistently called Eriba-Adad. (As in the OT, 'father' can have the broader meaning of 'ancestor' in Assyrian texts, but Assur-nadin-ahhe is not attested at all in royal inscriptions, even though Assuruballit names his ancestors going back six generations.) It is therefore possible that we are dealing with two different Assuruballits (CD, pp. 306, 340; TT, pp. 396-98) [1].

But a further synchronism between Assyria and Egypt also appears to uphold the conventional scheme. The Hittite king Tudhaliya IV wrote to one Tukulti-Ninurta who was almost certainly a king of Assyria. A king of that name (labelled Tukulti-Ninurta I by modern historians) is known from the Assyrian King List, which places him in the late 13th century BC. This is where the conventional Egyptian-derived dates for the Hittite Empire would put the king addressed by Tudhaliya.

Chronological revisionists challenge the conventional reconstruction of Mesopotamian history on several grounds. James et al. point to significant gaps (c. 1200-1150 BC and c. 1060-940 BC) in Assyria's archaeological record, an almost complete 'dark age' in Babylonia between the 11th and 8th centuries BC, and a string of stratigraphical and art-historical anomalies from both regions. They suggest that the Assyrian King List 'has been deliberately "smoothed out" to give the impression of an unbroken royal succession down to the 8th century BC', and point out that a 'notoriously problematic part of the King List overlaps with the dark period of Assyrian archaeology during the 12th to 10th centuries BC' (CD, pp. 271-308; cf. Rohl, TT, pp. 394-98). They point to inscriptional evidence that two dynasties ruled concurrently in Assyria after Tukulti-Ninurta I, and propose that a 'tuck' should be made here (and possibly earlier as well) to achieve a reduction. So far they have not published a detailed outworking of this, but a more thorough argument for parallel Assyrian dynasties has been made in the context of Rohl's revision by B. Newgrosh [2], and in the context of a shorter reduction by G. Hagens [3].

These proposals are by no means free from difficulties. So long as the Assyrian King List is believed to provide a reliable framework back to the mid-second millennium BC, drastic shortening of Assyrian chronology, and hence of ancient Near Eastern history in general, is ruled out. The history of Assyria currently poses the most stubborn obstacle to a major chronological revision.

However, in view of the strong evidence for chronological reduction from so many other areas, it would be rash to reject it on the basis of the Assyrian King List, whose historical accuracy remains unproven. It is also worth pointing out that the revision proposed by James et al. permits a number of new sychronisms which the conventional scheme does not allow (CD, pp. 137, 249-251, 255, 303, 308). As a counter-balance to the problems posed by the Assyrian King List, we will look briefly at favourable evidence from another area.

The case in favour: the Hittites

The Amarna correspondence reveals that the Hittite king Suppiluliuma I was a contemporary of Akhenaten; during the 19th Dynasty, Ramesses II fought against the Hittite ruler Hattusili III and the two later made a non-aggression pact. It is clear from these and other synchronisms that any redating of Egypt's New Kingdom must involve a similar redating of the Hittite Empire.

The Hittite Empire collapsed, in conventional terms, around 1190 BC (the end of the Late Bronze Age in that area). In central Anatolia, the Hittite civilisation's heartland, a long cultural gap followed this event. The gap apparently lasted until the emergence of the Phrygian kingdom in the 8th century BC. Yet, stratigraphically, there is no evidence for a long haiatus between the two cultures (e.g. no sedimentation layer between Hittite and Phrygian strata at Boghazk, site of the Hittite capital Hattu_a).

In northern Syria 'Neo-Hittite' states flourished after the Empire's collapse. This region saw a revival of Hittite traditions of art, architecture and language, with inscriptions written in the Hittite hieroglyphics of the Empire period. As Peter James points out, scholars have indulged in 'a chronological tug-of-war' over several elements of this Neo-Hittite civilisation. There is clear Late Assyrian influence on its artistic traditions, favouring a date in the 10th-8th centuries BC; but Imperial Hittite influence is also strong, leading some scholars to place the earliest Neo-Hittite phase much earlier (CD, pp. 121-124).

Around 1325 BC (conventional date) the Hittite emperor Suppiluliumas I conquered Carchemish on the Euphrates, and from then until the collapse of the Empire Carchemish was the base from which a branch of the Hittite royal family ruled Syria. This imperial line ruling in Carchemish is well documented almost until the end of the Empire period, after which there is a gap until the 10th century BC when Assyrian texts reveal Carchemish as a thriving Neo-Hittite kingdom [4].

Neo-Hittite Carchemish was ruled by a line of kings known as the Suhis dynasty (after its first known ruler Suhis I, c. 975 BC). In the revision proposed in CD, the Hittite Empire would not have disintegrated until the latter half of the 10th century BC and the last imperial rulers of Carchemish would have been contemporary with the Suhis dynasty. This may sound like a difficulty for the revision, but in fact there is evidence for just such an arrangement. Inscriptions suggest that the kings of the Suhis dynasty, who bore the title 'Lord of the Country of Carchemish', were contemporary with another line whose shadowy rulers claimed the higher status of 'Great King', a title borne by kings of Carchemish and Hattu_a in the Empire period. Since the publication of CD, David Hawkins, a leading authority on Hittite inscriptions, has outlined a scheme of parallel dynasties very similar to the one proposed there [5]. Of course Hawkins does not identify the 'Great Kings' who lived in the 10th century BC with the imperial rulers of Carchemish because he accepts the conventional chronology. However, Peter James points out that one of those 10th-century 'Great Kings' has the Luwian name Ura-Tarhundas, which translates into Hurrian as Talmi-Teshub, a king of Carchemish known from the late Empire period.

If we identify the two, the history of Carchemish can be rationalized straight away. The void between the Empire and the 10th century BC would be removed; the mysterious Great Kings are explained as the last incumbents of Hittite imperial power in the area; and above all, continuity in art and archaeology between the two periods is restored. (CD, p. 137.)

Space prevents reference to the many other anomalies that would be resolved by lower dates for the Hittite Empire. In terms of OT history the redating presents no difficulties. Placing the Empire's collapse in the second half of the 10th century BC would simply mean that the 'kings of the Hittites' with whom Solomon traded and intermarried (1 Kgs 10:28; 11:1) belonged to the late Empire period.

This brief survey shows the inconclusive nature of arguments based on synchronisms. It should be clear, however, that the problems posed by the Assyrian King List should not be allowed to veto the whole project of constructing a shorter chronology for the ancient Near East.

2. Radiocarbon dating

It is often claimed that the scientific dating technique known as Radiocarbon dating supports the conventional chronology, but this is misleading [6]. Radiocarbon results are so far ambivalent. Many (especially from tests done before the mid-1980s) have too large a margin of error. Sometimes a single stratum (or even a single mummy) can yield contradictory results, some favouring conventional dates and some favouring much later ones. The issue has been complicated by a recent study arguing that Radiocarbon dates from samples that grew downwind from the Mediterranean are likely to be too early [7]. If this is correct, it means that, even if the majority of dates from the ancient Near East favoured the conventional chronology, we would have to treat that outcome with extreme caution. For a fuller discussion of Radiocarbon dating see the CD website: www.centuries.co.uk

3. Dendrochronology

This is another scientific dating method with the potential to provide absolute dates for the ancient Near East. It involves matching the annual rings in wood taken from excavations with a master tree-ring sequence. Work is in progress on a master sequence for Anatolia that currently spans over 1500 years. So far this has been joined only indirectly with more modern sequences of tree-rings that could provide a firm anchor for it, but many researchers are already treating it as a reliable yardstick.

As such it has been seized on as proof that a drastically shortened chronology cannot work. A Bronze Age shipwreck off Uluburun (Turkey) included among its cargo a scarab of Nefertiti, wife of Akhenaten. A log also from the ship's cargo was dated to shortly before 1300 BC by matching its rings with the Anatolian master dendrochronology. This was hailed as confirmation that Akhenaten reigned in the 14th century BC [8]. However, in a little-publicised statement, the match between the log and the master sequence was subsequently admitted to bedoubtful [9].

The task of matching finds to the master sequence is by no means straightforward, as the site of Tille Hoyuk (an outpost of the Hittite Empire) illustrates. Wood from a gateway was matched with the Anatolian master dendrochronology using two statistical criteria (called 'correlation' and 'trend') and dated to the 12th century BC. However, the Tille Hoyuk wood also matched another date very closely on both statistical tests, namely 981 BC: 'For 981 BC, the significance of the correlation was 99.99995% and of the trend, 99.5%. The investigators matched the wood to 1140 BC (correlation significance 99.9995%; trend significance 99.99999%).' [10] The later date for the gate's construction would clearly be incompatible with the conventional chronology, but would strongly support the CD revision. (Note that even the 1140 BC date involves extending Tille Hoyuk's existence several decades beyond the end of the Hittite Empire.)

More fundamentally, D. J. Keenan is suspicious of a methodology that can produce widely-separated dates with this high statistical confidence. He concludes: 'Anatolian dendrochronology should be regarded as suspect and in need of independent scrutiny.' [11]


4. Astronomical dating

The absolute dates given to Manetho's framework of Egyptian dynasties appear to be supported by astronomical dating methods. In view of the claims sometimes made for these, we must venture briefly into this somewhat technical area.

The Sothic Cycle

The Egyptians had a civil calendar of 365 days, one quarter of a day short of the actual solar year. Since they made no adjustment to allow for the difference (as we do by adding an extra day every leap year), this meant that New Year's Day of the civil year shifted slowly around the natural year. In an 'ideal' year, New Year's Day coincided with the first day on which the star Sirius (or Sothis) rose early enough to be visible before the sun's light obscured it (July 19/20th in our calendar). But after four years, New Year's Day would be a whole day earlier than this event (known as the heliacal rising of Sirius), and after eight years it would be two days earlier, and so on. It would take 1460 years (i.e. 365 x 4) for an 'ideal' year to come round again, and this period is called the Sothic cycle. According to the Roman writer Censorinus, the Egyptian New Year's Day coincided with a heliacal rising of Sirius in 139 AD, which means that the previous Sothic cycle began 1460 years earlier in 1321 BC. The one before that would have begun in 2781 BC [12].

The ancient Egyptians watched for the heliacal rising of Sirius because it heralded the annual inundation (when the Nile rose above its banks - an event that ceased with the construction of the modern dams), and preparations had to be made for this. In theory, understanding the Sothic cycle means that if an ancient Egyptian document records a heliacal rising of Sirius, and gives it a precise date in terms of the civil calendar, it is possible to reckon the number of days by which the calendar had slipped in relation to the natural year. Multiplying that number of days by four gives the number of years since the start of a Sothic cycle, from which a date BC can then be calculated (within an error of +/- 3 years). If the same document can also be dated to the reign of a particular pharaoh, then we know that his reign must have included the year in question.

Doubtful Dates

We have only four references to a heliacal rising of Sirius from before the Graeco-Roman period, and two of those lack important data. Of the others, the Ebers Papyrus appears to place the 9th year of Amenhotep I, an early New Kingdom pharaoh, in either 1517 or 1506 BC (depending on whether Sirius was observed from Thebes or Elephantine). This Sothic date would fix the New Kingdom at c. 1550-1070 BC.

However, several Egyptologists have questioned the exact meaning of the reference in the Ebers Papyrus (Rohl, TT, pp. 134-5 and references cited there). Manfred Bietak concluded: 'It would therefore be safer not to use it any longer for chronological reconstructions.' [13]

The other Sothic date which has been said to provide a firm linchpin is found in a collection of documents known as the El-Lahun (or Illahun) Papyri, a temple archive dating from the Middle Kingdom. This text dates from Year 7 of an unnamed pharaoh, widely believed to be Senwosret III of the 12th Dynasty. It provides a date of 1871 BC (if the observation was from Memphis, slightly later if from further south).

However, other texts from the same archive contain lunar observations which do not fit well with any date in the 19th century BC. Some scholars have opted to disregard the lunar data, while others have chosen to ignore the Sothic date (strictly speaking a prediction of a heliacal rising made 21 days in advance, so perhaps not an accurate record) [14]. Following the latter option and using only the lunar data, D. Lappin has calculated dates of 1699-1680 BC for Senwosret III, in very close agreement with Rohl's chronology for the 12th Dynasty [15].

More fundamentally, one of the basic tenets of Sothic dating is open to question - namely, that the system operated without adjustment throughout the pharaonic era. This involves an assumption that Egyptian priests never took any steps to correct the slippage between the natural year and the civil calendar (even though the latter named its three seasons 'inundation', 'sprouting (of crops)' and 'summer'). As Chris Bennett, a staunch critic of drastic revisions, admits: '...Any calendrical reform would completely invalidate the entire theory for estimating dates prior to such reform. Some evidence suggests that the Egyptians did experiment with the calendar.' [16]

A Kitchen-Rohl dispute

Finally, something must be said about an alleged fixed date which has been wrongly associated with Sothic dating. Responding to Rohl's revision, but with James et al. also in his sights, Kitchen refers to a short text which he claims 'destroys completely, finally and irrevocably all these ambitious but ill-founded attempts to move Egyptian chronology by centuries'. The text is a graffito from the reign of the 19th-Dynasty pharaoh Merenptah and it contains a precise date, given by Kitchen as 'Year 1, 3rd month of Akhet [=Inundation], day 3'. It records, in its own words, 'the descent made by the water of the great inundation', which shows, according to Kitchen, that the season called 'Inundation' in the civil calendar actually corresponded with the annual Nile flood in Merenptah's first year. 'Thus, as we know that the calendar was correct in the 2nd century AD, and in the 13th century BC..., there can be no doubt that Year 1 of Merenptah in this little text fell in the 13th century BC - and not 250/350 years later, as some would like. All such "revisions" are excluded.... The game is up.' [17]

In a short article Rohl responds on two fronts to Kitchen's criticism. Having examined and photographed the text himself, Rohl argues (in agreement with a 19th-century scholar) that the correct reading is 'Year 2, 2nd month...', not 'Year 1, 3rd month...'. Secondly, he questions Kitchen's view that the word translated 'descent' refers to the waters of the Nile flooding the fields; he argues that it would be more natural to take it as a reference to the flood receding. Rohl points out that either of these re-readings of the graffito would change the timing of the event by 30 days, requiring 120 years (30x4) of a Sothic cycle to have elapsed. Together they would displace Merenptah by 240 years from his conventional date [18].

So who is right? Given the crudeness of the graffito, the correct reading may never be known, but both Kitchen and Rohl are wrong to apply the principles of Sothic dating. To be fair to Rohl, he stresses that he does not accept the usefulness of Sothic dating, but merely wants to show the real consequences if Kitchen insists on applying it. But he would have done better to have pointed out a different error.

When Kitchen says the seasons of the civil calendar corresponded with reality in the 2nd century AD and in the 13th century BC, he must be referring to the 1460-year Sothic cycles that began in 139 AD and 1321 BC, the latter being not far from the start of the 13th century BC in which Merenptah reigned. (Note, however, that Kitchen's dates for Merenptah are 1213-1203 BC, over a century after the start of the Sothic cycle.) But, as Lynn Rose points out, the 1460-year Sothic cycle is actually irrelevant here. What matters is the time it would take the cycle of the seasons (the tropical year) to move around the civil calendar and return to its starting point. This period is actually 1507 years. Hence if the date 'Year 1, 3rd month of Akhet, day 3' was in phase with the inundation in 139 AD (as Kitchen has it), 'then, by retrocalculation of the tropical year... it would previously have been in phase with the inundation more than 1500 years earlier, in the14th century BC!' [19]. In short, Kitchen's logic would actually place Merenptah 150 years earlier than his conventional date - an impossible setting in any chronology.

All this highlights the uncertainties over exactly what Merenptah's Nile text is saying. While these remain, the inscription cannot be an obstacle to a major chronological revision.


1. For more detailed discussion see B. Newgrosh, 'Ashur-uballit, King of Assyria: A Question of Identity, Journal of the Ancient Chronology Forum 9, 2002, pp. 91-109. Back
2. B. Newgrosh, 'The Chronology of Ancient Assyria Re-assessed', Journal of the Ancient Chronology Forum 8, 1999, pp. 78-106; 'The Chronology of Ancient Assyria Re-assessed: An Update', Journal of the Ancient Chronology Forum 9, 2002, pp. 109-112. Back
3. G. Hagens, 'The Chronology of Tenth-Century Assyria and Babylonia', Journal of the Ancient Chronology Forum 9, 2002, pp. 61-70. Back
4. There is a reference from the reign of Tiglath-pileser I of Assyria (conventionally c. 1100 BC) to a king of Carchemish called Ini-Teshub, but no texts from that time attest his reign. In the CD revision, he can be recognised as a king Ini-Teshub known from the late Hittite Empire (CD, pp. 125, 135, 303). Back
5. D. Hawkins, '"Great Kings" and "Country Lords" at Malatya and Karkamis', in T. P. J. van den Hout and J. de Roos, eds., Studio Historiae Ardens, Istanbul, 1995, pp. 73-85. Back
6. See the response to critics by James, Kokkinos and Thorpe in Balmuth and Tykot, eds. (see n. 16 in booklet), pp. 36-38, and more briefly on the CD website: www.centuries.co.uk Back
7. D. J. Keenan, "Why Early-Historical Radiocarbon Dates Downwind from the Mediterranean are Too Early", Radiocarbon 44/1, 2002, pp. 225-237. Back
8. A. Dodson, 'Towards a Minimum Chronology of the New Kingdom and Third Intermediate Period', Bulletin of the Egyptological Seminar 14, 2000, p. 15 and references cited there. Back
9. See S. W. Manning, B. Kromer, P. Kuniholm and M. W. Newton, 'Anatolian Tree Rings and a New Chronology for the East Mediterranean Bronze-Iron Ages', Science 294, 2001, p. 2535, n. 38. Back
10. Keenan, as above, p. 232. Back
11. Keenan, p. 233. See also James, Kokkinos and Thorpe in Balmuth and Tykot, eds. (see n. 16 in booklet), pp. 38-40, and James, 'The Dendrochronology debate', Minerva, July/Aug 2002, p. 18, and the CD website: www.centuries.co.uk Back
12. In practice things are more complex than this. See P. F. O'Mara, 'Censorinus, The Sothic Cycle and Calendar Year One in Ancient Egypt: The Epistemological Problem', Journal of Near Eastern Studies 62/1, 2003, pp. 17-26. Back
13. M. Bietak, 'Egypt and Canaan in the Middle Bronze Age', Bulletin of the American Schools of Oriental Research 281, 1991, pp. 27-72 (quotation from p. 47). Back
14. See respectively: R. A. Wells, 'The Role of Astronomical Techniques in Ancient Egyptian Chronology: The Use of Lunar Month Lengths in Absolute Dating', in J. M. Steele and A. Imhausen, eds., Under One Sky: Astronomy and Mathematics in the Ancient Near East, Munster, 2002, pp. 459-472; L. E. Rose, 'The Astronomical Evidence for Dating the End of the Middle Kingdom of Ancient Egypt to the Early Second Millennium: A Reassessment', Journal of Near Eastern Studies 53/4, 1994, pp. 237-261. Back
15. D. Lappin, 'The Decline and Fall of Sothic Dating: El-Lahun Lunar Texts and Egyptian Astronomical Dates', Journal of the Ancient Chronology Forum 9, 2002, pp. 71-84; cf. Rohl, TT, pp. 339, 390-393. Rohl's chronology claims a high degree of support from astronomical methods; see also W. A. Mitchell, 'Ancient Astronomical Observations and Near Eastern Chronology', Journal of the Ancient Chronology Forum 3, 1990, pp. 7-26. Back
16. C. Bennett, 'Temporal Fugues', Journal of Ancient and Medieval Studies 13, 1996, pp. 4-32, note xviii; accessible online at: www.christiananswers.net/abr/docs/temporalfugues.rtf Back
17. K. A. Kitchen, The Third Intermediate Period in Egypt (1100-650 BC), Aris and Phillips, Warminster (2nd edition, reprinted with new material), 1996, p. xlv. Back
18. D. Rohl, 'Kenneth Kitchen's Atom Bomb', Journal of the Ancient Chronology Forum 8, 1999, pp. 43-47; Rohl does not say in which direction Merenptah would be displaced, but in fact he would reign 240 years earlier than his conventional date. Back
19. L. E. Rose, 'The Role of the Nile in Egyptian Chronology', Chronology and Catastrophism Review, 2001:2, pp. 26-30.