Centuries of Darkness 
by Peter James.
Chatto, 434 pp., £19.99, April 1991, 9780224026475
Show More
Show More

Gigantic academic blunder? The phrase appears without the question-mark on the last page of Centuries of Darkness. That title too, as we shall see, would have better conveyed the book’s message if it had ended in a question-mark. Fortunately, though, the message is already brilliantly expressed by the book’s dust-jacket, showing the design from the painted box of Tutankhamun in the form of a jigsaw. The jigsaw is divided into two parts by a black intervening gap, yet we can see that the two halves would actually fit together. The gap should not be there.

What Peter James – the main author and incidentally the designer of the jacket – and his four collaborators seek to prove is, put simply, that the entire early history of the civilised world has been similarly distorted. An intrusive, imaginary gap, partly of modern invention and about 250 years in length, has been inserted into the sequence, artificially lengthening it so that all dates before about 950 BC are two and a half centuries too early. All historically-based dates, that is: dating by scientific methods is a separate issue.

Four years ago, in another revolutionary work, Black Athena, Martin Bernal argued that outsiders have been responsible for most fundamental challenges to disciplines. The first interesting point about this book is that its origins are of a quite different kind. Peter James is a graduate student of University College London and his colleagues are archaeologists and historians of similar standing. They are thus perfectly proper, if youngish, members of the ‘guild’. This should make their professional colleagues take their work all the more seriously, while the intelligent layman will surely detect the difference between first-hand authority and the maverick citation of deservedly-forgotten sources which usually passes for research in the game of Confound the Experts.

How could such a massive miscalculation ever have come about – let alone won universal acceptance? James proceeds by a very indirect route before he approaches the heart of the matter, which is to be found (of course) in Egypt. After a brief sketch of the problem, the next eight chapters take us in a great clockwise circle from the Western Mediterranean, through the Balkans to Troy, Greece, Asia Minor, Cyprus, Israel and the Sudan. In each case, the supposed ‘fixed points’ for dating the regional sequence are examined with a cold eye. In each region, it emerges, the archaeologists have been dancing to the magic flute of Egyptian precision.

At this stage, the argument is based almost entirely on circumstantial evidence. In one instance after another, there emerges an unexplained ‘interlude’ of very slow development or apparent total stagnation, when the archaeological record becomes thin and the historical cross-references disappear altogether. In the more advanced cultures, this occurs at the beginning of the Iron Age; in the more backward, it comes near the end of the Bronze Age. Often the material record of the last period before the ‘interlude’ shows suspicious links with the first period after it. The authors find a few forerunners who have taken steps – if short and hesitant ones – in the right direction. Einar Gjerstad, a great and undoubtedly underrated archaeologist, appears as the hero twice over, in Italy and in Cyprus. The authors are so far content to pose questions in the form ‘What if ...?’ and ‘Why not ...?’ What if the ‘interlude’ is unreal and the sequence uninterrupted? Since the chronology of the later periods is in general more securely based, they cannot be extended backwards in time: so the earlier periods will have to come down. Why should they not? Because of Egypt.

Here a new and more formidable adversary looms up before them: astronomy. Everyone knows that the ancient Egyptians watched the stars. Very occasionally, their observations are recorded in a form that has chronological significance. The starting-point is a report that, in the year AD 139, something happened in Egypt that would only occur once in 1460 years. Sirius appeared on the eastern horizon just before sunrise on New Year’s Day. The Egyptians kept a 365-day calendar without leap years, while the phenomena of the solar system – like this ‘heliacal’ rising of Sirius – observe a year of about 365 1/4 days. The Egyptian calendar therefore slipped out of synchronisation with the Sun by one whole day every four years. So when the records mention such a rising of Sirius on a particular calendar date – any date – it is possible to place the year of that record. The only alternatives will be 1460 years earlier or 1460 years later, and these can he easily excluded. Thus the ninth year of the 18th-Dynasty Pharaoh Amenhotep I, which can hardly have been in 2977 or 57 BC, must have been in 1517 BC.

Or must it? James is quick to point out the most important hidden assumption here: the assumption that the Egyptians never tampered with their calendar. (There are other assumptions too, but they are secondary.) Why should we believe that an intelligent and scientifically curious priesthood, finding that their calendar was going further and further astray in relation to nature – in seven centuries, New Year’s Day would have moved from midsummer to midwinter – did nothing to about it for three thousand years? As a matter of fact, we know that, latterly, they did try to adjust: under Ptolemy III, for example, leap years were officially introduced, but they proved unpopular. Any earlier adjustment, if it won acceptance, would have been enough to wreck the modern chronological calculations.

There is another obstacle, though – the king-lists. Those grim catalogues of dynasties used to intimidate the readers of old-fashioned history books. But how was this catalogue assembled? It was not merely a matter of adding up the lengths of the successive reigns. The fact is that we do not know how long some of the pharaohs reigned. Worse, certain dynasties are known to have overlapped, at times when Egypt was divided. This happened especially in the Third Intermediate Period, conventionally dated from the 11th to the seventh centuries BC – the very period that this book is about. The truth is that the evidence of this period has been arranged to fit a pre-determined time-span, pre-determined by the few earlier dates that are thought to be astronomically fixed. But the rulers of the Third Intermediate Period cannot, even then, be fitted end-to-end: and so a degree of overlap has had to be accepted, even by proponents of the conventional chronology. In for a penny, in for a pound, says James: take away the astronomical ‘fixes’ and the whole period will shrink to its natural length, which is not four centuries but about a hundred and sixty years. This is the coup de grâce: if the Egyptian dates collapse, then all the half-civilised cultures round the Mediterranean, and the regions beyond which are dated by connections with these cultures, will move down with them.

There remains Mesopotamia, which has a dating structure of its own, partly independent of Egypt’s. But it proves to have the same weaknesses in accentuated form. The earliest celestial ‘fix’ is an eclipse of the Sun on 15 June 763 BC. Before that, everything is based on trust in the Assyrian and Babylonian king-lists. These betray familiar faults: they disagree with each other; they show the same kings’ names recurring; they portray as successive dynasties ruling houses which more probably reigned in parallel over a divided country.

Their real aim was not the truth but ‘the exaggeration of antiquity’, which forms the title of the 12th and penultimate chapter. The book ends with a few pages sketching the advantages of the new scheme which will emerge if it is accepted. The great Bronze Age civilisations of the Near East and the Mediterranean did indeed collapse, as everyone has always believed; but they did so two and a half centuries later than hitherto realised, and the subsequent recovery, instead of being delayed by a long period of stagnation, was rapid. Archaic Greece grew straight out of the ruins of Mycenae, and the neo-Hittite kingdoms from those of the Hittite Empire; the notorious warrior rulers of Assyria built fairly soon on the conquests of their supposedly remote predecessors; in Egypt, meanwhile, the Third Intermediate Period was a real interruption, but one which had only recently begun. Most interesting of all, the empire of David and Solomon, whose absolute dates are for special reasons retained, now coincides with the glories of the Canaanite Late Bronze Age, instead of languishing in the apparent squalor of the Early Iron Age

It all sounds rather attractive, What will be its reception and its long-term impact? There will certainly be the instinctive conservatives who, in every discipline, mobilise against radical change. The reaction of some of the Egyptologists can be expected to be especially dismissive. But the proper question is: do the gains of the new scheme outweigh the losses, not in terms of vested interest or amour-propre, but from the point of view of understanding the past? For, so far, the issue is one of persuasion rather than of demonstration. The conventional scheme has not been proved wrong; what has been demonstrated, I think quite adequately, is the possibility of its being wrong.

The main impact of the proposed change does not fall in prehistoric Bronze Age studies, where many of the important question are unaffected by a change in absolute dating. It comes in the archaeology of the early historical period, where the proposal is not for mere alteration, but for elimination. This reviewer, having written a book about a period which, according to James, does not exist, is right in the firing-line. But I wish to turn the argument to a quite different direction: that of science.

The same solution is advocated in a brief foreword to the book by Colin Renfrew. On the back of the dust-jacket where we began, his words are judiciously excerpted by the publishers, ending with the resounding phrase: ‘a chronological revolution is on its way’. The trouble (for the authors) is that it is not the same one. Indeed, its effects tend in the opposite direction, as a full reading of Renfrew’s foreword shows.

The intrusion of science into these questions began with the development, in the 1940s, of radiocarbon dating. It took a major step further in the 1960s, when it was found possible to text radiocarbon dates against the much more precise dates obtained from tree-ring sequences, by submitting the same samples to both methods of investigation. Now, from the later 1980s on, a more dramatic possibility has dawned: through extending the tree-ring sequences over an ever-wider geographical area, and further and further back in time, it is becoming possible to date certain episodes involving the use of timber, in antiquity and even in prehistory, to the year. What is more, these dates can be checked, not only against the rather erratic radiocarbon clock, but against sequences of comparable precision like the Arctic ice-core layers.A phenomenon like a major volcanic eruption, for example, can show itself both in the tree-rings and in the ice-cores.

This book makes intermittent reference to radiocarbon dates, and devotes an appendix to tree-rings: but both are used exceedingly sparingly. There is no mention of either technique in the chapters on Greece, Cyprus, Nubia or Mesopotamia. Yet there is some awkward evidence there, as the case of Greece will show. The ruins of the Mycenaean palaces, whose destruction is here brought down from about 1200 to about 950 BC, have yielded sequences of radiocarbon dates from charcoal samples. Once these dates are calibrated against the tree-rings, in the way just described, it can be seen that the end of the series, at both Pylos and Mycenae, comes out somewhere between the extreme outer limits of 1400 and 1100 BC. Some of the others are much earlier, and are usually explained as being from timber beams already old at the time of destruction. No doubt the authors will say that the same is true of these late samples: but they should have said so in the book.

Then there is the Thera eruption. Here the adjusted radiocarbon dates suggested a dating, not between 1500 and 1450 BC as the conventional chronology predicted, but about 150 years earlier. Then it was found that the tree-rings in distant California and the ice-cores in the Arctic concurred in a spectacular way: a major volcanic eruption, somewhere in the world, had undoubtedly occurred in the 1620s BC. No comparable effect could be detected in any other year that lay within the possible time-range of the Thera cataclysm. This evidence, too, finds no place in the book between the preface (where it is coolly received by the authors) and the already-mentioned appendix. Yet it has persuaded an ever-growing number of Acgean prehistorians to abandon their long-cherished conventional chronology.

This example points to a cautious move backwards in time in the dating of the Aegean Late Bronze Age; yet we have in this book a proposal for an abrupt extension forwards in time for the same period. The two possibilities are not irreconcilable: the period may have been very much longer than was thought (some eight centuries rather than five). But this recent evidence hardly provides a fair wind for the authors’ venture.

There is worse, from their point of view, to come. The tree-ring sequence for Asia Minor, built up by Peter Kuniholm and his colleagues, has progressed much further than this book acknowledges. In June 1988 Kuniholm was able to announce an impressive break-through: he linked up the 806-year-old sequence of timbers from an Iron Age burial-chamber at Gordion to some pieces of charcoal from a Bronze Age palace of the Hittite king Suppiluliuma I, some miles away. The interval between the two structures was 654 years, though neither could be precisely dated in years BC. For the conventional chronology, which would place Suppiluliuma on the throne in 1350 BC and the tomb-chamber at about 700 BC, the dates came as a shot in the arm. James and his colleagues, for whom the interval could hardly be longer than four hundred years, seem rather to have been shot in the foot. There is of course the escape-route suggested just now for Mycenae, that the Hittite king’s palace incorporated some very aged timbers: but that way out loses credibility if it is used every time.

In this instance, the authors were doubtless unaware of the new finding when they wrote the book. So, too, with an even later (1990) report from Kuniholm, that radiocarbon determinations from his Asia Minor tree-rings have been successfully matched, in 18 separate phases, with those from an oak sequence in northern Europe, which runs right down to modern times but which, for the pre-Roman period, had hitherto been historically unattached. This will open the way, for the first time, to giving absolute dates, not only to Suppiluliuma but to his exact contemporary Tutankhamun of Egypt, to the Kassite kings of Babylon and eventually, no doubt, to other regions as well.

From the scientific viewpoint, therefore, the timing of this venture seems unlucky. Whatever one makes of the scientific evidence so far available, one prediction is safe: in the near future – perhaps within less than five years – the volume of evidence will have increased to the point where a decisive judgment can be given. The indications are that the verdict will go against Centuries of Darkness, and send us back to – centuries of darkness. I admire the book enough to feel, quite sincerely, that that will be a great pity. But it does not mean that the authors’ time has been wasted: far from it. By exposing the myriad points of weakness in the conventional dating system for the later prehistory of Europe and western Asia, they will have helped to give it some of the flexibility that it will need to face the ordeal that undoubtedly awaits it. By their familiarity with the whole, vast field, and by the sober and reasoned way in which they put their case, they have earned the right to be listened to, and no doubt to be quite widely believed. Whatever the final outcome, posterity will have to say that it was a nice try.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN


Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences