Underworld: The Mysterious Origins of Civilization

by Graham Hancock

  ‘You’re quite right,’ I told Anton after I’d thought through his reasoning, ‘I don’t like it at all.’

  As he looked at me expectantly, I raised my left hand and began to enumerate the counter-arguments on my fingers.

  ‘Firstly, there’s the issue of the relative chronology. To make your argument work – I mean about the megalithic civilization of Malta being a thousand years older than the megalithic civilization of ancient Egypt – you have to buy into the orthodox archaeological datings for both places. But you ought to be the first to know that orthodox archaeological datings may not always be correct. In the case of Egypt we have actual structures, such as the Sphinx and the megalithic temples beside it, which may be much older than the third millennium BC7 – I’m sure you’re familiar with the debate. There’s the megalithic stone circle at Nabta, 200 kilometres west of Abu Simbel, which is at least 7000 years old.8 And then there are the accounts of the ancient Egyptians themselves – the Abydos King List, the Turin Papyrus and so on – which trace the origins of their civilization back 30,000 years into the past. Again, your relative chronology only works if you accept the orthodox position that all such accounts are baseless fictions – which I certainly don’t.’

  Secondly, I continued, Anton’s argument involved not taking Plato seriously on the epoch in which he had set the Atlantis events – supposedly 9000 years before Solon’s time, i.e., 9600 BC, i.e., about 11,600 years before the present, i.e., the end of the Palaeolithic. And I could see no good reason not to take Plato seriously on that – indeed, he could have hardly set his global deluge (described as affecting both the Mediterranean and the Atlantic) in a more flood-prone and cataclysmic epoch than the end of the Ice Age around 11,600 years ago. To conclude that Plato had not meant 9000 years before Solon’s time (9600 BC) but 1600 years before Solon’s time (2200 BC) seemed to me arbitrary, to say the least.

  Thirdly, the notion inherent in Mifsud’s reasoning that Plato must have been speaking of the Mediterranean west of Malta when he referred to the ‘true ocean’ leading to an opposite continent seemed to me to be highly suspect. I told Anton I was convinced that when Plato said this ocean was the Atlantic and placed Atlantis in it ‘opposite the Pillars of Hercules’ he knew exactly what and where he was talking about. So to my mind this on its own made Malta in the central Mediterranean an unlikely candidate for Plato’s island.

  But I hastened to add – and not just out of politeness – that none of this meant Mifsud was necessarily wrong. I could be the one who was wrong. Atlantis had been placed at other sites in the Mediterranean by other scholars – also at relatively late dates.9 And it had been placed almost everywhere else in the world from Indonesia to the South Pole.10 I happened to be one of those who believed in taking Plato as literally as possible – if I was going to take him seriously at all – but I recognized the validity of other approaches.

  Terminus ante quern

  Malta, 16 June 2000

  Besides, there was no real contradiction between our positions – for the simple reason that we seemed to be talking about entirely different things. To state, as Anton did, that Malta underwent a flood/earthquake cataclysm about 4200 years ago neither weakened nor strengthened my proposition that it would also have been subject to flood cataclysms – probably several times – during the meltdown of the last Ice Age between 17,000 and 7000 years ago. Likewise, Anton’s belief that a land-bridge between Filfla and south-west Malta collapsed through rifting processes 4200 years ago in no way contradicted the well-established fact that Malta’s north-eastern coast was connected to Sicily by a 90 kilometre land-bridge before it was swallowed up by the rising seas at the end of the Ice Age.

  Both periods are interesting for different reasons. But, I pointed out, Anton’s own research indicated that there had been human beings on Malta during the period of the great Ice Age floods at the end of the Palaeolithic. And although he seemed to accept the orthodox radiocarbon ‘sequence’ and chronology for the temples and the Hypogeum (of 3600 BC down to about 2200 BC), hadn’t he himself also written in Dossier Malta that:

  The terminus ante quern Carbon-14 dates given for these sites … are of before such and such a year in the Neolithic; whether this period was a year or several centuries cannot be established by the Carbon-14 date alone. The most logical explanation is successive utilization of such sites initially by Palaeolithic and subsequently by Neolithic Maltese.11

  And a few pages later:

  Since the megalithic temples have been assigned a terminus ante quern carbon date of before 3000 BC, nothing precludes that they were a carryover of a tradition which had started in the Palaeolithic. Indeed the bas relief images of bulls and a cow on the large blocks of stone lying just outside the Tarxien temple complex are themselves diagnostic features of Palaeolithic art.12

  I could sense Mifsud’s reluctance as a rigorous scientist to get drawn into idle speculation. But surely he was aware of the general direction in which his arguments were tending? If he was saying that there had been humans on Malta in the Palaeolithic, which he certainly was, and if he was suggesting that these Palaeolithic humans had initiated the development of the megalithic temples, which, again, he certainly was, then weren’t the floods at the end of the Palaeolithic of at least as much potential significance to Maltese prehistory as the floods and earthquakes that might also have occurred 4200 years ago?

  Skeletons in the Hypogeum

  Anton Mifsud’s attack on the orthodox chronology and interpretation of prehistoric Malta is made across several different fronts and sometimes produces contradictory data. This doesn’t seem to bother him. Once launched on an inquiry, he pursues the quest for data ruthlessly, as an end in itself, not to support particular arguments or positions.

  In the case of the Hypogeum, Mifsud’s approach, at first, was not directly concerned with chronology. Poring with the eye of a doctor over the early excavation reports of Zammit, Bradley and others, he was puzzled by what they had to say about the state of the human remains found inside the labyrinth.

  In summary, as we’ve seen, all the excavators and all subsequent archaeologists propose slightly different versions of the same theory that this great mass of remains had been ritually buried in the earthy matrix that was found filling the Hypogeum’s lower levels to a depth of about a metre when it was opened. Yet the instinctive reactions that they set out in the original reports that Mifsud had assembled, referred to in chapter 16, show that they were clearly startled, and in a few cases troubled, by the complete chaos and disorder in which the bones were found, commenting, for example: ‘From the upright position of an isolated radius it might be judged that the filling up of the cave was of a wholesale nature.’13 But how do we explain such a ‘wholesale’ filling-up of the Hypogeum with the remains of thousands of human bodies, all seemingly just dumped there ‘in a haphazard way’14 with no anatomical disposition? Isn’t it a bit of a mystery?

  Not according to the archaeologists who say they’ve seen mass ‘catacomb-style’ burials before in other parts of the world and on these islands – for example at Burmeghez, a natural cave in Malta,15 and at the Borchtorff Circle on Gozo, where rock-cut subterranean tombs encircle megaliths.16 So that makes this sort of ‘funerary behaviour’ part of a pattern that legitimate experts in the subject can already claim to understand. None of them would deny that the Hypogeum’s labyrinthine character is utterly different from the rock-tomb character of the other sites, or that the bones it contained were indeed in such an extreme state of disarrangement that any form of ‘regular interment was out of the question’.17 But the problem can easily be resolved within the prevailing ‘burial-place’ paradigm by proposing that excarnation – i.e., the removal of the flesh from the bones – was practised before interment and that the Hypogeum must therefore have been ‘a burial place in which the bodies were laid or heaped as skeletons’.18

  Oh really? Heaped up and tossed about so casually? As Mifsud counters, the Hypogeum cannot be legi
timately compared to either of the two other significant sites of prehistoric mass burial in Malta:

  At Burmeghez there is a predominance of anatomical relationship between body parts, a left-sided flexed position of the body, an orientation along the main axis of the cave, and, by way of a lithic assembly, a stony arrangement [large, purposefully laid slabs] protecting the upper body parts …19

  Like at the Borchtorff Circle, all the burials are of an evidently (and uncontested) ritual nature and were disposed in two phases – a pair of rock-cut tombs with a shared central shaft dated to the Zebbug phase20 (about 4000 BC-some centuries before the supposed beginning of the temple period at Gigantija c.3600 BC), and further subterranean rock-cut tombs of the Tarxien phase arranged in an approximate circle around a subterranean ‘megalithic assembly’.21

  In fact, the only site in Malta that Mifsud regards as comparable to the Hypogeum of Hal Saflieni in general appearance – and that was found on excavation to contain exactly the same sort of chaotic deposit – is the nearby Hypogeum of Santa Lucia (less than a kilometre away) that was excavated in the early 1970s and has since been sealed up, presumably for ever – at least so the authorities must intend – because it has been covered over by a modern cemetery.22

  Mifsud describes the Santa Lucia Hypogeum as:

  a smaller version of that at Hal Saflieni, with a megalithic entrance and an internal architecture similar to the temples above ground. The deposit inside this hypogeum consisted of human remains admixed with Neolithic pottery and amulets, in a matrix of red earth soil; the context is similar to that at Hal Saflieni. In the words of the Director of Museums at the time, the deposit inside the Santa Lucia Hypogeum was ‘as if the mass had been dumped inside the monument from the surface’. F. S. Mallia could not have been more precise, and the close proximity of the two hypogea enhances even further a similar mechanism operating in both monuments in the creation of the deposit in question.23

  Which brings us to the heart of the matter. Since Mifsud clearly does not believe that the carpets of disarranged human bones littered and dumped inside the hypogea of Hal Saflieni and Santa Lucia arrived there as a result of burial, then what ‘mechanism’ does he think was operating?

  A flood

  Like other good ideas that no one has ever had before but that everybody immediately gets the point of once the secret is out, Mifsud’s explanation for the mass of bones inside the Hypogeum is extremely simple:

  The accumulation of human remains at the Hypogeum in Hal Saflieni were not related to primary ritual burial, but were brought down into the Hypogeum labyrinth through the action of floodwater in a matrix of red earth and soil.24

  The first and most obvious evidence for this novel hypothesis comes in the massively disordered nature of the remains described in the excavation reports. The presence of these disarticulated, non-anatomically disposed remains in an entirely ‘unstratified’ deposit ‘made of the red earth one finds in our fields’ that was ‘always of the same type and contained objects of the same quality’, cannot in Mifsud’s view be explained by any form of deliberate burial – with or without prior excarnation. Only one agency, he argues, is capable of creating such a conglomeration in an unstratified earth matrix in which ‘the same quality of shards were found on the surface, at the bottom and in the space in between’,25 and in which ‘fragments of shards in parts of the Hypogeum fitted other fragments deposited in other caves far away’.26

  That agency is a massive flood – and such events, from varying causes, have not only been known to occur in the Maltese islands but also have left distinct traces of their passage in the form of animal and human bones, as well as assorted other materials, all muddled up together and evenly spread throughout deposits of silt or earth trapped inside caves or rock fissures. The classic example is Ghar Dalam, an extensive natural limestone cave near Birzebbuga in eastern Malta, which contains six distinct layers of flood deposits swept into its depths at different periods over the last 200,000 years. Exactly as in the Hypogeum, notes Mifsud, the organic remains in Ghar Dalam ‘were not distributed in an anatomical manner as they would have been in a ritual burial, but they were dispersed in random fashion inside the stratum of earth they lay in’.27

  What Mifsud is proposing in the case of the Hypogeum, therefore, is a one-off, one-time deluge that swept over the surrounding fields and habitations, and finally over a great surface-level necropolis that then existed in the area, carrying away all its mouldering dead in one fell swoop and dumping their skeletons and their grave goods, promiscuously mixed with fragments of pottery, the bones of large and small animals (including those of frogs and hedgehogs)28 and a motley collection of other objects, into the nearest possible sinkhole – in this case the Hypogeum itself.

  Moreover, Mifsud believes that this was the same flood – caused by the collapse into the sea of his proposed Filfla land-bridge and the resulting tsunami – that brought the temple-building culture of Malta, and all its activities within the Hypogeum, to an abrupt and permanent halt c.2200 BC. Since carbon-dates from Malta are as scarce as ice cubes in hell, as we shall see, it is interesting that the first ever radiocarbon-dating of the Hypogeum’s few surviving human remains – carried out in 1999 – does place them at the end of the Tarxien phase, c.2200 BC just as Mifsud argues.29 This new evidence from the Hypogeum, he concludes, further strengthens ‘the feasibility of a sudden cataclysm accounting for the sudden termination of the Tarxien people’.30 And he points to the well-known fact that the Tarxien temple itself was sealed at the end of the late Tarxien phase under a metre-deep layer of sterile silt.31 After several centuries of abandonment a new culture then appeared – one that had nothing to do with the temple-builders – and began building – on top of the silt layer.32

  Old Stone Age

  Although nobody in the world of archaeology seems to have noticed yet, the late Tarxien date (of between 2470 BC and 2140 BC)33 for the Hypogeum’s human remains contradicts the long-established convention, entered into dogma by J. D. Evans, that ‘the primary use of the Hypogeum [was] as a place of burial’. Since all archaeologists accept that the construction of the labyrinth began significantly earlier than 3000 BC, perhaps as early as 3600 BC (and since it even contained pottery of the Zebbug phase prior to 4000 BC), its ‘primary’ purpose can hardly have been to receive human remains that were not deposited in it until around 2200 BC (whether or not one accepts that they were deposited there by flood). It must, therefore, have had some other quite distinct function at the time of its origin – a function that scholars may hitherto not have guessed, since no serious attempt has ever been made to investigate alternatives to the burial scenario.

  It may have been an underground version of a temple, of course – and its ‘temple-like’ features have always been recognized – but if so, why is it such a unique and unusual temple? Why does it need all those winding corridors and levels and cists secreted within rock walls, and spooky sound-effects and red-painted chambers, and falls and traps?

  Whatever its purpose – probably it will never be known fully, or known at all – certainty that the Hypogeum was not primarily designed as a central place of burial for the dead of the temple-building culture as had hitherto been thought left the way open for Mifsud to explore other possibilities about its function and identity. And about its age. Because, as with the megalithic temples, so with the rock-hewn labyrinth, the argument of terminus ante quern applies, and nothing rules out the possibility that the Hypogeum may be a ‘carryover of a tradition which had started in the Palaeolithic – to quote again Mifsud’s own words, cited earlier. By the same token the flood that he believes swamped it with bones and debris at around 2200 BC tells us nothing whatsoever about the origins and antiquity of the structure itself – only that it was already there to be flooded in 2200 BC (and certainly not how long it had existed before that).

  Long but ultimately relevant excursion to two potentially unpronounceable temples

  Maltese is a lovely, l
ilting language to hear. Structurally it belongs to the Semitic family and is thus closely related to Arabic and Hebrew – indeed, Maltese friends tell me that their language and Arabic are often mutually comprehensible without need for interpretation. Modern Maltese also includes great numbers of Indo-European loan words that, for historical reasons, come mostly from Italian and English. Written Maltese uses the Latin alphabet but the pronunciation of the letters is often quite unusual in order to allow full expression of Semitic and uniquely Maltese cadences in speech. Thus Hagar Qim is pronounced something like Hajar-iim, Tarxien is Tarshien, Mgarr is Umjaar, Zebbug is Zebooj, and the potentially unpronounceable Mnajdra reaches the ear as a soft and mellifluous Munaidra.

  Floorplan of Hagar Qim temple. Based on Evans (1971).

  About 8 kilometres south-west of the Hypogeum, but separated from one another by less than a kilometre, the temples of Mnajdra and Hagar Qim stand on Malta’s south coast, overlooking a spectacular panorama of deep blue sea and Mediterranean sky in which the craggy little island of Filfla – blasted to smithereens by centuries of artillery practice – floats like a mirage. By night, roofless (though thought to have been roofed in antiquity),34 they gaze up at the wheeling constellations and take an interest in the peregrinations of the moon. By day they use a variety of shadows, peep-holes and cunningly contrived alignments to follow and to record the path of the sun.