We know less about the ancient world than we think we do
On 15 June 763BC, a near total eclipse of the sun was visible over a swathe of the Near East. As luck would have it, the event was noted in the official list of Assyrian high officials. This record provides the earliest absolute and uncontroversial date in ancient history. Using lists of kings and the chronicles of events, historians have counted the years back from this date to construct the chronology of ancient history.
Radiocarbon analysis (which measures the decay of carbon 14, an unstable isotope) and the predicable styles of pottery found in digs both provide corroborating evidence. Dating the layers of archaeological remains from the artefacts found within them is called stratigraphy and can yield quite precise results. The vast amount of pot shards that has been unearthed allows archaeologists to use statistical methods to screen out random noise and anomalous samples that have found their way into the wrong strata.
Of course, pottery and radiocarbon methods need to be calibrated to produce absolute dates. This has been done using samples of wood whose age can be determined by matching patterns of tree rings, a technique called dendrochronology. We can count back sequences of tree rings from the present day, all the way to 2000BC. By carbon dating the oldest samples of wood, we can tie the tree ring record to the results from carbon 14 decay.
By 1990, all these clues had yielded a multi-dimensional jigsaw which fitted together to almost everyone’s satisfaction. There were a few heretics like Peter James, who suggested in his book Centuries of Darkness that the conventional chronology included two hundred additional years around 1000BC. Thus remains that were conventionally dated to 1050BC actually occurred in 850BC. Although James’s book is an excellent read, it fails to convince.
Nonetheless, it has now turned out that the conventional chronology was not as secure as everybody else thought. While James was convinced ancient history was two centuries too long, new evidence has begun to pile up in the opposite direction: it now looks like the conventional chronology is up to 150 years too short. To put it another way, a cataclysm that everyone thought occurred in 1500BC actually happened before 1620BC. The event in question was the massive eruption of the island of Thera in the Aegean Sea.
Conventional chronology dated the end of Minoan age in Crete to 1450BC. Archaeologists assumed that the Thera eruption (on the modern island of Santorini) and its resulting tsunami had destroyed the Minoan fleet leaving them vulnerable to raiders from the mainland. Certainly, the havoc wrought by the volcano can clearly be seen across the Eastern Mediterranean. When Thera exploded, it blasted 60 cubic kilometres of rock into the atmosphere which settled over Asia Minor.
The resulting layer of ash and pumice is used to date the sites where it is observed. And the eruption had other effects. Sulphur dioxide released by the volcano spread across the northern hemisphere and fell to earth as acid rain, or more significantly as acid snow. At the poles, not all of that snow has yet melted and, from the 1990s, it provides a new strand of evidence to date the eruption.
Ice cores, drilled from the icecap of central Greenland, record the depth of each annual snowfall. The ice holds within it information on the constitution of the atmosphere going back tens of thousands of years. Like tree rings, each layer can be counted so as to give an absolute rather than relative date.
Big volcanic eruptions show up as spikes in the sulphur-content of the annual fall of snow: Krakatau in 1886; Tombura in 1815; Vesuvius in AD79. Despite the presence of literate civilisations in Egypt, the Levant and Babylon, no written record of the Thera eruption exists, but the ice cores should overcome that deficiency and provide an absolute date for the cataclysm.
Actually, the fact that the Thera event went unrecorded is less surprising than it seems. Mankind has been remarkably unobservant of enormous volcanic eruptions. An event in 1257AD, less than 800 years ago, is indelibly imprinted into both the Greenland and Antarctic ice cores. It was greater in size even than Tombora and thus the largest eruption in the last ten thousand years. But remarkably, no one knows where it happened. Only in 2012 has Mt Rinjani in Indonesia emerged as a likely candidate. Another big eruption, as recent as 1809, remains unidentified.
By 2000, the Greenland ice cores had revealed that Thera could not have happened when everyone thought it had. The most likely anomaly in the ice dated from 1640BC, but this turned out to be from a volcano in Alaska. At the same time, carbon dating an olive tree buried in the Aegean eruption yielded a date of around 1620BC. Sulphur traces in the ice have been found that correspond to this date, although they are not as strong as might be expected.
Now, the dendrochronologists have piled in. The Thera eruption would have caused unusually cold weather which stunted plant growth across the globe. Evidence from bristlecone pines in the western United States, oak trees in Ireland and Swedish pines all point to a cold snap in 1627BC. This is consistent with what we’d expect from a big volcano blowing its top in the Mediterranean. Evidence from the Antarctic ice cores should be in shortly, but for a northern hemisphere volcano, this is unlikely to be conclusive.
The lack of a definitive date for the Thera disaster is frustrating, but we can now be reasonably sure it occurred 120 years earlier than thought. The implications of this for ancient history are immense. The chronology of the New Kingdom of Egypt was thought to be rock solid. Finding that they need to find room for a dozen more decades has been too disconcerting for Egyptologists to tackle so far. There is a good chance that the extra years belong in a period after the well-documented New Kingdom called the Third Intermediate Period.
For historians of Babylonia, the crisis has been less existential. Absolute dates for the second half of the second millennium are based on ancient observations of the planet Venus. We know from modern calculations that a particular configuration of Venus recorded during the eighth year of the reign of a certain King Ammisaduqa must have occurred in 1702BC, 1646BC, 1582BC or 1550BC.
Other events in Babylonian history, such as the reign of King Hammurabi (famous for his law code) and the sack of Babylon by the Hittites are arranged around whichever absolute date is most convenient. That some of these possible Venusian dates differ by 120 years, about the same length of time that the Thera eruption has been moved back, is highly suggestive to say the least.
So, where does all this leave biblical chronology? That remains very unclear. But the redating of Thera shows that we know a lot less about when things happened in the ancient world than we thought we did.