A little over 3,500 years ago, an emporium at Çeşme-Bağlararası, Turkey, had grown into a prosperous port of trade on the Aegean coast. But the fortunes of this settlement changed abruptly in the immediate aftermath of what may be the greatest natural disaster ever witnessed by humans. This unfolded when the volcano forming the ancient island of Thera – modern Santorini – erupted with sufficient force to leave a global fingerprint. Volcanic particles ejected into the atmosphere can be found in polar cores, while the cooling effect created by such material made its mark on tree growth rings in North America. Closer to Thera, ash carried by the wind carpeted sites lying hundreds of kilometres to the north and east. Çeşme-Bağlararası was among them. By the time this grey rain drifted down, though, its inhabitants had already felt the fury of the volcano. Recent archaeological work at the site has revealed that a tsunami slammed into the settlement. This was just the first in a series of calamities to befall its occupants, shedding poignant new light on the human cost of an eruption that transformed life in the Aegean.
An unexpected settlement
Puzzling out the anatomy of the disaster that engulfed Çeşme-Bağlararası was only possible following archaeological analysis that was as ingenious as it was meticulous. Work at the site was carried out as part of a much larger programme underway in the area. ‘It forms part of our Izmir Region Excavations and Research Project,’ says Vasıf Şahoğlu, head of the Ankara University excavations. ‘This was initiated by my late professor and mentor, Hayat Erkanal, in the early 1990s. The aim was to understand the prehistory of western Anatolia, because study of the area has always been dominated by an interest in the Classical period. We knew there were important prehistoric cultures in the area, so we started excavations at a series of sites, to open up a wider perspective on the development of the region. At Çeşme-Bağlararası, we have been investigating Bronze Age activity from roughly 2500-1300 BC. The site was discovered totally by chance by Çeşme archaeological museum in 2001, after ancient pottery was found at a construction site. This lay at the heart of the modern town – in the street just behind the marina, at what is basically the second-biggest summer tourist destination in Turkey after Bodrum. In the early days this made excavation challenging, as it’s really valuable land, so this part of the project was quite an experience.’
‘Even though we could only work in limited areas, we now know quite a lot about the site. Occupation started in the Early Bronze Age, at the middle of the 3rd millennium BC. There are various gaps in the habitation history, but in most cases we think this is connected to how the site developed, rather than them being genuine absences. Long-lived sites in the region normally formed mounds, because ground level rose as new phases of activity were superimposed on top of the remains of earlier settlements. But Çeşme did not grow vertically in the same way. Instead, it seems to have shifted horizontally, probably due to changes in the coastline, or the course of rivers that flow either side of the settlement. So, the Middle Bronze Age site, for example, was not built directly on top of the Early Bronze Age one; instead it lies next to it. We think many seemingly missing periods of activity lie elsewhere under modern Çeşme.’
‘This movement along the coast is important for understanding the part of the site that was hit by the tsunami. Our early work there between 2002 and 2005, alongside Çeşme museum, unearthed a really nicely preserved late Middle Bronze Age site. We found streets, houses, and even a facility for making wine – something that is unique for Anatolian archaeology. The excavations then stopped, before being reinitiated under my direction from 2009 to 2019, with the University of Ankara and a permit from the Ministry of Culture. And what we found was very interesting. In this part of the site, the remains were much more disturbed. We realised that these structures had been abandoned when the settlement shifted along the coastline, and at some point after that the ruins were partly filled with large quantities of debris and sediment, containing finds belonging to a later phase of occupation. So then, of course, the question was: “How did this happen?”’
An ash layer up to 2cm thick that seemingly lay on top of this mass of sediment and debris was equally intriguing. Although Çeşme is 227km from Thera, that is still well within the range of sites where large quantities of ash from the eruption have been detected. Could the Çeşme deposit also be explained by the volcano? There is still passionate debate among scholars about when precisely the eruption happened, with some favouring a date in the mid to late 17th century BC, while others prefer the 16th century BC. By either timescale, though, the pottery found among the detritus in the ruins was correct for objects lost at around the time that Thera blew. Even better, as different volcanoes produce material with different geological signatures, there was potential to secure a definitive link with the eruption.
‘I was invited to take cores at Çeşme-Bağlararası as part of a PhD looking at coastline changes,’ remembers Beverly Goodman-Tchernov, now head of the University of Haifa Marine Geosciences Department. ‘My background is sedimentology and geoarchaeology. Some people like ceramics, but I’m into everything that surrounds the ceramics, so I like sand, dirt, mud, clay; everything that allows me to reconstruct the environment that the site existed in. As it turned out, I didn’t have time to complete the Çeşme cores for my PhD, but one thing did come out of it. As I was preparing a few samples, volcanic pumice bubbled up, and I thought, “That’s cool.” There’s a laboratory in Austria that invites people to submit samples of ash or pumice, as they are creating a database of material from Thera. So, in around 2005, I sent it to them. Disappointingly, the analysis came back as “not Santorini, no match.” As it was negative, I left it there.’
‘Years later, when Vasıf was directing excavations and found this really nice ash line, he did the same thing. This time the laboratory said, “It’s not a good match for Santorini, but we have a perfect match at Çeşme!” So Vasıf emailed me to ask about the pumice sample. We both thought the result for his ash layer was surprising and discouraging, because the dating seemed perfect. Anyway, I went and sampled the whole exposed ash section at Çeşme. We decided the best approach was to go forward and keep trying to figure out where the ash had come from. It was a slow process, but we continued thinking about it over the years, we got radiocarbon dates, and we persisted. Eventually, the lab decided to try separating out the bigger pieces of volcanic material in our sample from the smaller pieces. After all, while the fine ash can get into the upper atmosphere and travel long distances, that doesn’t happen with larger material. That was when everything came together. We saw the Santorini signature in the fine ash, while it turned out the larger pumice came from eruptions that were both much older and much more local. So, there was the solution – ash from the Thera eruption was mixed in with volcanic material that had already been present at Çeşme and had nothing to do with Santorini.’
Now that the ash layer seemingly capping the fill could be linked to Thera, what about the fill itself? ‘One way to explain how a thick deposit of material contemporary with the Thera eruption ended up in the ruins of an earlier part of the site,’ says Vasıf, ‘is if it was washed there by a tsunami. You need to imagine buildings that were still visible, but deserted, after the settlement moved along the coast. These abandoned structures included a stretch of fortification wall, which may have already been partially dismantled to construct new houses. Our scenario would see the tsunami striking the weakened rampart and demolishing part of it. Houses right next to the fortification were also partially destroyed and filled with debris from collapsing walls and other material, which travelled about 5 or 6m into the ruins. In one part of the fill, within the debris, there were the remains of a young adult male human.’
‘Because we can see that the damage is coming from a single direction, and because nearby walls could sustain very different levels of damage, what we uncovered is a poor fit for something like an earthquake. Once we had the idea of a tsunami, it became possible to reconstruct a provisional narrative of what happened when Thera erupted. Ash was sent up into the atmosphere, and a tsunami was also triggered. It was the tsunami that reached Çeşme first, with the ash then falling on top of the devastation left in the water’s wake. That all seemed to fit neatly. But when I was writing up the site, I studied the finds very carefully to make sure of the chronology. This produced a surprising result: date-wise, the deposits directly above the ash contained exactly the same sort of archaeological material as the deposits directly below the ash. By that reckoning, they should all belong to the same period. I spoke to Beverly, who had also noticed something unexpected about the soil above and below the ash.’
‘This brings us to one of the major issues for tsunami sediments at an archaeological site,’ says Beverly, ‘which is how do you recognise them? Obviously, you’re most likely to encounter them on coastlines. And what do you have at a coastal archaeological site? Well, you have coastal materials. Because the elements of the tsunami deposits can’t be anything other than what already exists in the landscape, you could walk past a 10m-thick wall of wave-deposited debris and not realise what it is. If you do the analysis, though, you start seeing things that have come from deeper water and been transported inland. So, I examined our apparent tsunami deposit to see if it contained those kinds of marine components.’
‘One good marker is what’s known as foraminifera, which I think of as the heroes of the ocean that no one has ever heard of. They are single-celled organisms, many of which create calcium-carbonate shells or skeletons that can preserve quite nicely, making them wonderful indicators of the local environment. Some live in the water itself, others are bottom dwellers, and there are tens of thousands of species. On average, they are about the size of a grain of sand. A good example of why they are so handy is that if you find on land a whole bunch of well-preserved foraminifera that normally live at a depth of 100m underwater, you can be certain that they didn’t get there on their own. Instead, there has to be some kind of transport mechanism. What is more, research has shown that when these foraminifera are carried in rapid events like a tsunami, there is often so much deposited material and so much decomposition that chemical processes will make them stain in strange colours that you don’t normally see. So, they are incredibly useful.’
‘Initially, I was frustrated with the sediments at Çeşme, because we were working on the belief that the ash layer was the terminal marker for the eruption. After that, everything was calm, everything was quiet, it was over. But when I was looking at the samples and doing the analysis, I found myself doubting whether we could really have a tsunami at all. Not because there were no marine components – there were plenty. The problem was that they were present in the deposits both above and below the ash layer. A noticeable shift between the sediments containing these marine markers and those with far less did happen, but it was much higher up in the stratigraphy. It was Vasıf who said, “What if you’re comparing the sediments to the wrong thing? What if they all match because what we’re seeing is a series of tsunamis occurring below and above the ash layer?” Once he said that, and once I looked at the data again, it was the “wow” moment when everything started to fit together.’
While the possibility of multiple tsunamis carries horrifying implications for the Bronze Age inhabitants of Çeşme, it also provided the key to unlocking how the archaeological deposits had built up. Re-examining the deposits both above and below the ash layer revealed that there were distinct packages within them, which all carried the markers of a tsunami event, but at the same time were different from one another. These discrepancies could involve subtle differences in content, as well as more obvious variations, such as the size of the objects that had been deposited. This last point can be linked to various factors, including when structures were collapsing, but as a rule of thumb it is the size and speed of a tsunami that determines how big the material it can shift is. Bringing all of these variables together allowed four distinct layers to be identified, two above the ash layer, and two below. This in turn indicates that Çeşme was on the receiving end of at least four powerful waves.
If a series of tsunamis made landfall at Çeşme, the next question is how far apart in time they were. It is here that the archaeological evidence from the site has perhaps the most important implications for our understanding of how the Thera eruption played out. Volcanologists and geologists have long studied deposits from the eruption, and when the Çeşme team looked into their work more closely, they found a school of thought had developed that divided the eruption into four phases. In which case, could there be a link between the four eruption events, and the four powerful waves that struck Çeşme? The archaeological evidence certainly suggests that while the gaps in time between the tsunamis varied, they can all be reasonably linked to the volcano. Radiocarbon dates from material among the detritus gives the earliest possible date for the series of inundations as 1612 BC. While this would fit both of the competing chronologies for the Thera eruption, it is perhaps a better match with the catastrophe occurring in the 16th century BC.
Piecing together the events that engulfed Çeşme after the eruption started not only exposes the impact of the natural disaster, but also allows a glimpse of survivors’ early responses to it. The first tsunami to reach the site was the one that breached the ruinous fortification wall and left a human victim buried among its debris. This included plenty of massive material, testifying to the power of the tsunami. Careful study of the deposit showed that in the aftermath some ash fell, but the resulting seams are only a few millimetres thick and do not form a continuous layer. A reasonable conclusion is that there was only a short period of time for ash to accumulate – probably a matter of hours – before the next tsunami struck. The layer associated with this second event contained less bulky objects, but still spread some way into the ruins. It was after this second incursion that the continuous ash layer once thought to overlay the one and only tsunami built up. Because more ash settled, it is likely that there was a longer delay before the next tsunami arrived, probably at least double the length of time between the first and second.
When the third tsunami arrived, it was smaller than its predecessors. It also carried charred or still-burning animal bones and other material, presumably from fires started as a consequence of the eruption, painting a picture of widespread devastation. What happened after that tsunami struck is particularly fascinating. There is not much ash on top of this set of debris, but this time it does not seem to be a product of only a short interval between inundations. Volcanologists tend to see the ash ejection as primarily a feature of the first phase of the Thera eruption, so perhaps by the time the third tsunami arrived, the ash fall had more or less concluded. Alternatively, the wind direction may have changed, resulting in the ash being sent elsewhere. A third possibility is that the fine particles sent into the atmosphere are likely to have brought about increased rainfall in the days, months, and years following the eruption, which may have washed much of the ash away.
There is ample evidence for human activity following the third tsunami. Numerous irregular pits and trenches were dug down into the material that had been washed onto the site. Some of these pits are about the right size for a person, and so could be attempts to remove the dead or injured from the debris. If so, the human skeleton found by the archaeologists was the only victim in this area who was not found and recovered in the immediate aftermath of the catastrophe. Other holes seem to have been geared towards salvaging masonry or other material needed to make repairs elsewhere. Naturally, rescue and repair efforts would only start after survivors felt safe enough to return to the damaged areas. That also suggests a lull after the third tsunami, perhaps of somewhere between a day and a week or two. It was only then, when recovery efforts were under way, that the fourth tsunami struck. In terms of power, it was on a par with the first one. Not only did it fill the pits and trenches with sediments and rubble, but it also mounded a thick layer of debris on top of the material left by the other inundations. No more pits were dug after the fourth tsunami, and there is no evidence of any human activity at the site for several centuries.
Counting the cost
‘Probably there had been a sense of “it’s over” after each tsunami,’ says Beverly. ‘But after the third one, when there really was a long period of quiet, it is horrifying to imagine what it must have been like for the survivors when that fourth tsunami came in. And of course, we’re only looking at a very small area at Çeşme. If you take the picture that is emerging there and extend it through the Aegean, it really brings a new dimension to our sense of the impact of the eruption. Not only do we have destruction of settlement and loss of life on Thera itself, but we also have tsunamis that were felt at a regional scale. To that you can add the evidence for climate change on a global scale for a couple of years, with the volcanic particles in the atmosphere bringing cooler, wetter years. That in turn will have brought changes to crops, as will the salt water washed onto fields by the tsunamis, creating an economic impact. When you start to bring all of this together, you see how a natural event became a social disaster. No one was left untouched by it.’
This raises the question of whether the volcano could also have had a hand in the mysterious collapse of the great Bronze Age Minoan civilisation centred on Crete. ‘This has been a source of debate for a long time’, says Vasıf. ‘Recent articles have perhaps tended to undermine the impact of the volcano on the Minoans. We know that their palaces were damaged at around the right time, but some were renewed afterwards and continued in use. Another way to look at this, though, is to focus on how the Minoans were almost entirely dependent on maritime trade for many of their resources. Imagine what that meant. All of the boats at sea when the eruption happened must have been destroyed. All of the coastal and harbour settlements through the Aegean were wiped out. Our research forces us to see the eruption as a much greater tragedy than has previously been suspected. Of course, people tried to rebuild afterwards and keep things going, that’s what humans do after disasters, but ultimately the Minoans had been severely weakened and couldn’t rekindle their past glories. Eventually, that handed the Mycenaeans an opportunity to emerge from mainland Greece and take over. I think that the volcano had a major impact on the ability of the Minoan civilisation to sustain itself.’
FURTHER READING For more information about the results of the research into tsunamis at Çeşme-Bağlararası, see V Şahoğlu, J H Sterba, T Katz, Ü Çayır, Ü Gündoğan, N Tyuleneva, İ Tuğcu, M Bichler, H Erkanal, B N Goodman-Tchernov (2021) ‘Volcanic ash, victims, and tsunami debris from the Late Bronze Age Thera eruption discovered at Çeşme-Bağlararası (Turkey), PNAS 119: 1. CWA is grateful to Vasıf Şahoğlu, Beverly Goodman-Tchernov, and James Fattal.