Dispatch 12: Analysis Begins

October 6, 2000

[Note: Nationalgeographic.com does not research or copyedit dispatches.]

The following dispatch was supplied by Dr. Fredrik Hiebert, Chief Archaeologist of the Black Sea 2000 Expedition.

Since the report of the discovery of the habitation site (site 82) and its initial reporting the following has happened: We were granted permission from the Turkish authorities to move to a further stage of investigation—sampling and actual testing of some of our observations. Thus, when we went in and investigated the site further, we found that suggested mud slabs were actually stone, and that the suggested stone artifacts were actually wood, etc. We now have samples from the site which we collected. (They are in my refrigerator at the moment.) I am sending them for scientific analysis.

Scientific Method in Action

We are in the process of learning about this flooded coastline. Every exposed rock and stone we see was clearly part of the landscape when this was dry land. It is very exciting to be the first to see this landscape since it flooded. With the discovery of this site, we now are in a position to fit it into our larger picture of the ancient cultures of the Black Sea area.

Archaeologists use the scientific method. Our assignment of this site as an ancient habitation site is a hypothesis. To test this hypothesis, we first need to have a model (something which is only partially known or unknown) in our heads of what the objects we could see were, and how they came to lie together on the floor of the Black Sea.

While the video images taken from the ROVs look just like ancient structures that we had seen on land, we had to consider other scenarios that could have produced a similar arrangement. Other interpretations that we must consider are that we were observing a natural geological outcrop that had accumulated wood over the years, a broken-up shipwreck, or even the debris from a modern depth charge.

Our job as scientists is now to devise tests that would allow us to exclude these interpretations and/or to offer support to our hypothesis the site was a human-built and used archaeological site. Using still cameras, we compiled a photographic map of all the features that we could see. Armed with this tantalizing picture, we approached the Turkish Ministry of Culture for permission to sample the site for scientific analysis. We were granted permission and the Northern Horizon returned to the site location. The opportunity to visit this submerged site was the thrill of a lifetime.

Twist of the Depths!

We jury-rigged a special recovery device for Little Hercules—an experimental addition that was put together on the boat.

It worked! We watched on the video while a flexible probe touched and lifted some of the objects. The texture, density, and shape of the objects on the screen became clear as Little Herc circled only a few feet above the Black Sea floor. For now, this is the closest that we will ever get to actually “seeing” the site with our own eyes. Later, a sampling bag allowed us to raise a few carefully chosen artifacts and other objects to the surface for examination and testing. In addition to these objects, we collected samples of the mucky deposit on the surface of the site.

What did this device allow us to learn? First of all, we found that the sedimentary blocks that appeared to be clay are actually stone, not mud. Second we retrieved several of the artifacts, and they turned out to be wood, not stone.

Many Neolithic and Bronze Age buildings have such stone foundations, just as many have wood and mud type architecture. We have to understand that people in the past used many types of building techniques, depending on the situation (I like to use the Three Little Pigs concept) during the same period. A culture can build buildings out of sticks, brick and stone. It is quite likely that the stones could have made a solid foundation for a building that was partially constructed under the ground, like a shallow pithouse. This new hypothesis is based upon learning from our testing and revising our model. Following this new scenario, the original surface might have been eroded by water to leave the surface that we see today.

Our closer look at the wooden artifacts confirms our first impression that they are the result of shaping by humans. They have smooth, symmetrical shapes and unmistakable traces of drilling to produce the holes visible on the videos. We don’t know a lot about ancient wooden tools since wood is rarely preserved in any form, but we can compare these objects with the tools for woodworking that have been found at sites on land, and with wooden objects produced by craftsmen in today’s traditional societies.

Scientific Analysis Begins

With samples in hand, we rapidly developed a plan for the analysis of materials that might represent an ancient land environment but which have also been in a cold, wet, Black Sea environment for an unknown number of years. Once we had consulted with an expert on conserving waterlogged wood samples, we proceeded with our highest priority, dating the wood samples using the supersensitive accelerator mass spectrometer (AMS) radiocarbon technique. Only tiny samples are used for this technique. We chose fragments of six worked and natural wooden objects to submit to the laboratory. (Extracting these fragments did not destroy the objects.)

One test of the integrity of an archaeological site is whether a series of radiocarbon dates fall in a cluster in time. If we are told that the dates are quite recent, or represent a wide range of dates, we will have to reconsider our hypothesis of how the site was formed.

Dates will be reported to us in a matter of weeks. In the meantime, we are reaching out to a variety of scholars who we hope will be able to help us understand what an undersea site might look like. Using a new, refined model of what to expect if site 82 is an ancient structure, we plan to study the artifacts and deposits at a number of levels. The muck will be analyzed to determine its microscopic structure, its phosphate content (a signal for human occupation and trash), its possible content of charred or waterlogged seeds, bones, and shells, and any occurrence of pollen or silica cell wall remains of crops or other land plants. The wood itself will be identified and compared to modern and prehistoric forest cover in northern Turkey. At the same time, we will be examining the contours of the Black Sea floor to most precisely locate site 82 in its geographic context, both during the period it would have been on dry land, and since the time it must have flooded.

Our visual “fix” on site 82 using remote video technology was immediate and compelling. Now, we are excited to be able to test and evaluate our impressions, and to tie these finds to our other research projects on ancient Black Sea peoples.

—Dr. Fredrik Hiebert

Go to 1999 Dispatches