Sicilian Atlantis or Telepylos?

Revealed on Twitter last week was the discovery of a “lost city” just 40 km off the coast of Sicily! This article is why that idea is sunk. So to speak.

When I first decided to write this, I expected to show a few screenshots of Twitter posts and comments about a crazy Atlantis claim. But the more I read about under sea mapping methods and the caveats that go with them, the more I realized that there are a lot of these “anomaly hunters” out there looking at Google Earth and various GIS datasets. And this could serve as a reference point on how to look at undersea “city” claims.

Going by the Twitter handle, @TheImmortician, Andre Chaisson has identified a set of “anomalies” using bathymetric data from GEBCO and EMODnet. He says he’s on the hunt for Atlantis but has come across a “submarine city” much larger than Atlantis.

Instead of Atlantis, he’s calling this the mythical city of Telepylos, from Homer’s Odyssey. That city of the Laestrygonians, a race of cannibals and giants who lived in a rocky stronghold in the far north.

Why not Atlantis?

Well, according to Chaisson, it’s too big. Therefore it must be Telepylos. Chaisson claims he’s located a city “surrounded by a deep canal, 3 stadia wide in most places, enclosing an inner city are of 80 square kilometers–nearly twice the size of Atlantis.”

He goes on to say, “while Homer never detailed the city’s [Telepylos] interior layout, its dimensions, aligning perfectly with whole-number units, make its anthropogenic origin undeniable.” He’s also duly impressed that the site aligns “perfectly with the cardinal directions.”

In Homer’s Odyssey, only 45 of Odysseus’ men escape in a single boat of what was a 12-ship fleet until they encountered the Laestrygonians. This is because most of the ships were moored in the harbor and easy targets from stone-wielding giants, some of whom waded into the harbor and speared their sailors.

Yes, TheImmortician of Twitter, accepts a fictional story written almost 3000 years ago as an eyewitness account. Wait until he hears about the time Bruce Wayne traveled from Gotham City to Hogwarts where he learned the ways of the Force!

But Why Telepylos?

He doesn’t say. At least not on Twitter (that I saw). He does, however, “publish” an essay on Academia, where all the good (and bad) pseudoarchaeologists go to publish their works. There he writes, [Homer’s] “vivid depiction [of Telepylos] includes towering cliffs, a narrow harbor entrance, and dramatic landscapes, suggesting a location of strategic maritime importance.”

He also remarks both on Twitter and his essay that the precise adherence to the “stade” as a unit of measure is also evidence. Which version of the stade, you ask? One is left to assume his version. Chaisson writes on Twitter:

In his essay he writes:

Most compellingly, the dimensions align perfectly with the ancient Greek unit of measurement, the stadia (176.4 meters), which was widely used in antiquity. This alignment not only suggests human origin but also potentially links the site to ancient Mediterranean civilizations that employed this unit. The consistent use of the stadia across the site further supports the notion that these structures were intentionally constructed, adhering to a standardized system of measurement. This alignment would be improbable if the features were purely natural formations

Andre Chaisson, https://www.academia.edu/125868234/Telepylos_Gateway_to_a_Lost_World

To his assertion above, I’d answer “yes” and “no.”

Yes, the consistency in alignment and measurements suggest human origin.
No, they don’t link to either the “stade” or ancient Mediterranean civilizations.

There are some problems with Chaisson’s claims. Most are pretty obvious but here are the ones I think stand out. I’m pretty sure I’m not the only person to think so.

The Stadion

He’s right that the stadion or stade was widely used as a unit of length in antiquity. The problem is, there are many definitions of what constitutes a stadion depending on region and period of time.

It was Lev Vasilevich Firsov that established an empirical version of the stadion based on 81 distances provided by ancient Greek mathematician Eratosthenes and Strabo, geographer in Asia Minor. Firsov used known measurements from these two historical figures, averaged the results, and arrived at 157.7 meters.

There are a handful of known versions of the stadion used in antiquity, but Firsov’s averages of the Eratosthenes/Strabo stadion seems the more reliable. The table below lists some of the versions of the stade used in antiquity, which depend on the the various versions of the Greek foot (pode) in modern millimeters.

Chaisson appears to use the Olympic stade as described by Engels (1985) rather than the description Hoyle uses, which is really the Attic stade. I’m assuming that Chaisson read somewhere that the Olympic stade was used in Southern Italy and that might be the reason for his reliance on it for this location underwater.

But why not use the Doric stade, which was used in Sicily and Greece as well as Persia? Or the Itinerary stade, which is consistent with the Stade of Eratosthenes when averaged?

Whatever version used, distance measurement had a flaw in that there was no empirical standard. And this created problems when one considered the curvature of the earth and, perhaps more importantly, the indirect nature of travel in antiquity. Shorelines have curves, bends, and jagged edges. Interior journeys had topography like mountains, hills, and bodies of water.

The Burden of Proof

Chaisson used QGIS to load multiple bathymetric data from GEBCO and EMODnet that was georeferenced to a basemap. Then I’m sure he used a measurement tool to measure distance. All well and good. I double checked and got pretty much the same measurements, give or take 50 meters here and there. His claim is that this consistency to a stade means this must have been an ancient human site and that it also means it cannot be digital or data processing artifacts.

But is that how the burden of proof really works? Not at all.

When you look at his measurements that show the “stade” of 176 meters, his start/end points are really pretty arbitrary. He could have taken the initial measurement from the outside of the so-called “canal” but he chose the inside. He could have taken measurements from several “points” that are really just moving the bullseye to fit the bullet-hole, as it were. This is known as the sharpshooter fallacy. There’s a lot of focus on points that work for his hypothesis, but no mention of features or potential measurements that do not.

But if you examine Chaisson’s bathymetry closely, you see that the rectilinear feature set is also roughly the length of the combined research area for Micallef, et al (2014), as they studied the mass movement dynamics of the Malta-Sicily Escarpment, which is adjacent to Chaisson’s “Telepylos.”

As one might imagine, this is already a research location so there is an inordinate amount of geophysical data like bathymetric lidar, side scan sonar, sub-bottom profiling, and magnetic resonance. And probably some I didn’t mention. I’m pretty sure Micallef et al also used gravity core sampling which is a really cool method of combining geophysics and geochemistry with physically retrieved core samples.

Interpolation and Data Processing Artifacts

The rectilinear feature set that Chaisson is calling an “ancient coastal city” is no accident. It’s a combination of varied resolution data sets that are giving the false impression of undersea geologic or man-made features. Transects of various data collection methods, with varied results and resolutions are potentially being overlaid. And there are definite signs of interpolation in the results. Note the areas I circled in yellow (above), which show identical patterns.

In side-scan sonar, data are interpolated when filling in “Not a Number” (NaN) values. These values happen because of missing or invalid data readings. Interpolation is also used to fill gaps in multibeam or Shuttle Radar Topography Mission (SRTM) data. This is the way seamless shaded relief maps of the ocean floor are created. Sometimes second or even third datasets are allowed to show through a layer mask when the data are available.

But when you start merging different surveys into a single dataset or data output, the results can sometimes get messy and artifacts can be more noticeable than the topography. Typically, the resolution of multibeam data and smoothing techniques are used to remove artifacts.

I could go into a lot of detail on various data collection methods and the types of artifacts. But I suspect there are a palimpsest of them on this section of the seafloor due to the interest in the Malta-Sicily Escarpment and the tectonic activity that seems to create large movements of sediment there.

The Actual Sea Level

Chaisson argues that the Mediterranean sea level at the time of the Last Glacial Maximum was held over through the Younger Dryas period and on to 8600 BCE when it was finally inundated by the rising Mediterranean Sea. This is contrary to what is actually supported by probably a few dozen lines of evidence that say otherwise. Chaisson doesn’t really say why he revises it though his occasional reference to Atlantis suggests that he needs the rise in sea levels to be more recent to fit some notion about this fictional city.

Or maybe he sees 20,000 years ago as to early a date for Odysseus to be traipsing around the Med with a fleet of ships. Because the Neolithic period of Greece is more reasonable than the Paleolithic? At any rate, the red dot in each frame of the graphic above shows the region of his “Telepylos” at various periods of time. Among the lines of evidence that support the current understanding of sea level chronology in the Mediterranean Sea are:

• Dated marine terraces and shorelines.
• Coral reef formations and ledges with shellfish remains that are dated.
• Sediment cores that include pollen, seed, and phytolith samples that are dated.
• Incised valleys from rivers prior to inundation.
• Glacial moraines — that accumulation of sediment left behind from melting glaciers.
• Paleoclimate records from cave deposits

Not the First to Notice

There’s actually another person, Luigi Usai, that found this set of data processing artifacts. He put it on the internet by at least 2022. I’m guessing the data sets underlying the EMODnet bathymetry isn’t much older.

Conclusion

While it’s frustrating to see so much imagination and fantasy go into creating a narrative about a lost city–one that was a fictional city to begin with, I keep reminding myself this is coming from a person that has a true interest in ancient history and perhaps archaeology. As cool as a “lost city” might be, the methods of under sea mapping and the way the technology is being used to find genuine undersea settlements and habitation zones is way cooler.

Sure, these aren’t cities of “giant cannibals” or people of lost high-technology or whatever else some people are hoping to find. I’ve been a bit dismissive of much of Chaisson’s work. Though I know some who would say “not harsh enough.” I think he genuinely wants to understand the past, which is good. But I also think he has some pre-conceived ideas about what the past is and, like many, he’s more interested in those data that are supportive of these ideas. His essay on Academia shows it. For all the words he typed–more than I have here–he only includes three references: Micallef et al (2014), Chow (1959), and Homer (~700 BCE).

I genuinely think those who believe in Atlantis and Lost Civilizations have an honest interest in the human past. Maybe we just need to show them that the human past can be a cool story even if these ideas turn out to be fantasy.

References and Further Reading

Engels, Donald (1985). The Length of Eratosthenes’ Stade. American Journal of Philology, 106(3), 298–311.

Foglini, Federica, et al (2015). Late Quaternary Coastal Landscape Morphology and Evolution of the Maltese Islands (Mediterranean Sea) Reconstructed from High-Resolution Seafloor Data. In, J. Harff, G. Bailey, and F. Luth (Eds), Geology and Archaeology: Submerged Landscapes of the Continental Shelf. Geological Society, London.

Firsov, Lev Vasilevich (1972). “Eratosthenes’ Calculation of the Earth’s Circumference and the Length of the Hellenistic Stade,” Journal of Ancient History, 121, 154-75.

Gulbekian, Edward (1987). The Origin and Value of the Stadion Unit used by Eratosthenes in the Third Century BC. Archive for History of Exact Sciences, 37(4), 359–363.

Lehmann-Haupt, C. F. (1929). “Stadion.” In, Friedrich von Pauly (Ed.), Paulys Real-Encyclopadie der calssischen Altertumswissenschaft. Stuttgart.

Micallef, Aaron, et al (2014)The Malta-Sicily Escarpment: Mass Movement Dynamics in a Sediment Undersupplied Margin. In, Sebastian Krastel, et al (Ed), Submarine Mass Movements and Their Consequences. Springer, New York.

National Oceanic and Atmospheric Administration (2024). Standard Ocean Mapping Protocol (Final). Federal Register.

Patterson, Tom (2014). Mountains Unseen: Developing a Relief Map of the Hawaiian Seafloor. Cartographic Perspectives, 76(76), 5-18.

Wesenberg, Burkhardt (1975). “Zum metrologischen Relief in Oxford.” Marburger-Winckelmann-Programm, 16.

Yang, Fanlin, et al (2007). A Post-Processing Method for the Removal of Refraction Artifacts in Multibeam Bathymetry Data. Marine Geodesy, 30(3), 235-247.

Special Notes

I usually attempt to use public domain and creative commons images, though I do occasionally use images within the terms of fair use for educational purposes. If you find an image that belongs to you and object to my using it, contact me and I’ll make the necessary adjustment(s). I occasionally use affiliate links, particularly in reviews. These links are marked “no_follow” and “sponsored.”