Lost civilizations and the science of knowing what we don’t know

The most extraordinary claims need the most extraordinary evidence

If you find value in #ComplexityThoughts, consider helping it grow by subscribing and sharing it with friends, colleagues or on social media. Your support makes a real difference.

→ Don’t miss the podcast version of this post: click on “Spotify/Apple Podcast” above!

---

→ At the end of the post there is a poll. Your feedback can shape future posts, cast a vote!

---

Can a Netflix series teach us something important not about the ancient past, but about how we create and evaluate knowledge today?

Human history is a complex adaptive system: countless interacting agents, environmental pressures and dramatic changes, cultural innovations and feedback loops. Reconstructing it is like reverse-engineering a living organism from a few scattered bones: inference under sparse and biased data. And we know how difficult it is:

“Model-free“ analysis of a complex system

“Model-free“ analysis of a complex system. Part II

Functionally, the challenge is also epistemic:

• In science, we distinguish between what is established (multiple independent lines of evidence converge) and what is conjectural (plausible but not yet tested, or tested and rejected).

• In public debate, those categories can blur, especially when a compelling narrative appeals to our fascination with lost knowledge.

I am not an archaeologist, of course. My interest here comes from a scientist’s standpoint: as someone used to testing hypotheses, trying to falsify them and dealing with uncertainty when working with empirical systems.

That’s why Graham Hancock’s Netflix show, Ancient Apocalypse, caught my attention: not just for its content (which I didn’t guess a priori, while I was also unaware about Hancock and his story), but also for how it frames disagreement with the archaeological mainstream as an attempt to “silence” him. That’s curious, isn’t it?

Overall, for people like me loving (hi)stories about ancient places and mysterious sites, that show provides a great opportunity to discover amazing places around the world.

Göbekli Tepe: a site older than Stonehenge. Source: History.com

---

The Hancock’s Hypothesis

The show rotates around the central claim — by Graham Hancock, journalist and author according to Wikipedia — that a single, advanced, global civilization existed until a catastrophic event — namely, the Younger Dryas — about 12,800 years ago. In the following we will refer to this take as the Hancock’s Hypothesis.

There was a lost advanced world spanning civilization that existed until the end of the last ice age. — G. Hancock

In his telling:

• This civilization possessed advanced knowledge in agriculture, mathematics and astronomy.

• It was destroyed by a cataclysm, possibly a comet impact.

• Survivors dispersed globally, teaching “primitive” — in Hancock’s framing, not mine — hunter-gatherers to build megalithic structures aligned with celestial events, as well as underground cities due to fear of future disasters.

Parts of this story rest on established science; while others are deeply contested. The point is that the show does not make the effort to present a balanced overview of the debate: on the one hand we have the scientists who are accused to hide “the truth” — providing scientific evidences that are never discussed — while, on the other hand, there is Hancock providing “the truth that someone does not want to tell us”.

Underground cities in Cappadocia, Turkey. Figure from cabinstanbul.com

In the following, let’s try to make sense of a few fundamental pieces of the story, trying to distinguish what is an hypothesis supported by scientific evidence from what is storytelling.

Younger Dryas

This is a well-documented abrupt cooling period from ~12.9 ka BP to ~11.7 ka BP, recorded in Greenland ice cores, speleothems and lake sediments. A recent study suggests the onset of YD in the North Atlantic around 12,870 ± 30 B.P., and termination that may have started first in Antarctica at ∼11,900 B.P. followed by the North Atlantic between ∼11,700 ± 40 and 11,610 ± 40 B.P. Therefore, the estimated dates are quite accurate.

[…] a north-to-south climate signal propagation via both atmospheric (decadal-time scale) and oceanic (centennial-time scale) processes. — Cheng et al, 2020

The cause of the YD is debated. Recent studies attributes this to meltwater flooding into the North Atlantic disrupting deep-water formation and weakening ocean circulation systems like the AMOC.

The drainage pathways of meltwater stored in glacial lakes located along the southern margin of the Laurentide Ice Sheet. Figure from PNAS.

An alternative view, known as the extraterrestrial‑impact hypothesis, has been proposed but remains controversial and largely unsupported due to poor replication and lack of compelling geological evidence:

A key element underpinning the controversial hypothesis of a widely destructive extraterrestrial impact at the onset of the Younger Dryas is the claim that 29 sites across four continents yield impact indicators all dated to 12,800 ± 150 years ago. This claim can be rejected: only three of those sites are dated to this window of time. At the remainder, the supposed impact markers are undated or significantly older or younger than 12,800 years ago. Either there were many more impacts than supposed, including one as recently as 5 centuries ago, or, far more likely, these are not extraterrestrial impact markers.

The above study comes one year after the one claiming evidences for the impact hypothesis:

We present detailed geochemical and morphological analyses of nearly 700 spherules from 18 sites in support of a major cosmic impact at the onset of the Younger Dryas episode (12.8 ka). The impact distributed ∼10 million tonnes of melted spherules over 50 million square kilometers on four continents. Origins of the spherules by volcanism, anthropogenesis, authigenesis, lightning, and meteoritic ablation are rejected on geochemical and morphological grounds. The spherules closely resemble known impact materials derived from surficial sediments melted at temperatures >2,200 °C. The spherules correlate with abundances of associated melt-glass, nanodiamonds, carbon spherules, aciniform carbon, charcoal, and iridium.

Figure from phys.org

This is an important fact to take into account, since part of the Hancock’s Hypothesis is based on the assumption of such an impact.

Sea-level rise

Since the Last Glacial Maximum (~26–20 ka B.P.), global seas have risen ~120–130 m, with rapid pulses early in the Holocene. Below I show two plots about this.

The first one places the Last Glacial Maximum in a much longer context, spanning the past 120,000 years of glacial–interglacial cycles. It shows how global mean sea level has oscillated with the waxing and waning of ice sheets, from highstands near today’s level (e.g., MIS 5a, ~80 ka) to lowstands over 120 m below present. During Marine Isotope Stage 3 (~50–30 ka), sea level hovered around −40 m before plunging toward the Last Glacial Maximum lowstand.

Long-term sea-level change. Global mean sea level over the past 120,000 years, showing the highstands of interglacial periods (near present sea level) and deep lowstands during glacial maxima. Marine Isotope Stage (MIS) 3 (~50–30 ka) remained ~40 m below present before the drop into the Last Glacial Maximum lowstand (>120 m below present). Figure from Nature.

The second plot zooms in on the last 30,000 years with high-resolution reconstructions from coral, sediment and glacial-isostatic modelling. It captures the sharp fall into the LGM lowstand (−125m to −130m around 26 ka B.P.) and the rapid, punctuated rise of ~120 m into the Holocene.

Deglaciation in detail. High-resolution reconstruction of global mean sea level for the past ~30,000 years. The Last Glacial Maximum (~26 ka BP) reached ~125–130 m below present, followed by a rapid, stepwise rise of ~120 m into the Holocene, dramatically reshaping coastlines worldwide. Figure from PNAS.

Together, the two figures reveal both the long-term rhythm of Earth’s ice-age cycles and the dramatic deglaciation that would have reshaped coastlines — and any coastal societies — within the span of human memory.

In the show, my attention was especially caught by the theory that before sea level rise there was a land corridor connecting connecting Sicily to Malta, allowing my ancestors to reach Malta to build megalithic sites like Ggantja (an amazing place!).

Ggantija Temple on Gozo. Source: Wikimedia.

Since I am Sicilian, I have tried to check by myself that statement, at least in terms of what I am able to do: look for data and plot them. A quick search pointed me to a dataset that maps land–water transitions relative to the modern coastline using the ICE‑6G_C ice-sheet reconstruction (VM5a) at 10‑arc‑minute resolution, providing gridded topography, land mask and ice mask data for key time slices between 0 and 26,000 years ago (PMIP‑4 archive). I do not know if this is the best dataset out there, but it did the job: below, I show the world maps I have obtained for land-water transitions in three distinct epochs — 25 ka B.P., 13 ka B.P. (the starting of the Younger Dryas) and 11.5 ka B.P. (around the termination of the Younge Dryas) — compared to the modern time.

The colors indicate how each location’s status (land or water) at thousands of years before present differs from today: orange marks land areas that have since been submerged (land → water), yellow marks areas that were under water but are now land (water → land), green marks land that remained land (land → land) and blue marks water that remained water (water → water).

Cool! Isn’t it? Especially if one realizes that this comes from publicly available data.

Here it comes the interesting part: a zoom around Malta:

Can you imagine my disappointment? It seems that this land corridor might have been there, but not around or during the Younger Dryas: at that time the sea level was already high enough to prevent any walk between Sicily and Malta. Therefore, the dates provided by advanced reconstructions do not match the dates provided by Hancock in the show. There are three explanations for this mismatch:

• The data is correct, but I was unable to use it correctly;

• The data is not correct;

• Hancock’s Hypothesis is wrong.

From a search through the scientific literature, it seems that the PMIP‑4 archive is well accepted.

White Sands footprints

Four human footprints found in alkali sand at White Sands National Park in New Mexico. The prints date from 23,000 to 21,000 years ago, long before humans were thought to have first entered North America. Figure and caption from berkeley.edu

Dated to ~23–21 ka using multiple independent methods, these show humans in North America earlier than previously thought, calling for a major revision to the “Clovis-first” model. This is amazing since, to the best of my knowledge, it was assumed (based on available data) that the first humans settling in the North America were migrating from Asia through the Bering straight, as shown below,

Map of early human migrations based on the Out of Africa theory; figures are in thousands of years ago (kya). Figure from Wikipedia.

thanks to emergence of a land bridge after the Last Glacial Maximum rise in eustatic sea level (see the animation below).

Figure from Wikipedia.

However, early presence is not necessarily evidence of advanced civilization: it still fits well with current models of late-Pleistocene human dispersal and does not require some lost world-spanning ancient civilization to be explained. I am reasonably sure that scientists supporting the “Clovis-first” model will accept the new evidence, just because it is scientific evidence.

Figure from Science.

---

Building a narrative vs. doing science

The show catches the attention of the public during every single moment. The narrative follows approximately this structure:

1. Framing. It starts with an appealing mystery: lost civilizations, ancient wisdom, myths, or a cataclysm.

2. Selection. Hancock visits specific sites that can be made to fit the story, while apparently ignoring others that don’t. But this is impossible to understand from the show alone: one must do some educated online search. I recognize that the average public would not even care to question what they listen to, let alone to perform their own searches to put what they listen to in the correct — or at least less biased — context.

3. Interpretation. Hancock offers subjective readings of artifacts without standardized dating, stratigraphic controls or peer-reviewed methodology. This is a fact: most of the information presented during the show is usually interpreted in an “alternative way” with respect to the “mainstream archeology”, as clearly indicated by the narrator. I feel there is always a great pushing over this duality.

4. Amplification. Hancock makes use of similarities in iconography or alignments as proof of contact or dating, without showing independent, replicable cultural transmission pathways. Other traditional methods adopted by scientists for the same purpose are not provided.

This distorted set of constellations is what our distant ancestors saw in the night skies of 20,000 B.C. Humans have always used the brightest stars to trace patterns in the sky, but those stars are generally our closest neighbors in the galaxy and those with the highest proper motions. (Image credit: SkySafari app). Figure and caption from space.com

To my eyes, this approach is the opposite of scientific method, which is more about to try to prove oneself wrong. In Bayesian terms:

Let’s assume that A denotes an hypothesis H while B denotes supporting data D.
If the prior P(H) for a global Ice-Age civilization is extremely low (because such a civilization would leave abundant, convergent evidence), then the likelihood P(D| H) must be extraordinarily high, based on strong, reproducible data, to change our minds. And finally, one compares all the theories with their evidence against each other, and usually consensus emerge on the one able to explain the most while using the minimum number of assumptions or free parameters. That is, according to the principle of the Occam’s razor. However, this standard isn’t met here.

I think that this is not just about archaeology. It’s about information ecosystems and how narratives propagate. Scientists must explain why evidence doesn’t fit a claim without belittling curiosity, without risking that “no evidence” is conflated with “cover-up”.

TV shows should make an effort to equip people with tools to evaluate reported information. From dating methods to understanding replication, people should be enabled to distinguish speculation from consensus, without the need for censorship. In fact, consuming misinformation and disinformation can have dangerous consequences (e.g., during a health emergency). Accordingly, documentaries should label conjecture clearly and show how experts test claims; I do not think that free speech means to build a show where only one side of a debate is presented. Such a situation might occur only when the true underlying interest is monetization, a problem that is common to all information ecosystems nowadays.

---

Take-home Message

Fascinating stories about our origins should inspire curiosity. But the most extraordinary claims need the most extraordinary evidence. When we shortcut the scientific method, we don’t just risk misunderstanding the past: we risk eroding trust in the very tools that let us know anything at all. Watching shows like “Ancient Apocalypse” might be very entertaining, as long as the intent is not to confuse the public about how science works.

---

A final note. It is unfortunate that more scientific documentaries are not as engaging, even if, honestly, I have enjoyed watching “Our Universe”:

and “Alien Worlds”:

---

→ Please, remind that if you find value in #ComplexityThoughts, you might consider helping it grow by subscribing, or by sharing it with friends, colleagues or on social media. See also this post to learn more about this space.