You Can’t Change the Past, But….

The past has already happened, but our understanding of it is anything but fixed. History is reconstructed from whatever happens to survive: a broken pot, a few marks scratched into bone, fragments of a clay tablet, timbers preserved beneath wet soil, or sometimes little more than a stain where a building once stood. Every discovery gives us another piece of that enormous puzzle, and new technologies increasingly allow researchers to return to evidence we already possessed and find things earlier generations could not see.

Three recent discoveries illustrate just how dramatically that can change our view. In Germany, researchers studying marks made nearly 40,000 years ago have found evidence that some of Europe’s earliest modern humans may have been using conventional signs to store information. Scholars working with artificial intelligence have reconstructed much of a long-lost Babylonian hymn from cuneiform fragments scattered across collections. And another German excavation has produced a remarkably preserved Celtic burial chamber whose oak timbers can be dated to the actual year it was built.

None of these discoveries rewrites the past by changing what happened. They change the picture we are able to reconstruct from what happened, sometimes turning something vague or almost invisible into a surprisingly detailed glimpse of real human lives.

Was Protowriting Developing 40,000 Years Ago?

People were making deliberate marks on objects long before anyone invented an alphabet. Lines, dots, crosses, notches, and other geometric signs appear on artifacts from the Upper Paleolithic, but determining what those marks meant has always been difficult. Decoration is one possibility. Counting, ownership, identity, memory aids, or symbolic communication are others.

In 2026, linguist Christian Bentz and archaeologist Ewa Dutkiewicz took a much broader look at the problem. Rather than concentrating on one spectacular artifact, they used computational methods to study thousands of geometric marks found on more than 200 objects associated with the Aurignacian culture of Central Europe. Many came from caves in southwestern Germany and date to roughly 43,000–34,000 years ago, a period that also produced some of the world’s earliest known figurative sculpture and musical instruments.¹

Patterns began to emerge when the objects were compared as a group. Certain signs appeared more often on particular kinds of artifacts, while sequences of marks showed enough regularity to suggest that people were applying shared conventions rather than simply adding random decoration. On some measures, the information carried by those sequences even showed similarities to much later proto-cuneiform notation.¹

That makes protowriting a useful way to think about what may have been happening, as long as we remember how much remains unknown. Nobody can currently read these marks, and there is no evidence that they encoded spoken sentences the way later writing systems did. What the research suggests is something more fundamental: people may already have been developing agreed-upon visual signs capable of carrying information outside the human mind.

One particularly intriguing artifact is the small ivory figure sometimes called the Adorant, discovered at Geißenklösterle Cave. It is roughly 40,000 years old and carries several arrangements of notches and dots. Similar geometric markings occur on other artifacts from the region, and studying them collectively gives researchers a chance to ask whether these were individual acts of decoration or pieces of a larger system of visual communication.¹

If that interpretation continues to hold up, the significance goes well beyond simply moving a date farther backward. Writing did not suddenly appear when someone in Mesopotamia decided to invent cuneiform. The cognitive and cultural habits that made writing possible, such as using standardized marks, associating symbols with information, and storing knowledge outside the brain, may have roots extending tens of thousands of years deeper into human history.

The objects were sitting there all along. What changed was our ability to compare them at a scale that earlier researchers could not.

AI Reassembles a Lost Voice from Babylon

By the time we reach ancient Babylon, humans had taken external information storage to another level entirely. Mesopotamian scribes produced enormous quantities of writing on clay tablets, recording everything from taxes and legal contracts to mathematics, medicine, mythology, prayers, and poetry.

Clay has one wonderful archaeological advantage: once hardened, it can survive for thousands of years. Unfortunately, tablets also break.

That leaves modern scholars with a monumental jigsaw puzzle. Pieces of the same composition may have been excavated at different times, stored in different museums, catalogued separately, and studied by different specialists. Recognizing that two small fragments belong to copies of the same text can require years of familiarity with cuneiform writing and Babylonian literature.

Artificial intelligence is beginning to make that search considerably faster.

Assyriologists Anmar A. Fadhil of the University of Baghdad and Enrique Jiménez of Ludwig Maximilian University of Munich recently reconstructed much of a previously unknown hymn celebrating Babylon, its patron god Marduk, his great temple Esagil, and the people of the city. The original composition was probably around 250 lines long, with roughly two-thirds of it now recovered completely or in part from 20 surviving manuscript witnesses.²

The reconstruction benefited from the Electronic Babylonian Literature project, which has been digitizing enormous numbers of cuneiform texts and developing computer-assisted methods for comparing them. AI-supported searches helped identify fragments and copies that appeared to belong to the same composition, allowing scholars to recognize relationships that could otherwise have taken years of manual comparison.³

This is not a case of an AI system independently translating an ancient hymn. Human Assyriologists still had to determine how the fragments related, resolve damaged passages, understand the grammar, recognize literary allusions, compare variant manuscripts, and interpret the resulting composition. The technology helped them find promising connections within a body of material too large for any individual scholar to hold completely in memory.

What they recovered is wonderfully human.

The hymn praises Babylon’s monumental buildings and religious importance, but it also describes the Euphrates watering the surrounding countryside and bringing life to the fields. It discusses women serving as priestesses and praises the generosity of Babylon’s people, including their treatment of foreigners.² ³ Instead of giving us only kings, wars, and monumental architecture, the text adds another layer to how Babylonians imagined their own city and community.

The number of surviving copies reveals something else about the past. This was apparently not an obscure work that disappeared immediately after it was written. Evidence from manuscripts and school exercises suggests that the hymn circulated within scribal education and was once familiar enough to be copied and learned.²

At some point, however, that familiarity disappeared. The clay survived, but the literary connections between the surviving pieces did not.

For roughly two thousand years, parts of the hymn existed in collections without anyone realizing that they belonged together. Digital tools did not create a new Babylonian text; they helped modern scholars recognize a text that ancient scribes already knew.

A Celtic Burial Chamber Gets an Actual Year

The third discovery takes us back to Germany, but this time only about 2,600 years into the past.

Near Riedlingen in Baden-Württemberg, archaeologists began investigating a large Early Celtic burial mound in 2023. At its center, they uncovered something extraordinarily rare: a wooden burial chamber whose oak floor, walls, and roof had survived beneath the ground. Under ordinary conditions, much of that timber would have disappeared centuries ago. Waterlogged, oxygen-poor soil had instead preserved enough of the wood for archaeologists to study the chamber’s construction in remarkable detail.⁴

The mound itself was monumental, measuring about 65 meters across and perhaps originally rising six meters above the surrounding landscape. Inside, however, the burial chamber was surprisingly intimate—roughly three by four meters and only around a meter high. Archaeologists associate it with the elite society surrounding nearby Heuneburg, one of the most important Early Celtic centers in Central Europe.⁴

The preserved oak gave researchers something even more useful than an intact structure: tree rings.

Dendrochronology works by comparing the pattern of wide and narrow annual growth rings in wood against established regional sequences. Because climate affects tree growth from year to year, those patterns can function almost like barcodes. When enough of the outer portion of a timber survives, researchers can sometimes identify not merely the century or decade when a tree grew, but the exact year in which it was felled.

At Riedlingen, the match was exceptionally precise. The oak used in the chamber had been cut during the winter of 585/584 BCE and again in the summer of 584 BCE, allowing the Baden-Württemberg State Office for Cultural Heritage to date construction of the burial chamber to 584 BCE.⁴ ⁵

That level of precision changes the way an archaeological site feels. “About 2,600 years ago” belongs to the hazy distance of prehistory. “584 BCE” suddenly sounds like a building project undertaken by actual people. Trees were selected and felled, heavy timbers were moved, carpenters shaped the wood, and a group of people organized enough labor to build a chamber and bury it beneath an enormous mound.

The tree rings don’t tell us their names, but they pull their actions into much sharper focus.

Even the Grave Robbers Left a Story Behind

The Riedlingen burial has another layer that makes the site particularly fascinating. Ancient grave robbers eventually tunneled into the mound and removed many of the objects they considered valuable, apparently concentrating on metal. What they left behind turned out to be enormously valuable to archaeologists.

The waterlogged chamber preserved wooden objects, fragments associated with furniture or a wagon, textiles, fur, woven baskets, and other organic materials that would normally have decayed. Human remains found scattered in the robbery shaft belong to a young man approximately 17 to 19 years old, apparently the person for whom the chamber had originally been constructed. Archaeologists also recovered a strip of birch bark carrying a rare engraved image of a horse.⁴

The robbers even left evidence of their own activities. A birchwood torch and baskets found within the disturbed area appear to have been associated with the looting itself. Once the tunnel filled and returned to wet, low-oxygen conditions, many of the objects they had discarded survived.

That means archaeologists are reconstructing two different human events at the same location. The first occurred in 584 BCE, when a community built an elaborate burial chamber for a young man. The second came generations later, when another group dug into the mound, crawled through the darkness, dismantled parts of the grave, and removed whatever they considered worth taking.

The same site preserves both the people who created a monument and the people who later violated it. Archaeology becomes much richer when we stop thinking only about the object and start asking what sequences of human behavior produced everything around it.

The Past Keeps Giving Us More to Work With

These three discoveries are separated by tens of thousands of years and have almost nothing obvious in common. One involves geometric marks made by Ice Age people, another concerns fragments of Babylonian literature, and the third begins with oak timbers preserved beneath a Celtic burial mound.

What connects them is the changing relationship between evidence and the tools we use to understand it.

Computational analysis allowed researchers to compare Paleolithic signs across hundreds of objects and look for patterns that would be difficult to recognize one artifact at a time. Artificial intelligence helped Assyriologists search an enormous digital corpus for fragments that might belong to the same Babylonian composition. Dendrochronology turned preserved wood into a calendar precise enough to tell us that trees were being cut for a burial chamber during different seasons of 584 BCE.

In each case, the evidence existed before the breakthrough. The marks were already carved, the clay fragments were already sitting in collections, and the trees had been lying beneath the ground for more than two millennia. What changed was the quality of the questions researchers could ask and the tools available to answer them.

That is one of the reasons history and archaeology remain so exciting. Learning about the past is not simply a matter of memorizing a finished body of facts. Our reconstruction is always provisional because the evidence is incomplete, and because tomorrow’s researchers may have ways of seeing information that today’s researchers cannot yet extract.

Some future discoveries will come from spectacular excavations. Others may come from museum drawers, old photographs, forgotten field notes, DNA, chemical residues, satellite imagery, computer models, or technologies that have not yet been invented. Occasionally, the breakthrough will come because someone notices that a familiar object deserves a completely different question.

We cannot change the past, but we can continue getting better at understanding the people who lived there. Every new piece gives us a little more of a world that once seemed permanently out of reach.

References

¹ Bentz, Christian, and Ewa Dutkiewicz. “Humans 40,000 y Ago Developed a System of Conventional Signs.” Proceedings of the National Academy of Sciences 123, no. 9 (2026): e2520385123. DOI: 10.1073/pnas.2520385123.

² Fadhil, Anmar A., and Enrique Jiménez. “Literary Texts from the Sippar Library V: A Hymn in Praise of Babylon and the Babylonians.” Iraq 86 (2024): 21–78. Published online March 17, 2025. DOI: 10.1017/irq.2024.23.

³ Ludwig Maximilian University of Munich. “LMU Professor Enrique Jiménez Discovered Hymn to Babylon.” July 1, 2025. Electronic Babylonian Literature project and University of Baghdad collaboration.

⁴ Landesamt für Denkmalpflege Baden-Württemberg. “Neue Erkenntnisse zur keltischen Grabkammer in Riedlingen.” March 12, 2026.

⁵ Landesamt für Denkmalpflege Baden-Württemberg. “Die Dendrologie von Riedlingen-Galgenberg: Eisenzeitliche Grabkammer und deren Beraubung aus der Sicht der Bäume, des Holzes und des holzverwendenden Menschen.” February 26, 2026.

Keep Exploring with Forgotten Skillz

Interested in hands-on learning through nature, history, traditional skills, craft, stewardship, and human ingenuity? Follow Forgotten Skillz, explore our upcoming workshops and outings, and subscribe to the newsletter to keep the conversation going.

Michael A. Evans is the founder and lead educator of Forgotten Skillz and CEO of Synchronicity Coaching Inc., a multidisciplinary organization creating educational programs and experiential learning opportunities across Long Island and beyond.

Through Forgotten Skillz, Michael develops hands-on experiences that connect nature literacy, cultural history, traditional crafts, ancestral technologies, environmental stewardship, and practical outdoor skills. His programs invite participants to observe, question, experiment, make, and better understand how people across cultures and throughout history have interacted with the world around them.

Beyond Forgotten Skillz, Michael contributes to youth education as a creative collaborator on the Little Laurie Science Stories book series and the Ninja Née Science Education Program, helping create experiences that foster curiosity, creativity, scientific thinking, and a lasting enthusiasm for learning.