10 Fierce Debates Rocking Ancient History (2025)

For regular videos on ancient cultures and forgotten civilizations, please subscribe. The past 12 months have been a reminder that ancient history and archaeology aren't just tomes and trowels and trenches. It can be intellectual hand-to-hand combat. Yeah, today we're diving headfirst into that academic crossfire. These are the 10 hottest debates, in my opinion, of the past 12 months among academics and scholars. Arguments so fresh, the conference coffee cups are still warm. Number one, ancient Egyptian race debate. In July 2025, Adeline Moraes Jacobs and colleagues published the first full genome from ancient Egypt, a man from Nuairat, radiocarbon dated to 2855 to 2570 BCE. Their take, about 80% North African ancestry plus about 20% fertile crescent ancestry, consistent with gene flow between Egypt and Mesopotamia during the Old Kingdom. Co-author Pontus Skoglund called it a rare preservation win that finally lets genetics weigh in on early Egypt's population history. The evidence base now combines that lone genome from Nuairat with earlier mummy samples from Abu Sir el-Melek published in 2017, which tilted toward eastern Mediterranean affinities. Everyone asked, how typical was that genetic cocktail? Harold Ringbauer of the Max Planck Institute asked the key question that this limited evidence can't settle. Did Levantine-like ancestry arrive with farming much earlier, or does it reflect movements specific to Egypt's state-building era? More pointedly, does this settle the argument over whether ancient Egyptians were black, white, or something else? The controversy flares because each data set is narrow and can be weaponized. Afrocentric historians contend that colonial scholarship erased Egypt's African identity. Critics of the 2017 study note its northern bias, while skeptics of the 2025 paper call one genome an anecdote, not a population. Traditional Egyptologists counter that race is an anachronism here, urging focus on culture, not color. How might this be resolved? Scale and context. A geographically balanced ADNA project from Lower Egypt to Nubia spanning dynastic millennia could map real population shifts, and future debates may finally pivot from racial labels and who should get all the credit to the far richer story of Egyptian cosmopolitanism. Who should own the world's antiquities? The countries where they were created, or the museums that now display them? That question boiled over again this summer when France introduced a bill that finally opens the door to returning colonial-era artifacts to their original countries, at the same time insisting the national collections remain inalienable. Weeks earlier, the Netherlands handed another trove of Benin bronzes back to Nigeria, adding weight to calls for permanent restitution. And in London, politicians on the right warned the British Museum that any deal to reunite the Parthenon marbles with Athens would undermine British heritage, drawing scorn from classical archaeologists who called the threat culture war theater. Pro-repatriation scholars argue that many marquee pieces left Africa, Asia, and the Mediterranean through coercion or outright looting. Returning them restores cultural memory and can jumpstart local tourism economies. They point to successful recent handovers, like Germany's 2023 Benin shipments, as proof that major collections won't collapse when they let go. Universal Museum defenders counter that encyclopedic displays tell a global story in one place and ensure conservation standards some source nations still lack. They also warn that once a return precedent is set, political campaigns could empty Western galleries piece by piece. A workable piece may hinge on hybrid models, renewable loans that rotate objects home for long exhibitions, joint conservation labs funded by both parties, and 3D scanning that puts full-resolution replicas online for anyone, anywhere. Number three, where did the Proto-Indo-European language begin? In February 2025, a major nature study argued that the ancestor of all Indo-European languages most likely formed among Caucasus Lower Volga CLV groups about 6,500 years ago. It labels this ancestor language Proto-Indo-Anatolian. The Anatolian languages then split off from it and Proto-Indo-European, PIE, which did not include the Anatolian languages, crystallized later among the Serednisty communities north of the Black Sea. The well-known Yamnaya then spread PIE widely, but they weren't the first speakers. The case rests on hundreds of newly sequenced ancient genomes, which show strong links between Yamnaya and those earlier CLV populations, and it offers a neat solution to an old puzzle. Hittite and the other Anatolian languages never showed the classic step genetic signal, and there are elements of their language group that differ from all other Indo-European languages. Immediate pushback came from linguist Paul Hegarty, lead of a groundbreaking 2023 phylogenetic paper, who promoted a different theory. He argues the new model is a retreat with some smoke and mirrors in how it redefines Indo-European. He points out that the new study contains no linguistic analysis, calls the new DNA silent on language, and argues that early Indo-European farmers emerged south of the Caucasus in the northern part of the Fertile Crescent. What is the path to resolution? Well, secure genomes from 7,000 to 4,000 BCE Anatolia would probably fill the glaring gap between Neolithic farmers and documented Hittites, and allow scholars to integrate language data with ancient DNA maps. Until then, every new dataset seems to redraw the language family tree in some way, and keeps this 200-year debate galloping. 4. AI vs. Human Translation of Ancient Texts In April 2025, a Leiden team ran a bold test. Feed dozens of fragmentary Greek and Latin sentences to both a large language model and a panel of seasoned classicists, then compare their reconstructions to the published editions. The shock came when the AI's guesses aligned with the gold standard more often than the scholars agreed with one another. Three months later, DeepMind's Aeneas system raised the stakes, filling gaps in 90% of trial inscriptions and even proposing likely dates and find spots. That performance sits on a mountain of new evidence, millions of digitized papyri and ostraca, high-resolution scans, and transformer models trained on everything from Homer to market receipts. Yet the same engines sometimes invent smooth-sounding words, erase dialect quirks, or misanachronisms that jump off the page to a human eye. Enthusiasts say these tools free researchers to tackle bigger cultural questions. Skeptics come back and say that over-reliance could de-skill young scholars and lock algorithmic bias into future editions of translations. So what should be done? A realistic way forward might be to let the model propose readings, let humans vet every line, and then publish a transparent change log so anyone can trace how a smudge turned up as a syllable. But totally automatic translation might not be in the cards. Number five, were there true cities in the ancient Amazon? Back in January 2024, a science cover story raised some archaeologists' eyebrows. LiDAR sweeps over Ecuador's Upano Valley uncovered gridded avenues, six-meter-high platform mounds, and canal-fed garden plots dating to 500 BCE to 300 CE. Lead author Stefan Rostain dubbed them garden cities, arguing that tens of thousands of people once lived in a low-density urban landscape rivaling classical Mesoamerican towns. Since the 2024 science paper, coverage this year has amplified the Upano findings. Supporters say the scale is the story. Statistical modeling now projects 10,000 to 24,000 pre-Columbian earthworks across the basin. They have been amplifying LiDAR results that point to vast engineered landscapes, thousands of earthworks, gridded roads, garden plots, even orchards. But there has been pushback. Ancient Americanist Fernanda Ugalde criticized the lost city hype that followed the science paper, noting that Upano's platforms, canals, and roads have been known for almost 50 years. In other words, the novelty is overstated. Alejandra Sanchez-Poyo, an archaeologist who co-authored the 2023 LiDAR overview of the Upano Valley in Ecuador's National Institute of Cultural Heritage journal, notes that her team had already published remote sensing results before the science paper. We had the opportunity to work all of that data, she said, but since we weren't publishing in science it didn't create the same stir. Both of them caution that claims about cities should rest on excavation as much as on eye-catching scans. How can they settle this? Well, treat LiDAR as the map, not the verdict. As archaeologist Mark Robinson put it, LiDAR gives the magic landscape, but teams still need radiocarbon, GPR, and trench-by-trench ground-truthing to answer population impermanence. Rustin's group says that follow-up is underway. Number six, Celtic identity, ancient reality, or modern invention. In a new book published in 2025 titled The Celts of Modern History, author Ian Stewart argues that modern Celts are recent inventions grafted onto historical abstractions, a collage of disparate symbols from the pre-Roman past stuck together with imagination, enthusiasm, and academic linguistics. He refers to the idea of one Celtic people as a Victorian marketing triumph rather than an Iron Age reality. That claim hit a nerve. Historian Peter Beresford Ellis, a retorted in the Irish Times that Stewart overlooks essential studies and downplays hard archaeological links, Latin art styles, torc burials, chariot graves, and a family of clearly related Celtic languages stretching from Iberia to Bohemia. Stewart's supporters fire back. Classical authors use the term inconsistently, and a paper that just came out a few months ago has argued that Celtic languages originally spread without the In other words, delinking the genetics with the linguistics. What could settle this? I'm not an expert in DNA, but perhaps a continent-wide project sequencing Iron Age cemeteries matched with strontium isotope maps of migration would test biological connectedness. One day, hopefully, we will finally learn whether Celtic is an appropriate badge of identity or a brilliant piece of Victorian branding that stuck. 7. How early did writing appear in South India? The flashpoint is Kiladi in Tamil Nadu, where archaeologist K. Amarnath Ramakrishna has excavated. He argues that the site's earliest phase reaches the 8th to 6th centuries BCE, with Tamil Brahmi graffiti turning up in secure layers. If correct, this would push back the first appearance of writing in southern India much earlier. His 982-page report leans on AMS radiocarbon dates from Kiladi and neighboring sites. A 2024 roundup of 73 samples even flagged an early date of 685 BCE at the site of Shiva Galai. But in May 2025, India's Archaeological Survey, ASI, sent Ramakrishna's report back for revision. What? Why? Well, the covering letter says certain parts lack scientific rigor and need, quote, more authentic, unquote, documentation, and that his proposed dating of the site's layers, quote, requires concrete justification. It even suggests the earliest period should be at the maximum pre-300 BCE. A few weeks later, culture minister Gajendra Singh Shekhawat told reporters Ramakrishna's report is not technically well supported and needs further validation. Ramakrishna refused to amend his findings, telling the press it would be criminal and immoral to change a completed excavation report. He insists the chronology rests on stratigraphy and lab dates, not opinion. Meanwhile, the ASI also asked retired superintending archaeologist P.S. Sriraman, who ran a brief 2017 season that reported no significant findings, to file an official report of his own, a move that critics say muddies the waters. So the issue here is whether early Tamil Brahmi marks at Keeladi sit in securely dated contexts that truly come from a time before Emperor Ashoka in the 3rd century BCE. How will this be resolved? I'm not sure what the correct procedure is about the submission of reports and their robustness, but as regards the ancient writing, I don't think the issue will be totally resolved until more can be found to corroborate it. The problem is, what if there never is any further evidence found? Then this report may be a sticky wicket for some time to come. Number eight is classics too Eurocentric? In a new book called What is Ancient History, historian Walter Scheidel argued that teaching the ancient world should go truly global, and even wondered just how much language expertise do ancient historians actually need, suggesting that minimal engagement with Greek and Latin is certainly good enough. He proposes folding Greek and Roman studies into a broader survey of world antiquity, freeing curricula from what he calls a Euro-American foundation myth. To back up his position, he cites enrollment figures showing classics departments shrinking in the US and the UK, syllabi that still center Homer and Caesar while skimming Persia, India, and Mesoamerica, and the rise of high-quality translations plus AI reading tools that, in his view, make deep language mastery less urgent. But this got some classicist dander up. Critics like Oxford classicist Edmund Stuart reply that abolishing intensive language training breathtakingly guts the discipline, because entire arguments in Thucydides or Tacitus can pivot on a single particle. Others warn that global ancient history taught only in English risks swapping Eurocentrism for Anglocentrism, while still leaving students unable to read cuneiform hieroglyphs or classical Chinese. Supporters counter that insisting on six years of Greek and Latin has already priced most students out, and that comparative modules would show how Athens, Xi'an, and Teotihuacan faced similar state-building challenges. Where could this all be headed? Universities are piloting hybrid tracks. Two years of language boot camp, followed by thematic cross-civilization seminars, plus shared open access data sets, so that a student fluent in, say, Akkadian can still contribute to something in Roman studies. If such models raise enrollment and preserve linguistic rigor, the Eurocentrism debate may finally cool. Number nine, Petra's treasury tomb. Brand new find or reboot? In October 2024, a joint team working with Jordan's Department of Antiquities, ACOR, and the University of St. Andrews announced the discovery of a rock-cut tomb at Petra in Jordan, beneath al-Khazneh, the treasury, the single most iconic facade in Petra, and said it held the remains of 12 people plus grave goods. St. Andrews' release called it historic, stressing that complete Nabataean burials at Petra are rare, because so many tombs were robbed or reused over centuries. Maybe you remember it appeared in an episode of Expedition Unknown with Joss Gates. Gates said, this is perhaps the most significant tomb ever found at Petra, and a discovery of historic proportions. Specialists bristled. Dr. Lucy Wadeson, a Nabataean funerary expert, called the coverage a little hyped up, noting comparable tombs at the site. Bioarchaeologist Dr. Megan Perry pointed out that there's nothing unusual about finding tombs with burials. More important is the question of what the dates and context actually prove. Then came a revelation. Dr. Suleiman Farajat, a Jordanian archaeologist who co-led the Department of Antiquities excavations in front of the treasury in 2003 to 2005, publicly reminded everyone that the very same tomb was already excavated back then. The new work he revealed was a revisit, not a first discovery. ACOR later clarified that 2024 season reopened the context and collected samples for analysis. Another archaeologist who works at Petra, Dr. Leanne Bettle, revealed that everyone working on this knew that the site had already been discovered, and in fact they were using the original excavation report as a guide, and that when Joss Gates and archaeologist Pierce Paul Creasman were on camera supposedly uncovering the tomb, that was all staged. She said all this was damaging to the relationship the American archaeologists have with the Jordanian ones. Where do things stand? ACOR says samples, radiocarbon, OSL, DNA, ceramics, were collected, and analysis is coming. And number 10, the Dead Sea Scrolls, redated? In June 2025, the PLOS One team, led by Marvin Popovich, unveiled Enoch, an AI that blends radiocarbon results with handwriting metrics to redate 30 Dead Sea Scroll fragments, pushing several, like pieces of First Enoch and the Community Rule, a century earlier than previously thought. Supporters called it the next step in biblical chronology. The data behind the claim are impressive. Fresh, high-precision C14 runs from Zurich and Oxford, multispectral images that let the program measure every pen stroke angle, and a Bayesian routine that fuses chemistry with paleography into a single probability curve. But pushback was immediate. Epigrapher Christopher Ralston told Bible History Daily the new dates for 4Q114, one of the fragments, stretch credulity, because internal linguistic clues point later. Scroll veteran Peter Flint warned in Live Science that a 30-scroll training set is promising but premature. By contrast, Jonathan Bendol likes the algorithm for spotlighting scribal quirks scholars miss. And archaeologist Yonatan Adler says critics rely too heavily on what we've always thought rather than measured data. What could settle things? Well, you could run Enoch and a panel of human editors in a blind test on some new fragments, then publish every intermediate step, models, code, error bars, in an open repository. So there you have it, my picks for the 10 hottest academic debates about ancient history in the last year. Are there any scholarly controversies that you think I left out? If so, tell me in the comments section below. You might like my little e-booklet Why Ancient History Matters. It's designed to persuade people that the subject is important, even in the modern world. You might also wish to use it to help spread the word, so feel free to share it with someone you know. It's free for anyone who wants it. I've left a link in the description box below the video for you to grab a copy. Catch you later.