Deciphering Secrets of Ancient Civilizations, Noah's Ark, and Flood Myths | Lex Fridman Podcast #487
Summary
Finkel’s central thesis is that civilization’s decisive information technology was not pictography but the agreed sign that could replay sound. The oldest surviving evidence is around 3500 BC, not necessarily the invention date; phonetic signs liberated writing to capture grammar, literature, and unfamiliar languages. Cuneiform then operated for more than three millennia as a flexible protocol that looked “complex, bewildering-looking, off-putting” but could encode almost anything spoken.
Cuneiform’s durability came from early standardization, backward compatibility, and institutional control. Third-millennium scholars built systematic sign lists by material, color, country, deity, and other categories, creating a “spinal cord” that later readers could retrieve and teach. Scribal elites reinforced the system: literacy conferred power, and “nobody would ever concede any interest in the idea of literacy for all.”
The surviving record is a radically biased sample, making confident reconstructions of ancient economies and institutions hazardous. Thousands of Ur III accounting tablets from roughly 2100–2000 BC may derive from only two storage rooms, while Finkel argues that Ashurbanipal’s supposedly destroyed library may largely consist of leftovers after conquerors removed its tablets. His warning is blunt: assuming “what we have is what there was” is a foundational analytical error.
Finkel controversially thinks writing may reach back to Göbekli Tepe around 9000 BC. His evidence is a round green stone with an arched back, flat underside, and carved signs that he interprets as an official stamp, combined with the organizational demands of monumental construction. Calling it a “raindrop from which I infer writing,” he argues that earlier records may have existed on leaves or other perishable media—an inference Lex repeatedly flags as highly controversial.
The 1700 BC Ark Tablet makes the Mesopotamian ancestry of the Noah narrative difficult to dismiss while Finkel rejects a global Younger Dryas flood. Its roughly 60 lines prescribe a giant round coracle for Atra-Hasis, predating the biblical account by at least a millennium; the Gilgamesh parallel also shares the sequence of three released birds. Finkel assigns “negligible” probability to a global cataclysm, favoring a local disaster transformed into an “irresistible” literary countdown.
The Royal Game of Ur shows durable appeal arising from a calibrated mix of luck, strategy, portability, and short playtime. A 20-square game found at Ur around 2600 BC spread across much of the known world and lasted nearly 3,000 years without written rules. Finkel reconstructed playable rules from a second-century BC tablet, and the revived game is now played worldwide, including in Iraqi cafés; its chance-strategy mix still provokes the comic marital friction seen at his Getty demonstration.
For Finkel, museums are intergenerational memory infrastructure, not merely exhibition spaces. The British Museum’s roughly 130,000 cuneiform tablets—and the potentially millions still underground—are “stockpiling for future examination,” because objects acquire new meanings as analytical distance grows. His broader warning is that electronic addiction reduces human vitality and that diminished vocabulary weakens precise expression: language is a precision tool, and reading is how people obtain the words that improve the quality of thought.
Deep dive
1. Writing became transformative when pictures learned to carry sound
Finkel places the oldest surviving evidence around 3500 BC in Mesopotamia, where clay was already an everyday building material. His hedge is essential: “We have no idea when anything began”; the evidence identifies the earliest known stage, not the moment of invention.
Initial signs could convey simple pictographic messages—foot, leg, barley—and numerals. The decisive leap came when a picture could also represent the sound of its word, giving scribes reusable phonetic components from which they could record language rather than merely point at objects.
Once signs carried sound, writing could preserve grammar, proverbs, literature, contracts, medicine, magic, military campaigns, and almost every other textual genre. Finkel calls this “a pretty gigantic step”: a trained scribe could even transcribe an unfamiliar foreign language by ear without understanding it.
Long before the alphabet was imagined, cuneiform was already “very beautiful, very flexible” and capable of writing Sumerian, Babylonian, and other Bronze Age languages. Despite its apparent complexity, it lasted well over three millennia—possibly nearer four—before the simpler alphabet displaced it.
2. Standards and gatekeepers kept cuneiform legible for millennia
As signs proliferated, an early visionary apparently recognized two scaling risks: uncontrolled growth would fragment the system, and nobody could remember the inventory without a retrievable record. Early in the third millennium, scribes therefore organized signs and words into lists covering wood, reeds, colors, countries, gods, and other categories.
Finkel treats this invention of lexicography as inseparable from writing’s success. The early “visionaries” standardized, preserved, and taught the system so effectively that scholars living around Alexander’s era could still recognize tablets written in much older forms.
Complexity also protected an elite franchise. Temple clerks, royal secretaries, archivists, and educated professionals controlled knowledge unavailable to most people; Finkel imagines universal-literacy advocacy as “anathema” to a scribal class with no incentive to surrender its position.
Education sorted pupils by ability. Functional writers handled marriages, contracts, letters, and accounts; stronger students entered medicine, law, priesthood, royal service, architecture, army logistics, astronomy, theoretical grammar, or commentary. At the top were scholars analyzing the structures and relationship between Sumerian and Akkadian.
3. Finkel thinks pictography may predate Mesopotamia by millennia
The conventional sequence runs from pictures to phonetic signs, but Finkel finds it logically awkward. If people shared a language, why begin with the least flexible method instead of reusable sound units such as the frequent Sumerian prefixes mu, ba, and e?
His alternative is that pictographs developed much earlier for communication between traders who did not share a language. Three drawn sheep could cross a linguistic boundary just as a crossed-out cigarette does today; Sumerians later repurposed this long-standing international signage to write their own spoken language.
At Göbekli Tepe, dated here to about 9000 BC, Finkel points to a round green stone with an arched back, flat underside, and carved signs. He interprets it as a stamp used to ratify some object, arrangement, or official act—and therefore as evidence of writing.
Lex presses the controversy, and Finkel concedes that many Assyriologists “would leave the room.” His Sherlock Holmes formulation is that one can infer Niagara Falls from a raindrop; the seal is his raindrop, while any fuller record may have been written on flat leaves or other perishable material.
4. Ancient archives preserve accidents, leftovers, and administrative noise
Clay’s unexpected advantage is survival: hundreds of thousands of tablets now sit in collections, while Finkel suspects millions remain protected underground. Yet abundance can create a false impression of completeness; Assyriologists can read much of the material while still misunderstanding what proportion of the original record it represents.
Ur III archives illustrate the sampling problem. Thousands of matchbox-sized account tablets from roughly 2100–2000 BC record tithes and offerings through what Finkel calls the ancestors of the Inland Revenue, but the worldwide corpus may have originated in only “two rooms.”
Ashurbanipal deliberately assembled knowledge at Nineveh into a library, “a giant brain” comparable to Alexandria. Nineteenth-century excavators found burned, broken tablets scattered across floors and corridors, and later scholars spent generations joining fragments of Gilgamesh and other works.
Finkel reverses the standard destruction story: Babylonian and Elamite conquerors, sharing the language and script, would have taken the library’s tablets home rather than annihilating them. What survives may be duplicates, broken pieces, and dropped material—not the library itself. He labels the view controversial but insists it is “common sense.”
5. A trilingual inscription supplied the chisel that opened cuneiform
Cuneiform writes syllables rather than isolated consonants. Finkel’s English illustration breaks “museum” into mu, se, and um; for a consonant, vowel combinations could produce forms such as ab, ba, ib, bi, ub, bu, eb, and be.
The difficulty compounds because Sumerian signs retained multiple readings, perhaps inherited from tonal distinctions. Students learn ba one, ba two, ba three, and other values, then infer the right reading from context—without spaces marking where one word ends and another begins.
The Behistun inscription gave decipherers a Rosetta Stone-like bridge: Darius’s career appeared in Old Persian, Babylonian, and Elamite. Repeated royal formulas—“I am Darius the great king,” followed by titles and ancestry—aligned the three columns and exposed corresponding words and names.
Recognizing Babylonian as Semitic unlocked living linguistic relatives. A word such as Akkadian nāru, “river,” could be compared with Arabic and Aramaic nahr, sending scholars into Hebrew and Arabic dictionaries for roots fitting the context. Without that trilingual inscription, Finkel is unsure how quickly the script could have been cracked.
6. Multivalence made decipherment harder—and shifted credit toward Hincks
Henry Rawlinson physically copied the mountain inscription, climbing high above the plain with a young assistant and producing squeezes used by decipherers. Finkel nonetheless demotes the conventional “Father of Assyriology” to its “Stepfather,” arguing that Rawlinson’s initial book was largely wrong.
His preferred hero is Edward Hincks, a Northern Irish clergyman with five daughters whom Finkel calls “possibly a card-carrying genius.” After reaching a dead end on Egyptian hieroglyphs, Hincks examined cuneiform and recognized that one sign could carry several sounds and meanings.
Reading therefore becomes constrained hypothesis testing: start with the first sign, cycle through its possible values, compare the next sign, and search for a plausible syllabic word. The variables are numerous but controlled; once a student masters the corpus, sign construction, and recurring readings, “off you go.”
Finkel entered the field by accident. He planned to become an Egyptologist, but his lecturer died after one class; the department redirected him to the terrifying, “sarcastic, cruel” Professor Lambert. Within one or two lessons, Finkel knew cuneiform would be his life’s work.
7. Sumerian is a solitary survivor from a vanished linguistic world
Babylonian is an extinct Semitic language related to Hebrew, Aramaic, Arabic, Ethiopic, and Syriac, which gave decipherers comparative anchors. Sumerian is agglutinative—particles attach before and after a root—but has no demonstrated relationship to another known language.
Finkel finds that isolation “bewitching.” Since languages normally belong to families, Sumerian must once have had relatives; writing captured one representative just in time, while related languages across Asia or elsewhere disappeared beyond recovery, unless “something amazing happens.”
He is categorical that Neanderthals and early Homo sapiens possessed language. Shared hunting, child-rearing, Ice Age survival, bodily pain, and individual identity required communication; once a name could attach to a person, labels could rapidly extend to animals, skins, babies, and the surrounding world.
His mechanism is cumulative creativity: “You have an idea, and the idea then drives the brain, and the brain has another idea.” Monumental architecture, trade, warfare, and ordinary coordination all become evidence against treating early people as cognitively lesser or “like gorillas.”
8. Translation fails when grammar hides probability and political risk
Finkel rejects the premise that one word precisely equals another across languages. A dictionary gloss such as “to divide” may cover ten English nuances whose endpoints barely resemble one another; context in a proverb, letter, medical text, or technical treatise changes the best translation.
His prized corrective is the Chicago Assyrian Dictionary, begun in the 1920s and built through decades of collaboration. He jokingly ranks it with the electric guitar as America’s greatest cultural achievement because its entries preserve contextual meaning rather than pretending “lexically A means B.”
The crucial unresolved problem is modality. An omen conventionally translated “If a lizard runs across the breakfast table, the queen will die” cannot realistically have promised certainty: a royal diviner would look foolish if she lived and dangerously implicated if she died.
Finkel therefore reads such lines as “could die,” triggering ritual action to avert danger. The diviner was part philosopher, part political adviser, judging what the king wanted and softening bad news “like a violin concerto.” Although these modal distinctions are not reflected in Akkadian grammar or writing, Finkel speculates—without proof—that speech may have marked them through a drawn-out vowel.
9. Mesopotamian divinity was practical, plural, and taken for granted
A Babylonian street drama shows divine life modeled on ordinary human conflict: Marduk is in bed on a roof with another goddess while his wife Sarpanitum hears them from the room. The women fight in the street and trade insults such as “slop bucket”—something Finkel compares to Verdi opera without music.
The pantheon often placed Anu, Enlil, and Ea near the top, then integrated hundreds of local deities into a theological “phone book.” Gods acquired household jobs and specialties; babies were probably placed under particular divine tutelage, while worshippers offered sacrifices or small “bribes” to remind occasionally inattentive protectors to perform.
Finkel distinguishes believing from assuming: Mesopotamians did not debate gods and ghosts but “took them for granted.” The dead passed into the netherworld; families buried relatives beneath courtyards and poured food or liquids through openings as continuing symbolic provision.
Lex challenges Finkel’s claim that prolonged grief is “infantile,” arguing that humans remain infantile and that finitude makes moments precious. Finkel then blames monotheism for dogmatic “I’m right and you’re wrong” conflict; Lex agrees it is dogmatic but emphasizes religion as an ancient tool for confronting suffering, loss, and mortality.
10. Gilgamesh preserves oral performance inside written literature
The best-known Gilgamesh epic comprises 12 tablets from Ashurbanipal’s library, but its hero was, in Finkel’s account, a historical king of Uruk whose reputation outlived him. Stories circulated orally before versions began appearing in Sumerian and Babylonian around 1800 BC.
Separate tales were eventually woven into a long, Homeric-scale narrative about adventure, humanity, divine power, and immortality. Finkel calls it “very Hollywoody,” but its sophisticated written form still exposes the mechanics of live storytelling.
Formulae such as “Gilgamesh opened his mouth to speak and addressed his friend Enkidu” are unnecessary for a silent reader. Finkel sees them as frozen performance cues once supplied by a narrator announcing which character was about to speak.
He extends that oral lineage back to fireside gatherings, famous hunts, battles, floods, discoveries, and stories mapped onto the stars. Cave paintings demonstrate that early populations already contained extraordinary artists; Finkel even imagines a small number of masters traveling between caves rather than every family producing its own Rembrandt.
11. The Ark Tablet anchors Noah’s flood in older Mesopotamian literature
A visitor brought Finkel an eight-by-three-inch tablet that he recognized as part of a flood narrative. Dating to about 1700 BC, its roughly 60 lines begin after the gods decide to erase an intolerably “noisy” population and one dissenting god warns Atra-Hasis.
The warning is also a blueprint: it specifies shape, materials, and quantities for a vessel carrying Atra-Hasis, his wife, three sons, and male and female animals. Its shock is geometric—the ark is not coffin-shaped but a gigantic circular coracle, based on boats actually used on Mesopotamian rivers.
George Smith had already found a later flood episode at Nineveh in 1872. Gilgamesh’s survivor Utnapishtim sends out three birds after the vessel rests, matching Genesis closely enough that Finkel calls the relationship “literary dependence,” not two independent reports of one event.
The 1700 BC tablet resolves the direction of borrowing for Finkel because it predates any plausible biblical composition by at least a millennium. Mesopotamia also supplies the environmental mechanism: its rivers could bring too little, too much, or catastrophically too much water, whereas “you never get floods in Jerusalem.”
12. A local disaster became an identity charter and universal countdown
Finkel believes a devastating Mesopotamian flood—perhaps a tsunami across the flat land toward the Persian Gulf—seeded the tradition, but not a worldwide event. He assigns “negligible” probability to Graham Hancock’s proposed Younger Dryas cataclysm connecting America, China, Mesopotamia, and other cultures.
In his reconstruction, Judean refugees spent roughly three generations in Babylon under Nebuchadnezzar. Sharing Aramaic and many cultural features with locals, they might otherwise have assimilated; authorities therefore drew up a historical charter, recycling Babylonian creation and flood material so that divine irritation at overpopulation became punishment for sin under the Hebrew God.
The narrative’s commercial durability lies in its clock: one ordinary person receives impossible instructions and must save life before disaster. Atra-Hasis was neither a king nor a sailor or boatbuilder; he receives a blueprint and must execute it. Finkel draws a direct line from Utnapishtim through Noah to Hollywood’s recurring “24 hours to think of a cure” structure—“tick, tick, tick.”
“Noise,” he adds, may euphemize overpopulation because death did not exist before the flood. Later fragments institute barren women, infertile men, and social roles that do not reproduce, producing what he calls a sophisticated Malthusian response rather than a literal complaint about interrupted divine naps.
13. Building the coracle tested engineering and oral credibility
Three specialists in reconstructing medieval Arab boats built a roughly one-third-scale vessel on a Kerala lagoon. Computer calculations set the maximum workable size: curved wooden ribs and woven structure had to support an extremely heavy waterproof coating of bitumen.
Iraqi bitumen could not be obtained, and the Indian substitute leaked slightly during the maiden voyage. Finkel’s retort—“You ever been in a rowing boat without water in the bottom?”—treated bailing as normal operation, not disproof; he believed the vessel could have reached ports.
More important, the prescribed material ratios scaled like a real coracle. Finkel imagines oral performers facing technically knowledgeable river communities who would demand, “So how big was it then?” Precise quantities gave storytellers credible answers, placing the tablet “on the cusp of purely oral literature to purely literary literature.”
14. The Royal Game of Ur balanced chance, strategy, and social friction
Excavations at Ur uncovered several 20-square boards, dice, and pieces in royal graves dating around 2600 BC. Related boards later appeared across Mesopotamia, Egypt, Syria, Lebanon, Jordan, Turkey, Greece, and Crete; Tutankhamun’s tomb contained examples, and the design endured for nearly 3,000 years.
The idiosyncratic layout evolved toward 12 central squares with four on each side at one end. Players apparently entered from a corner, turned into a shared avenue, and raced off the far end; rosettes likely granted safety or another throw, while stacking and blocking may have introduced tactical choices.
A second-century BC tablet—about 2,300 years later than the early Ur boards—preserved piece names and information about throws. Finkel worked backward to reconstruct playable rules; his version now circulates online and is even played again in Iraqi cafés.
Its staying power comes from a blend: unlike pure-luck Snakes and Ladders or pure-strategy chess, it offers “a measurable quota of strategy” under dice-driven uncertainty. A fragment lamenting “Oh my astragal” hints at gambling with knucklebones, while Finkel’s Getty demonstration rapidly had affectionate couples joking about lawyers.
15. The British Museum treats storage as a promise to the unborn
Finkel defines the British Museum not as an art museum but as a material narrative of humanity—achievement, suffering, failure, exchange, and survival without privileging one nation or religion. Its second constituency is “the unborn”: future generations whose questions cannot yet be anticipated.
That mission explains keeping objects off display. The approximately 130,000 cuneiform tablets are not idle inventory but “stockpiling for future examination”; as distance grows, interpretations shift, precedents sharpen, and contemporary parallels become visible in ways unavailable to earlier observers.
He calls the museum a “lighthouse” amid ignorance and imagines guiding hurried Martians through its rooms to explain Earth. Lex returns to Finkel’s governing metaphor: artifacts are raindrops from which future scholars may reconstruct the otherwise vanished waterfall.
Finkel’s modern anxiety is that electronic life behaves like an addictive drug and weakens precise language. Lex counters that slang always existed and that history will preserve the era’s consequential achievements, not its emojis; Finkel holds his ground that rich vocabulary comes from reading and that exact language enlarges thought by removing “wiggle room.”