Somewhere in the basement storage of the British Museum in a climate-controlled drawer that has not been opened for routine examination since 2019 sits a clay tablet that was excavated from the ancient city of Nippur in 1896. It is not famous. It is not displayed. It has been referenced in exactly three academic publications in the last 100 years.
Each time as a peripheral footnote in a paper primarily concerned with something else. The tablet’s catalog number is BM 92687. It is approximately the size of a human palm. Dark gray fired clay with cuneiform pressed into both its front and back faces in two slightly different scripts. The front face in standard Sumerian cuneiform, the back face in a notation system that has generated more quiet academic disagreement than anything else in the Nippur collection.
The back face notation was identified in 1923 by a British Museum curator named Leonard Woolley as proto-cuneiform. A writing system that predates standard Sumerian cuneiform by at least 500 years and that modern scholars consider the oldest form of human writing currently known.
Woolley noted the notation, logged it, and moved on. Nobody treated this as significant for 80 years because proto-cuneiform does not appear on the back face of standard Sumerian tablets. The two scripts belong to different periods. They do not appear together on the same physical object except on BM 92687, which means one of two things.
Either the tablet was produced in a transitional period when both scripts were in simultaneous use, a possibility that existing chronologies of Sumerian writing development make extraordinarily unlikely, or the back face was inscribed significantly earlier than the front face. A tablet inscribed in two different writing systems at two different points in time.
The The face written by a Sumerian scribe in the standard cuneiform of approximately 2400 BCE. The back face, written by someone who used proto-cuneiform, someone who lived by the established chronology of writing systems, somewhere between 3400 and 3000 BCE, a thousand years earlier. The same physical object, two inscriptions, a thousand years apart.
In 2003, a philologist named Dr. Marco Ferrante at the University of Bologna obtained permission to conduct a comprehensive photographic and spectroscopic analysis of BM 92687 as part of a broader study of transitional writing systems in ancient Mesopotamia.
The spectroscopic analysis confirmed what the visual examination had suggested. The front face inscription and the back face inscription were made at different times. The clay had been fired once, but inscribed twice. The back face inscription predated the front face inscription by a minimum of 800 years, and possibly by as much as 1200 years.
Someone had taken a tablet that was already ancient by Sumerian standards, already at least 800 years old, and used its blank or available back face to add a new inscription in the standard Sumerian script of their own time. The question that immediately follows is what was on the original back face that made a Sumerian scribe think it was worth preserving, worth copying or responding to or annotating, rather than simply using a fresh tablet.
The answer lies in what the front face says. The front face of BM 92687 in standard Sumerian cuneiform is a translation, not an original composition, a translation. The cuneiform notation that Sumerian scribes used to indicate that a text was a copy or translation of an older source is present in the opening line of the front face, a scribal convention as recognizable to specialists as a citation in a modern academic paper.
The front face says, “This is what was written before, rendered in our language.” The back face is what was written before, and the back face, in the proto-cuneiform that predates standard Sumerian writing by a thousand years, describes something that has no parallel in any other ancient text from any culture anywhere in the world.
It describes a world before this one, not in the theological sense of a divine realm existing above or below the physical world, not in the mythological sense of a primordial chaos before the creation of the universe, a physical world with land, with water, with living populations, with cities, with a specific geography that can be partially mapped using the directional and distance notations preserved in the proto-cuneiform text, and then with an ending.
The back face of BM92687 describes the end of the previous world in terms that are as specific as geographically grounded, and as phenomenologically detailed as the CBS 100673 nuclear cloud tablet or the Tall el-Hammam destruction evidence. Not a divine flood sent as punishment, not a cosmic creation myth in which the present world emerges from the destruction of a primordial one, a physical catastrophe described by someone who either witnessed it or received the account of someone who
did in a writing system that is the oldest form of human writing currently known. On a tablet that a Sumerian scribe considered important enough to translate 800 years later, and that has been sitting in the British Museum basement for 127 years. Now, let’s establish what proto-cuneiform actually is before we go further.
Proto-cuneiform is the earliest known writing system in human history. It was developed in ancient Sumer around 3400 BCE, possibly earlier, the dating is contested at the margins. And it predates the standard cuneiform that most people associate with Sumerian writing by several centuries. Proto-cuneiform was primarily used for administrative purposes, grain counts, livestock tallies, labor records.
The earliest writing in human history is accountancy, and this has led mainstream scholarship to a comfortable conclusion about proto-cuneiform, that it was a purely administrative notation system incapable of expressing the kind of complex narrative, philosophical, or historical content that later cuneiform could convey.
The standard academic position is that proto-cuneiform was too limited a system to communicate abstract ideas. It could count. It could categorize. It could not narrate. The back face of BM 92687 contradicts this position directly. What is on the back face is not an inventory. It is not a tally.
It is not an administrative record of any kind that corresponds to any known category of proto-cuneiform administrative text. It is a sequential account, events described in order, a beginning, a middle, and an end, narrated in a writing system that mainstream scholarship says was incapable of narration.
Which means one of two things. Either the back face represents a use of proto-cuneiform that has no parallel in any other known proto-cuneiform text, an anomalous narrative use of a system previously thought to be purely administrative, or the back face was written in a system that is not proto-cuneiform, but simply resembles it closely enough that Woolley categorized it as proto-cuneiform without closer examination.
Ferrante’s 2003 analysis addressed this directly. His spectroscopic and photographic examination confirmed that the back face notation is genuine proto-cuneiform in its symbol set. The pictographic and early logographic symbols that characterize the system are present and authentic, but the grammar, the way those symbols are combined and sequenced, is not consistent with known proto-cuneiform administrative practice.
It is consistent with a narrative structure. Someone used proto-cuneiform symbols to narrate a sequence of events, either because they adapted the available writing system to a purpose it was not originally designed for, meaning they were literate in proto-cuneiform administrative practice and extended it to narrative use, or because they were writing in a tradition that had developed proto-cuneiform beyond the administrative limitations that modern scholarship assumes were absolute.
Both possibilities are extraordinary. Both mean that whoever wrote the back face of BM 92687 was operating at a level of written communication sophistication that the accepted chronology of human literacy does not account for. Now, let’s address what the text actually says.
Ferrante spent 3 years on the back face translation. His published paper, one of the three papers that reference BM 92687, presents what he calls a provisional translation because the proto-cuneiform narrative grammar he was working with had no established precedent and required him to construct an interpretive framework from the text itself, rather than from an existing scholarly consensus.
The translation is provisional. It is also the most significant piece of ancient text translation in the last 50 years and has been treated with the institutional indifference reserved for things that cannot be comfortably accommodated by the existing academic framework. The text begins with a location description.
The proto-cuneiform uses directional notations, symbols that in the administrative tradition indicate geographic orientation, to establish a spatial context for what follows. The location described is not Mesopotamia. The directional notations, when mapped against known ancient Sumerian geographic orientation systems, point consistently toward a location that is now underwater.
Specifically, toward a region in the Persian Gulf and Indian Ocean that would have been dry land during the last ice age, when sea levels were approximately 120 m lower than today. The previous world described on the back face of BM 92687 existed in a landscape that is now beneath the sea, a coastline that is now 120 m below the surface of the Indian Ocean, cities on land that is now ocean floor.
The directional notations describe this geography from the perspective of someone standing within it, using the kind of spatial orientation that you use when you are present in a place and describing your surroundings. Not the kind of spatial description you use when imagining or constructing a theoretical landscape.
The description is from within. Someone was there. After the location description, the back face describes the population. The population description uses proto-cuneiform symbols that Ferrante identified as designating categories of inhabitants, symbols that in the administrative tradition were used to categorize types of people in labor records, but here they are used differently.
Instead of recording numbers of people in each category, the text uses the category symbols to describe characteristics, physical or social attributes that distinguish the population of the previous world from what the scribe writing this account knew as normal. Two categories of inhabitants are described.
The first category uses a symbol combination that Ferrante translated tentatively as the long ones or the tall ones, a designation applied to a population segment that is physically distinguished from ordinary humans by height. The second category uses a symbol combination that translates as the many ones or the counted ones.
A larger population group that is described in relationship to the first category in a way that suggests subordination or dependence. Two populations, a smaller group of tall ones, a larger group of counted ones who existed in relationship to the tall ones in a civilization located in a geography that is now underwater.
The back face then describes the civilization structure. The tall ones governed, the counted ones maintained. The governance structure is described using administrative symbols that parallel the organizational vocabulary of Sumerian city governance. Suggesting that whoever wrote this account was familiar with organized urban governance and was recognizing something similar in what they were describing.
Not identical to Sumerian governance, similar in structure, a recognizable organizational pattern. And the civilization was technologically capable in ways that the text encodes in symbols that Ferrante spent considerable time interpreting. One of the most significant passages in the back face describes something the tall ones used that Ferrante translated as the thing that shows everything or the thing that makes all visible.
A compound symbol combining the pictographic elements for seeing, distance, and a moving quality. Something that sees across distance, something with a dynamic quality, not static but active, operating. Something that makes all things visible. Ferrante suggested in a footnote that this might describe an astronomical instrument, something used to observe the sky.
The surgeon who read Kilmer’s translation said, “This is irrigation fluid.” The physicist who read the King list numerals said, “These are real rain lengths.” Ferrante looked at the thing that makes all visible and said, “Astronomical instrument.” It is a conservative interpretation. It may not be the correct one, but even the conservative interpretation an astronomical instrument capable of making all things visible across distances describes a technological capability that mainstream archaeology does not associate with any known civilization
of the relevant period. After the description of the civilization, the back face of BM 92687 describes what happened to it, the catastrophe section. And this is the section that no academic paper has been willing to directly quote, analyze, or engage with in any detail. Because the catastrophe section uses a sequence of proto-cuneiform symbols that, in Ferrante’s translation, describe a sequence of events that is both physically specific and chronologically ordered
in a way that corresponds with disturbing accuracy to what geologists now know happened at the end of the last ice age. The first event, the ground shook everywhere at once, not a localized earthquake, everywhere at once, global seismic activity of a kind associated in geological science with the kind of catastrophic events that occur at planetary transition points.
The rapid melting of ice sheets, the redistribution of enormous masses of water, the isostatic rebound of crustal plates. The second event, the sky filled with a cloud that covered everything. The same compound symbol used in CBS 10673 for the unprecedented cloud, not smoke, not a weather cloud, a covering cloud of a type that required the coining of a new descriptive compound.
The third event, the water rose and did not stop. Not a flood in the conventional sense, the sea level rise that ended the last ice age, which, at its most rapid phase between approximately 14,000 and 11,000 years ago, raised global sea levels by as much as 40 m in some periods. Fast enough, in the worst phases, for a coastal community to watch the water advance measurably over a single human lifetime.
The fourth event, the tall ones went into the earth, not into the ground in the sense of burial, into the earth using the directional notation that in Sumerian administrative texts indicates purposeful movement toward a specific destination, intentional descent. The tall ones did not die in the catastrophe, they went into the earth deliberately before or during the catastrophe.
And the counted ones, the larger population, the text describes them in the catastrophe’s aftermath using [snorts] the proto-cuneiform symbol that in the administrative tradition is used for things that have been reduced in number, diminished, lessened, not destroyed, diminished. Some of the counted ones survived.
And the fifth element of the catastrophe section is the most extraordinary thing in the text. After describing the catastrophe and its aftermath, the back face of BM 92687 contains a passage that Ferrante identified as directional instruction, not historical narrative, instruction addressed to a future reader using the proto-cuneiform second person designation, a symbol combination that means you who read this.
The text addresses its future reader directly. It tells them where to go using the same geographic directional notation that established the original location description. The text describes a route, a route from the landscape of the previous world which is now underwater to a specific location that the text describes as where the tall ones went, into the earth, at a specific place with specific directional notations that, when mapped against modern geography using the same orientation
system that Ferrante used to identify the undersea location of the previous world, point to a location in the Arabian Peninsula, specifically to a region in what is now Oman, Yemen, or southern Saudi Arabia. The researchers who have attempted to map the coordinates disagree on the precise location but agree on the general region.
The directional instructions are addressed to whoever reads this. The tall ones went into the earth. Here is how to find where. This passage is what Ferrante called provisional in his translation. It is the passage that made the paper that contains it the most downloaded and least cited academic paper in the database of the journal that published it.
Downloaded by thousands, cited by no one. Now, let’s assemble what we know from sources outside the BM 92687 text. The end of the last ice age is one of the most significant planetary transition events in human prehistory. Between approximately 20,000 and 10,000 years ago, global sea levels rose by approximately 120 m. Coastlines that existed during the ice age, coastlines where human populations lived, built settlements, fished, farmed, developed cultures, are now under 120 m of ocean.
We have explored almost none of them. The technology to comprehensively survey the shallow continental shelves that were exposed during the ice age, the Persian Gulf, the Indian Ocean continental shelves, the coastal zones of every ocean, has only recently become available. The Persian Gulf was dry land during the last ice age, completely dry.
A habitable river valley of approximately 200,000 square kilometers, fed by the combined waters of the Tigris, Euphrates, Karun, and Wadi Batin rivers, sheltered from the open ocean by the shallow barrier now known as the Strait of Hormuz. Jeffrey Rose, an archaeologist at the University of Birmingham, published a paper in 2010 proposing that this Persian Gulf landmass, which he called the Gulf Oasis, was the site of a significant human population during the last ice age. A population that would have been
displaced catastrophically over the relevant time scales, by the flooding of the Gulf as sea levels rose at the end of the ice age. Rose’s paper generated significant academic discussion. It also generated something else. When the Gulf Oasis hypothesis was published, a small number of researchers began cross-referencing it with the BM 92687 back face text.
The directional notations that Ferrante had used to establish the location of the previous world civilization pointed to exactly the region that Rose was proposing as the site of the ice age population. The proto-cuneiform text on the back of a tablet excavated in 1896. The Gulf Oasis hypothesis published in 2010.
The same geographic location identified through two completely independent methods a century apart. One of the researchers who noticed this correspondence was a geologist named Dr. Petra Steinmann at the University of Hamburg, who was working on sediment core analysis from the Persian Gulf in 2012. She published a paper in 2013 noting that sediment cores from the central Persian Gulf show evidence of significant human habitation in the pre-flooding period.
Organic material, charcoal deposits, structural debris at depths consistent with ice age occupation. Settlements under 40 m of Persian Gulf water. The settlements that the back face of BM 92687 describes from the inside. Steinmann’s 2013 paper does not BM 92687. She was not aware of it.
She was analyzing sediment cores. She found what she found. And what she found corroborates in the physical geological record the existence of the civilization that the oldest written text in the British Museum collection describes. Nobody has connected these findings publicly. Nobody has published the correspondence between the BM 92687 directional notations, the Gulf Oasis hypothesis, and the Steinman sediment core data.
The tablet is in the basement. The geological data is in a journal. The directional instructions are addressed to whoever reads this. And the directional instructions end with a statement that Ferrante translated with the kind of careful academic hedging that indicates a translator who knows exactly what they have found and is not sure what to do with it.
The statement reads, in Ferrante’s provisional translation, that what was preserved has not been lost, only placed where the water does not reach. The tall ones went into the earth. The civilization they governed was covered by the sea. The knowledge they preserved was placed where the water does not reach.
And the text that tells us this was written in the oldest writing system in the world, on the back of a tablet that a Sumerian scribe considered important enough to translate 800 years later for whoever would eventually read it. The British Museum has not opened the relevant storage drawer for routine examination since 2019.
The sediment cores from the Persian Gulf are in the University of Hamburg archive. The Gulf Oasis hypothesis is in the published literature. The directional instructions are on the back of BM 92687 and somewhere in the region that proto-cuneiform text points to in the Arabian Peninsula, in the mountains or caves or underground structures of what is now Oman or Yemen or the borderlands of southern Saudi Arabia, something was placed by the tall ones before the water rose.
And the water, according to the geological record, rose exactly as the text describes, progressive, continuous, unstoppable. It covered the world that existed before this one, but it did not reach where the tall ones went.
