Codice Maya de Mexico

The oldest surviving book of the Americas, and the only one of the four known Maya codices that is still in the hemisphere that made it. Ten pages survive of an original twenty, painted on three-ply fig-bark paper faced with gypsum, recording the synodic cycle of Venus in a hybrid notation that mixes Maya bars and dots with the single dots of Central Mexico. Radiocarbon and pigment analysis place it between 1021 and 1154.

Its entire modern history is a demonstration of what an absent find-spot costs. In 1965 or 1966 the Mexican collector Josue Saenz was telephoned by a man who called himself Gonzales, flown from Villahermosa in a light aircraft whose compass had been covered with a cloth, and set down on a remote airstrip in the foothills of the Sierra de Chiapas, where villagers showed him a codex, a mosaic mask and other objects said to have come from a dry cave. His own authenticator called them fakes. He bought anyway. The cave has never been located, identified or excavated.

Because there was no context, the manuscript was attacked as a forgery for forty years - most damagingly by J. Eric S. Thompson in 1975, who never saw it. It was defended by Michael Coe, who had shown it at the Grolier Club in New York in 1971 and who, asked at the time where it came from, would say only that it was a real hot potato lent by an anonymous owner. Vindication came from materials science: no modern pigment, Maya blue made with palygorskite, a gypsum ground, radiocarbon dates on the paper, and the demonstration that the sharp edges critics read as scissor cuts were breaks in the plaster and the work of arthropods. On 30 August 2018 the Instituto Nacional de Antropologia e Historia ratified it as authentic and renamed it the Codice Maya de Mexico. The find-spot is still unknown, and everything a proper excavation would have told us about the book is still gone.

Four books survived the burning, and this is the only one still in the Americas

The Maya wrote books. They made them from the inner bark of the wild fig, beaten into paper, faced with a fine plaster and folded like a screen, and they filled them with astronomy, ritual and prophecy. During the conquest these were read by Spanish clergy as falsehoods of the devil and burned in bulk, after which the men who had burned them expressed surprise that the Maya were distraught.

Three books survived, and all three left the hemisphere: the Dresden, the Madrid and the Paris codices, carried to Europe in the sixteenth century as curiosities. Until 1971 that was the entire corpus. It had been assumed that no more would ever be found, because fig-bark paper does not survive the wet, acidic soils of Mesoamerica.

This manuscript is the fourth, and it is the only one still in the Americas. It is also, on the Mexican determination of 2018, the oldest surviving book of the Americas - older than any of the three in Europe.

What it says is astronomy. The pages record the synodic cycle of Venus: ninety days of invisibility at superior conjunction, two hundred and fifty as Evening Star, eight days invisible at inferior conjunction, two hundred and thirty-six as Morning Star, five hundred and eighty-four days in all, and five of those cycles running exactly to eight solar years. Venus in Mesoamerica is not a goddess of love but a dangerous, warlike thing, and the figures on these pages are gods with spearthrowers, knives and captives. The book was a working instrument in the hands of a priest, for predicting when a hostile power would appear in the sky.

A telephone call, a covered compass, and a cave nobody has ever found

In 1965 or 1966 a man giving the name Gonzales telephoned Dr Josue Saenz, a Mexican collector, three times, and told him to take the first flight to Villahermosa in Tabasco and ask no questions. At Villahermosa strangers met him and put him into a light aircraft. The compass had been covered with a cloth. Saenz knew the country well enough to work out roughly where they were going, and they landed on a remote airstrip in the foothills of the Sierra de Chiapas.

Villagers brought out looted antiquities: a codex, a mosaic mask, and other pieces, which they said had been found together in a dry cave. They did not say where the cave was.

Two accounts of the transaction survive and they do not agree. Michael Coe, told the story by Saenz, records that Saenz was allowed to carry the objects back to Mexico City to have them authenticated before buying. Karl Meyer, who investigated the mask, records that the villagers refused and demanded a decision on the spot, that Saenz offered two thousand dollars, and that the villagers produced a Parke-Bernet auction catalogue to demonstrate that they knew what Mexican antiquities were fetching in New York. Saenz then wrote a post-dated cheque for a higher figure.

What is not in dispute is that his authenticator, the respected art expert Jose Luis Franco, judged the material to be fake, and that Saenz bought it anyway.

Sixty years later, that airstrip conversation is still the entire provenance of the oldest book in the Americas. The cave has never been located. No archaeologist has entered it. Nobody has ever been named as the finder.

Seven weeks in New York gave it a name it has now lost

Michael Coe, then the leading American Mayanist, examined the manuscript in Mexico City and was reasonably satisfied it was not a forgery. He obtained Saenz's permission to show it in an exhibition he was organising at the Grolier Club in New York, where it was on view from 20 April to 5 June 1971. From those seven weeks in a private bibliophiles' club the book took the name it carried for the next forty-seven years.

The exhibition catalogue Coe published in 1973 listed the owner as a private collection, New York. That was not true; the owner was a collector in Mexico City. Asked directly, that same year, by the journalist Karl Meyer where the codex had come from, Coe would say only that it was a real hot potato lent to the Grolier Club by an anonymous owner, and would say nothing more.

It is worth being fair about what that means. Coe was not the trafficker, and he was, on the record, right about the object: he backed its authenticity from the start and lived to see a national institute agree with him. But the concealment of an owner in a printed catalogue is exactly the practice by which unprovenanced material acquires respectability, and the reader should notice that the first published statement about this manuscript's ownership was a false one.

Thompson's attack, and the reason it landed

In 1975 J. Eric S. Thompson, the most authoritative Mayanist of the previous generation, published a paper declaring the manuscript a forgery. He had not seen it.

His arguments were not foolish. He pointed to iconographic anachronisms, and to a strange absence: Maya codices tell a reader what a date portends, and this one, while it tracks Venus, carries no omens and none of the cardinal directions that saturate other Maya inscriptions. Claude-Francois Baudez later developed the point - the book, in this reading, is useless, and a real ritual manuscript is not useless.

Thompson also had an answer to the radiocarbon date. Blank fig-bark paper does turn up in caves; looters find it and discard it because buyers do not want it. An enterprising forger, he suggested, would simply acquire old paper and paint on it, and the paint would be what mattered. That objection was correct in principle, and it took forty years and a laboratory to close.

The reason the attack landed so hard is the reason this record sits in an archive about provenance. There was nothing to answer it with. No excavation, no context, no associated material recorded in place, no findspot, no witness, no photograph in the ground. The only alternative to Thompson's story was a collector's account of a covered compass and a cave, and an account is not a document.

The paper, the plaster and the blue

What eventually answered Thompson was materials science, and it is worth setting out, because it is a model of how an unprovenanced object can still be interrogated.

The support is amate, fig-bark paper, and it is three-ply, which is what allows a screenfold to bend without splitting. Both faces were prepared with a thin ground of gypsum, calcium sulphate, between two and three tenths of a millimetre thick, to give a smooth painting surface. Only one face was ever finished.

The palette has been identified: lamp black; red from hematite; brown from cochineal; and Maya blue, the famously stable pigment made by binding indigo dye into the clay mineral palygorskite. In 2007 the Institute of Physics at the National Autonomous University of Mexico ran non-destructive tests and identified the palygorskite. No modern inorganic material was detected anywhere in the manuscript.

Radiocarbon has been applied three times over forty-five years: a 1972 measurement gave 1035 to 1431, a 2012 test gave 1050 to 1284, and the Mexican programme narrowed the range to 1021 to 1154. Those are dates for the paper. The Mexican studies went further and established that the pigment is contemporaneous with the support, which is the finding that answers Thompson's old-paper argument directly.

The physical objections went the same way. Sharp edges that critics read as knife cuts turned out to be breaks in the gypsum ground, and elsewhere the crisply chewed margins left by small arthropods that had lived in the book. The manuscript passed through at least three episodes of high humidity in its life. Its pages were torn apart by rough handling in the trade; only pages four, five and six are still joined to one another.

What the object says about itself

Read as an object rather than as a case, the manuscript is a technical performance.

Each surviving page carries a standing deity facing left, a column of repeated day signs down the left edge, and ring numbers across the top margin. The ring numbers do arithmetic. On page seven a bar and three dots inside the ring instruct the reader to count forward eight days, the exact period of Venus's invisibility at inferior conjunction; on page eight the figure sixteen inside the ring is added to eleven dots each standing for twenty, giving two hundred and thirty-six, the exact period of visibility as Morning Star.

The gods are violent. Page one has K'awiil taking a captive; page two a death god; page six a skeletal deity, identified by John Carlson as a version of Tlahuizcalpantecuhtli, the Central Mexican Morning Star, holding a decapitated god by the hair with a jagged flint blade set in his nose; page eight a bird deity who has shot the temple in front of him with an atlatl dart; page ten a death god launching a dart into water to strike a shell. The numeration is hybrid, mixing Maya bars and dots with the single dots of Central Mexico and Oaxaca, and the headdresses have close parallels in the Venus pages of the Codex Vaticanus B.

This is why the book matters beyond the courtroom. It is surviving evidence that Maya and Central Mexican systems were being used together in a single manuscript in the eleventh or twelfth century, in the decades when Tula and Chichen Itza were declining. The style has been connected backwards to the late painting at Chichen Itza, where, as here, the final painted line wanders freely from the underdrawing, and forwards to the Mixtec codices, whose headdresses appear here first.

Forty years of argument, and one study that ended it

The literature on this manuscript is easy to summarise: a queue of scholars publishing for and against, none of them able to settle it, because the question was not really about the object.

Against: Thompson in 1975, Susan Milbrath in 2002, Baudez in 2002. For: David Stuart, John Carlson, and Victoria and Harvey Bricker, who pointed out that the contents are not copied from any known codex and yet are consistent with an accurate pre-Hispanic calendar - a difficult thing for a forger to manage. In 2008 Ruvalcaba, Zetina and colleagues published a non-destructive study which found no modern materials but also flagged what they called induced degradation, and concluded, honestly, that their work neither supported nor refuted authenticity.

In 2015 Coe, Stephen Houston, Mary Miller and Karl Taube published a full-size facsimile and a comprehensive study, using photographs taken by Enrico Ferorrelli for National Geographic in 1987, together with a full set of hand drawings released uncopyrighted so that anyone could work from them. They answered the calendrical objections, showed that the suspect cuts were plaster breaks, that the layout followed the sketch-and-grid practice of Maya mural painting, and that the pigments included Maya blue, which is exceptionally difficult to reproduce. Their closing question was whether the painter of this book could have lived in the twentieth century, and their answer was no.

Then, on 30 August 2018, the Instituto Nacional de Antropologia e Historia published its bulletin ratifying the manuscript as authentic and as the oldest book of the Americas, renaming it the Codice Maya de Mexico. Nobody has seriously argued the other way since.

A seizure, or a donation? Nobody has established which

The legal history of this object is a small scandal of imprecision, and it deserves to be set out exactly.

One account holds that in 1976 the Attorney General of Mexico invoked the treaty of July 1970 between the United States and Mexico on the recovery and return of stolen archaeological, historical and cultural property, and that this produced a seizure of the codex and its official return to Mexico. That treaty has been invoked very few times, which would make this one of its significant applications.

Most other accounts say simply that Josue Saenz donated the manuscript to the Mexican nation, and give neither a date nor a donor. Trafficking Culture, which went looking, records the timing and circumstances as unclear, and notes that the volume usually cited for the return does not in fact cover it.

These are not the same event. A treaty seizure is a coercive act premised on the object being stolen property of the Mexican state. A donation is a gift, with the tax and reputational consequences a gift carries, and it implies that title was the donor's to give. One of them happened, and this record cannot say which.

There is a further wrinkle in Saenz's own words. He told an interviewer that he stopped buying pre-Hispanic art in 1972 because a new Mexican law made all unregistered antiquities the property of the nation, and added that the law ended lawful collecting without solving the problem of preserving the nation's archaeological heritage. A man who says that has understood precisely what changed. What he does not say is what it meant for the codex he had bought six years earlier.

The mask that may or may not have been in the same box

A mosaic stone mask was bought in 1966 by Mildred Barnes Bliss, co-founder with her late husband of Dumbarton Oaks in Washington, and it is there still.

Coe published that the mask and the codex came out of the same cave and were shown to Saenz together at the airstrip. The route the mask took afterwards is better documented than the codex's: Saenz, on Jose Luis Franco's opinion that it was a fake, tried to stop his cheque and found it already cashed, then passed the mask to a dealer with instructions to sell it without any guarantee of authenticity, and it moved through a New York dealer into the United States. Gordon Ekholm, curator of Mexican archaeology at the American Museum of Natural History, examined it under a microscope, found what he took to be evidence that a beard had once been attached - a detail he thought a forger would not invent - and pronounced it genuine, which by Meyer's account considerably enhanced its value. It is now widely accepted as authentic.

Whether it and the codex were ever in the same box is not established. Trafficking Culture points out that the small plane and the cave appear nowhere in the literature around the codex at the time of the 1971 exhibition, and that the connection rests on Coe's later telling and on a dealer's word to a journalist.

The consequence is stated best by the researcher who assembled the case. Because it is unknown whether the mask and the codex were found together, and because both were looted, the opportunity to study what may have been an entirely unique Maya funerary context has been lost. Two objects, both now accepted as genuine, both in public collections, and the relationship between them permanently unrecoverable.

The counts do not agree, and neither do the names

A reader coming to this manuscript from the literature should be warned that the basic figures differ between authorities, and this record does not smooth them over.

The number of surviving leaves is given as ten in the current Mexican account and as eleven by Trafficking Culture. The original length is given as twenty pages in one place and twenty-one in another. The surviving pages are numbered one to ten in the modern convention even though the first eight and the last two of the original are the ones missing, which is a scheme guaranteed to mislead. The 1971 exhibition is named as Ancient Maya Calligraphy in one account and as The Maya Scribe and His World in another; the latter is certainly the title of Coe's 1973 book. Eight blank, unstuccoed sheets of fig-bark paper are associated with the manuscript, and one of them produced the 1230 plus or minus 130 radiocarbon date that Coe published in 1973 - a date, that is, measured on a blank sheet rather than on a painted page.

And a claim now mirrored across the internet, that the manuscript first appeared at an auction in the late 1960s, cannot be traced to any record of a sale. The only auction in the story is a Parke-Bernet catalogue held up by looters at an airstrip to prove they knew what the New York market paid. A sales catalogue in a seller's hand is not a provenance.

What was actually lost

The codex is authentic, it is in Mexico, it has been published in facsimile twice and it was shown to the public in its own country for three weeks in the autumn of 2018 - the first time most Mexicans could see the oldest book of their own hemisphere. On any reasonable measure this is among the better outcomes in this archive.

Set against that: no one knows where it was found, what it was found with, whether the mask belongs to it, what deposit it came from, whether the cave holds anything else, or whether the missing pages survive. The eight blank sheets exist but their relationship to the painted ones is a matter of assertion. Forty years of scholarly effort went into proving that the object is old, which an excavation would have established on the first day, and that effort was subtracted from the work of reading it.

Michael Coe put the position more sharply than any of his critics. The irony of the whole business, he wrote, is that if the codex had been found in an old library it would be accepted by even the most rock-ribbed scholar as genuine. The same is true of a trench. The doubt was never about the paint. It was about the hole in the ground that nobody has ever seen.