The Derveni Papyrus – The 44-Year Silence Behind Europe’s Oldest Book

On 15 January 1962, archaeologists pulled the oldest book in Europe out of a funeral pyre in northern Greece. Within six months, the conservator hired to save it had sliced the scroll apart with a razor blade and lost the original sequence of its pages. The university then sat on the text for 44 years before finally publishing an edition that ignored the best digital scan available.

Discovery of the Derveni Papyrus and the 1962 Funeral Pyre Find

In January 1962, a road crew widening the national highway between Thessaloniki and Kavala broke ground on an unlooted nobleman’s grave at the Derveni pass, ten kilometres north of the city. Greek Archaeological Service staff moved in under the direction of Charalambos Makaronas. Cist graves at the site contained weapons, a decorated bronze krater and the remains of a funeral pyre resting on top of the covering slabs.

Archaeologist Petros Themelis was inspecting those ashes when he recovered the upper portion of a carbonised papyrus scroll. Most of the lower half had burned away during the original cremation. What survived had been baked in an oxygen-starved environment, turning the plant fibres to charcoal without reducing them to dust.

The text was a philosophical treatise on the birth of the gods, written in a mix of Attic and Ionic Greek, tentatively dated to between 340 and 320 BC.

It was, and still is, the oldest surviving manuscript in Europe.

Then the artefact arrived at the Archaeological Museum of Thessaloniki and broke. The stress of the move cracked the fragile roll into several pieces and exposed the written inner surface for the first time in roughly 2,300 years. No conservator had touched it yet.

Anton Fackelmann’s 1962 Conservation and Razor Blade Intervention

Makaronas knew his team lacked the technical skill to unroll a carbonised scroll without destroying it. He invited Anton Fackelmann, conservator of the papyrus collection at the Österreichische Nationalbibliothek in Vienna, to perform the conservation. Fackelmann’s reputation in the field rested on his earlier work with charred scrolls from the volcanic ruins of Herculaneum.

He arrived in Thessaloniki in July 1962 carrying the standard 1960s conservation toolkit. Plant sap and gum arabic went onto the fragile surface, then the treated roll was placed on a thick glass plate underneath an electric light bulb. The heat and the resulting static electricity were meant to force the fused layers apart.

When the layers refused to separate, Fackelmann reached for a razor blade.

By the end of the intervention he had cut the scroll into 266 individual fragments. Each curved piece was flattened, then sealed between nine pairs of glass plates. Pressing the carbonised layers flat crushed the brittle plant fibres at microscopic level, accelerating the deterioration of the surviving papyrus.

The recordkeeping is where the damage compounded. Fackelmann did not log the sequence in which he removed pieces from the main roll. Instead, the 266 fragments went into the glass frames in roughly descending order of size, ignoring the consecutive textual order in which the layers had been found.

He left no map of the original structure.

Independent researchers, most prominently the classicist Richard Janko, have argued that Fackelmann also used undocumented adhesive tape to physically bind unrelated pieces of papyrus together, fusing text from different columns. The official editors Theokritos Kouremenos, George M. Parássoglou and Kyriakos Tsantsanoglou dispute the scale of this practice in their published rebuttal of Janko’s review.

The one thin paper trail of Fackelmann’s work in the public record is a letter dated 6 November 1977, in which he wrote to the director of the Archaeological Museum confirming that he had treated the papyrus. Substantive technical detail did not accompany the letter. The adhesive chemistry, the precise timing of each razor cut, the original layer sequence: Fackelmann kept all of it in Vienna.

The 44-Year Publication Delay at Aristotle University

In 1964, Stylianos G. Kapsomenos published the final five columns of the text, columns 22 through 26, which were the innermost section of the roll and had suffered the least heat damage. The global academic community read this as a signal that a full edition was coming soon.

That assumption was wrong.

What followed was a 42-year publication blackout. After Kapsomenos died, stewardship passed to Kyriakos Tsantsanoglou and George M. Parássoglou at the Aristotle University of Thessaloniki, and the glass plates went into the museum’s secure storage.

The university structure permitted the lockdown. There was no external oversight committee with authority to set a deadline, and no obligation for the editors to share raw data with outside scholars.

In 1978, the museum commissioned Makis Skiadaressis to take a new set of infrared photographs of the fragments. With the pieces still sealed between Fackelmann’s glass plates, the camera captured severe reflections off the glass surface, obscuring edges and hiding faint traces of ink. The team accepted this flawed survey as their baseline, and it would still be their reference image set 28 years later.

Pressure to publish kept building. In 1982, Tsantsanoglou and Parássoglou submitted a notice to the journal Gnomonpredicting their comprehensive edition would be in print by mid-1984. They missed the deadline by 22 years.

The 1982 ZPE Leak and the Bootleg Derveni Papyrus Transcript

By the early 1980s, frustration in the wider community of classical scholars had reached breaking point. The official editors had refused requests to view the physical glass plates and had missed their own predicted 1984 publication target. No sign of imminent completion had reached the wider academic press.

Then the system broke from inside.

In 1982, an anonymous individual with access to the Aristotle University files supplied a preliminary transcript of the papyrus to Reinhold Merkelbach, the editor-in-chief of the German journal Zeitschrift für Papyrologie und Epigraphik(ZPE). Two possibilities sit on the documentary record. Either the leak came from a dissenting member of the official Thessaloniki team, or it was intercepted by a third party with access to the university archives. The cover letter and routing data are missing.

Merkelbach made an editorial decision that doubled the breach. He printed the leaked transcript as an unnumbered appendix placed after page 300 of his volume, so any reader could physically tear the section out and circulate it independently. No page numbers tied the document back to ZPE.

The text was incomplete. It contained missing columns and shifted column numbering, defects that would only be visible once the official edition appeared. Yet for the next 24 years, every working scholar studying the Derveni Papyrus was reading from the bootleg version Merkelbach had printed.

Apotheke 2 Storage and 2006 Multispectral Imaging Breakthrough

When the Archaeological Museum of Thessaloniki closed for remodelling in 2003, all nine glass plates were removed from public display. They were deposited in the main storage facility known to museum staff as Apotheke 2 (also recorded in some files as Apotheke 8).

The storage location was poorly chosen. Glass plates went on shelves near exposed heating ductwork. Ambient temperature inside the room frequently rose above 30 degrees Celsius.

That figure mattered specifically. Fackelmann’s 1962 gum arabic adhesive can soften and become tacky again at sustained high temperatures, risking further damage to the carbonised papyrus underneath. No published reading exists of what happened inside those frames during the remodelling period.

In April 2006, while the plates were still in storage, an independent research group from Brigham Young University, the Ancient Textual Imaging Group, secured permission to scan the papyrus. They used multispectral imaging, a technique that captures many separate digital photographs of the same surface, each filtered to record one narrow band of light spanning 380 to 1100 nanometres.

The BYU team filtered each capture to a narrow slice of wavelengths, between 10 and 40 nanometres wide. Where the 1978 infrared survey had been crippled by reflections bouncing off Fackelmann’s glass plates, this narrow-band approach successfully read through the glare. Letters that had been invisible to the naked eye appeared on the digital scans.

Six months remained until the official editors finalised their text.

The 2006 Derveni Papyrus Editio Princeps and Scholarly Disputes

Forty-four years after Fackelmann left Thessaloniki, the official edition finally appeared. In October 2006, Theokritos Kouremenos, George M. Parássoglou and Kyriakos Tsantsanoglou published the editio princeps of the Derveni Papyrus, the first official printed edition of the text. The book ran to a complete text, a translation, a commentary and a set of photographs of every fragment.

It also admitted to 113 unplaced pieces, some as wide as a third of a column.

The editors had based their final arrangement on the 1978 Skiadaressis infrared photographs. They had not integrated the April 2006 BYU multispectral data, even though those scans had been completed six months earlier and were in their possession.

An earlier signal had shown the team’s approach prioritised grammar over physical evidence. In 1997, Tsantsanoglou had published an edition of the badly damaged opening columns in Zeitschrift für Papyrologie und Epigraphik volume 118. That edition formally revised the scroll from 22 columns to 26 columns, restructuring the entire text based on the editors’ reading.

Then came the swerve.

Richard Janko, a classicist at the University of Michigan, set out to test the official arrangement against the physical record. Using digital enhancements of the published photographs, he built a physical model of the entire scroll, working from fibre patterns rather than Greek grammar.

Janko identified that a perfectly straight right edge on a piece known as fragment F3a was not a natural break, but a clean razor cut made by Fackelmann during the 1962 unrolling. He then located two narrow vertical strips, fragments I7 and I55, whose plant fibres matched perfectly across that razor cut. The strips fit into the section of the rolled scroll where one layer of papyrus had sat directly on top of the next.

This fibre match proved that the official editors had placed F3a, I7 and I55 in the wrong columns. The official team published a formal rebuttal in the Bryn Mawr Classical Review on 2 November 2006, disputing Janko’s findings and defending the methodology of the 44-year project. As of this writing, the argument has not been settled in print.

Source

Sources include: Cambridge University Press publication ‘The Derveni Papyrus’ (Betegh, 2004); the 2006 official publication rebuttal in the ‘Bryn Mawr Classical Review’; Richard Janko’s fibre-matching analysis in ‘Zeitschrift für Papyrologie und Epigraphik’ (2016); and multispectral imaging data from the BYU ScholarsArchive (2009).