Why Lost Books Survive: Archaeology and the Accidental Preservation of Ancient Literature

A Homeric scrap where it should not be

During excavations at the necropolis of ancient Oxyrhynchus in November and December 2025, a team led by Núria Castellano uncovered a Roman-period mummy in Tomb 65, Sector 22. Against the abdomen lay a papyrus incorporated into the embalming arrangement. When the object was examined during the following campaign by conservator Margalida Munar, papyrologist Leah Mascia and classical philologist Ignasi-Xavier Adiego, the writing was identified as Greek and the text as part of Book II of Homer’s Iliad, from the Catalogue of Ships.

There is nothing surprising about finding Homer at Oxyrhynchus. The city has yielded Homeric papyri since the earliest excavations there, alongside pieces of Sappho, Pindar, Menander, Sophocles, Euripides, Plato and many other authors. What is unusual is the archaeological position of this particular fragment. The University of Barcelona mission had previously found Greek papyri in comparable positions on mummies, but these carried magical or ritual texts. According to the excavators, this is the first literary Greek text they have documented as deliberately incorporated into the mummification process.

That distinction needs to be handled carefully. The secure observation is that a papyrus containing Homer was placed on the body as part of the funerary treatment. It does not automatically follow that the passage was selected because it was Homer, still less that the Catalogue of Ships had acquired some funerary meaning. At present, the find has been announced by the excavation team but the full archaeological and papyrological study is still in progress. The individual is known to have been an adult male, but further investigation of the mummy and its context was still under way when the discovery was reported.

The attractive interpretation is obvious: perhaps Homer accompanied an educated Greek-speaking reader into death; perhaps familiar poetry was converted into a protective text; perhaps the choice says something about cultural identity in late Roman Egypt. All are possible. None is yet demonstrated.

There is, however, another possibility, and archaeology has encountered it many times before. The papyrus may have been useful precisely because it was no longer valued as a book.

The archaeology of waste paper

Oxyrhynchus is an unusually appropriate place for such a problem because its contribution to classical literature has always depended upon rubbish.

When Bernard Grenfell and Arthur Hunt began excavating the site in 1897, much of what they encountered consisted of ancient refuse mounds accumulated around the city. Papyrus had been carried out with ordinary waste, layer after layer, and the dry conditions at the desert margin permitted material that would normally rot to survive. The result was an enormous and highly irregular archive of discarded writing: tax records, petitions, accounts, private letters, official correspondence, contracts and lists mixed with literary manuscripts. Literary works constitute only a minority of the material, yet that minority transformed modern knowledge of Greek literature.

The archaeological paradox is worth dwelling on. The texts survived not because the rubbish dump was a library but because it was not one.

A manuscript kept in active circulation was repeatedly handled. It could tear, be recopied and discarded, or eventually disappear through damp, insects, fire or the ordinary destruction of buildings. A sheet thrown into a dry rubbish mound might instead enter an environment in which biological decay slowed dramatically. Cultural rejection became the first stage of archaeological preservation.

The Oxyrhynchus finds therefore give us something very different from the remains of a deliberately curated library. They provide a sample, distorted but enormous, of written material that passed through the hands of inhabitants over centuries. Some famous authors reappeared there after having virtually vanished from the medieval manuscript tradition. Poems of Sappho survived in scraps. Plays known only by title acquired words again. Portions of otherwise lost works by Aeschylus, Sophocles, Menander and others emerged from refuse rather than from the shelves of a monastery.

This is one reason archaeologists must resist treating textual survival as a simple measure of ancient popularity. What survives is filtered twice: first by ancient practices of use and disposal, and then by the physical conditions of burial.

At Oxyrhynchus those filters happened, unusually, to favour papyrus.

When writing becomes raw material

Discard was not the end of a papyrus sheet’s usefulness. In Egypt, old writing material could enter a second economy in which its textual content was largely irrelevant.

During the Ptolemaic period, discarded papyri were among the materials used to manufacture cartonnage: layers of papyrus or linen combined with adhesive and plaster to form masks, panels and other coverings for mummified bodies. Once incorporated, an obsolete account, petition or letter ceased to function primarily as a text. It had become structural material. Modern papyrologists consequently gained access to documents because ancient craftspeople had recycled them.

Tebtunis in the Fayum provides one of the clearest demonstrations. Grenfell and Hunt's excavations there at the turn of the twentieth century produced written documents from the cartonnage and stuffing associated not only with human mummies but with mummified crocodiles. A large group belonged to the archive of Menches, village scribe of Kerkeosiris in the late second century BC. Administrative reports, correspondence and petitions that had once belonged to working bureaucratic files eventually became material used in the preparation of sacred animals. A separate group of forty-five Greek documents came from only five crocodile mummies buried together.

Few examples expose preservation bias more neatly. An administrative document may survive because someone regarded it as waste. The bureaucrat who produced it, the person who discarded it, the priest who reused it and the modern scholar who reads it assign entirely different values to the same object.

The new Oxyrhynchus mummy may belong conceptually to this world of secondary use even though its precise treatment differs from conventional cartonnage. The papyrus was placed against the body rather than merely recovered as an anonymous layer of papier-mâché. The presence of other Greek papyri with magical or ritual content in similar positions makes a ritual interpretation reasonable. Yet the Tebtunis material warns against assuming that the meaning of the writing always governed its reuse. What mattered to the embalmer could have been the sheet, the bundle, its association with writing in general, or its availability, rather than the particular verses it contained.

This distinction is especially important in the case of Homer. The Iliad was deeply embedded in Greek education in Roman Egypt, and fragments of Homer are abundant among Egyptian papyri. A discarded Homeric manuscript may therefore tell us as much about the ordinary circulation and replacement of literary texts as about the beliefs of a particular deceased person.

There is no need to choose prematurely between “sacred Homer” and “recycled scrap”. The archaeological context may eventually narrow the possibilities. For the moment it does not eliminate either.

The history of cartonnage also raises a more modern problem. Earlier attempts to extract hidden papyri sometimes required dismantling the object that preserved them. That exchanges one form of evidence for another: a text becomes accessible, but information about manufacture, construction and funerary practice may be damaged or lost. Heritage-science experiments have consequently investigated combinations of optical, X-ray and terahertz methods that might detect writing inside cartonnage non-destructively. Results have shown both the promise and the difficulty of the approach; no single imaging method works reliably for every combination of papyrus, ink, plaster and adhesive.

The methodological principle is now difficult to dispute. A hidden text is not automatically more important than the archaeological object hiding it.

When destruction preserves

Egyptian dryness is only one route by which ancient books survived. Elsewhere, preservation could result from events that should have destroyed them completely.

The Derveni Papyrus was discovered in 1962 near Thessaloniki in the remains associated with a funerary pyre beside a wealthy Macedonian burial. It is a philosophical and religious commentary on an Orphic poem, usually dated palaeographically to the later fourth century BC. Papyrus rarely survives in the climate of northern Greece. The Derveni roll survived precisely because it had been carbonised by fire. The burning that made the object brittle also converted it into a form capable of persisting archaeologically. It is now generally described as the oldest surviving papyrus book found in Europe.

Whether the roll was intentionally consigned to the funerary fire, accidentally caught in it, or played some more specific role in the rites has been debated. Once again, archaeological context matters as much as literary content. A philosophical text found beside a pyre is not simply a manuscript with an unusual provenance; its physical condition and deposition are part of the evidence for interpreting what happened.

At Herculaneum the same preservation mechanism operated on an altogether different scale. When Vesuvius erupted in AD 79, hundreds of papyrus rolls in the Villa of the Papyri were carbonised and buried. More than 1,800 papyri and fragments are associated with the collection. The fire and volcanic deposits destroyed the villa as a living environment while preserving the only large body of a classical library to survive archaeologically. Much of its known content is philosophical, with a particularly strong representation of works by the Epicurean Philodemus of Gadara.

For centuries, preservation and access were in conflict. Physically opening the blackened rolls risked breaking them into fragments. That problem has turned Herculaneum into an important testing ground for non-invasive imaging. X-ray phase-contrast tomography demonstrated that the internal geometry of carbonised rolls could be reconstructed and portions of writing recovered without mechanically unrolling them.

The technological development has accelerated. In June 2026, Giorgio Angelotti, Federica Nicolardi, Paul Henderson and W. Brent Seales published peer-reviewed work showing that fine surface morphology associated with writing can itself help distinguish inked from uninked regions of Herculaneum papyrus, an important result because carbon-based ink and carbonised papyrus can be frustratingly similar in ordinary X-ray attenuation.

The same month, a larger research team reported the complete virtual unwrapping of the preserved writing surface of PHerc. 1667 and the recovery from another sealed roll, PHerc. 139, of a title identifying it as Philodemus’ On Gods, Book 8. As of August 2026, that full-unwrapping study is available as a research preprint rather than a peer-reviewed journal publication, so its status should be described accordingly. Nevertheless, it illustrates how an archaeological object that was physically unreadable can increasingly be treated as a three-dimensional dataset before a papyrologist attempts a transcription.

Preservation, in other words, no longer necessarily requires physical access.

What survives is not what antiquity valued most

These cases complicate a familiar question: why did some ancient books survive while others disappeared?

For most Greek and Latin literature, survival depended upon repeated copying. A work remained alive because generations of readers, scribes, schools, libraries and eventually medieval manuscript communities considered it worth reproducing. Failure to copy could be fatal. Once the final usable manuscript disappeared, the work effectively vanished.

Archaeological manuscripts follow another history.

A papyrus at Oxyrhynchus could survive because it was thrown away. A Ptolemaic document could survive because an embalmer needed cheap material for cartonnage. The papers of a village official could reach a modern library because priests stuffed discarded documents into crocodile mummies. The Derveni roll survived a northern Greek climate because it burned. The Herculaneum books survived because a volcanic catastrophe carbonised and buried them.

This produces what might be called a taphonomy of literature. Archaeologists are accustomed to asking why some bones survive while others dissolve, why metalwork disappears through recycling, or why charred seeds are overrepresented in archaeobotanical assemblages. Written culture requires the same caution. The surviving corpus is not a neutral sample of what ancient people wrote, owned or considered important.

Even within Oxyrhynchus, the imbalance is instructive. Documentary texts vastly outnumber literary ones. That does not mean inhabitants spent nearly all their time reading tax assessments rather than poetry. Administrative documents were produced and discarded according to different rhythms from literary books. A school exercise, personal letter, damaged roll and municipal register each had its own path into the archaeological record.

The reverse distortion occurs in the medieval transmission of classics. Works copied for centuries may survive in comparatively handsome manuscripts, while once-popular books that fell outside later educational or religious priorities disappeared almost entirely. Papyrus finds can therefore expose a literary culture that manuscript tradition had filtered out.

For the historian of literature, an Egyptian rubbish mound and a medieval library answer different questions.

What does the mummy actually tell us?

The Oxyrhynchus Iliad fragment is therefore more useful if treated as an archaeological problem than as a literary curiosity.

We can say with confidence that a Greek literary papyrus containing part of the Catalogue of Ships was present in a Roman-period funerary context and had been incorporated into the treatment of a mummy. We can also place that observation within a local pattern: the current mission had previously recovered Greek ritual or magical papyri in comparable positions. And the wider papyrological record demonstrates that Homer circulated extensively in Roman Egypt.

What we cannot yet establish is why this particular sheet was chosen.

If further study shows careful folding, purposeful positioning or other associations that distinguish the Homeric papyrus from ordinary packing material, a ritual interpretation may become stronger. If the fragment proves to have been damaged, obsolete or bundled with miscellaneous waste, recycling becomes more plausible. Palaeography, physical reconstruction, fibre orientation, traces of folding, associated materials and the detailed stratigraphy of the embalming package may all matter more to that question than the literary prestige of Homer.

There is another uncertainty. Even if the papyrus was deliberately selected because of its text, we would still need to establish what quality of the text mattered. Was it Homer as an authoritative poet? The military catalogue itself? Greek writing more generally? An object associated with education? A possession of the deceased? These interpretations are not interchangeable.

That is why the most important aspect of the discovery may not be that part of a famous book turned up inside a mummy. It is that the papyrus caught a written object in transition between two lives: first as a manuscript carrying Homeric verse, then as material incorporated into the treatment of the dead.

Much of the lost literature of antiquity reaches us at precisely such moments, after a book or document has ceased to perform the function for which it was originally made.

For archaeologists, this changes the question. Instead of asking only which texts survived, we must ask what happened to an ancient text after people stopped reading it.

In many cases, that is the reason we can read it now.

Selected Sources and Further Reading

1. University of Barcelona. 2026. “The UB's archaeological mission in Oxyrhynchus has found Homer's ‘Iliad’ inside a Roman-era mummy.” 20 April 2026. Primary institutional announcement of the Tomb 65 discovery.
2. University of Oxford, Oxyrhynchus Papyri Project. “Introduction”; “Waste Paper City”; “Scribes and Scholars.” Background on the archaeological recovery, documentary corpus and literary papyri of Oxyrhynchus.
3. UC Berkeley, Center for the Tebtunis Papyri. “The Tebtunis Papyri Collection.” Documentation of papyri recovered from human and crocodile mummy cartonnage, including the Menches archive.
4. Gibson, Adam, Kathryn E. Piquette, Uwe Bergmann et al. 2018. “An assessment of multimodal imaging of subsurface text in mummy cartonnage using surrogate papyrus phantoms.” Heritage Science 6:7. DOI: 10.1186/s40494-018-0175-4.
5. Kouremenos, Theokritos, George M. Parássoglou and Kyriakos Tsantsanoglou. 2006. The Derveni Papyrus: Edited with Introduction and Commentary. Florence: Casa Editrice Leo S. Olschki. ISBN 978-88-222-5567-9.
6. UNESCO Memory of the World. “The Derveni Papyrus: The oldest ‘book’ of Europe.” Nomination documentation and archaeological background.
7. Mocella, Vito, Emmanuel Brun, Claudio Ferrero and Daniel Delattre. 2015. “Revealing letters in rolled Herculaneum papyri by X-ray phase-contrast imaging.” Nature Communications 6:5895. DOI: 10.1038/ncomms6895.
8. Bukreeva, Inna et al. 2016. “Virtual unrolling and deciphering of Herculaneum papyri by X-ray phase-contrast tomography.” Scientific Reports 6:27227. DOI: 10.1038/srep27227.
9. Angelotti, Giorgio, Federica Nicolardi, Paul Henderson and W. Brent Seales. 2026. “Ink detection from surface topography of the Herculaneum papyri.” Scientific Reports. DOI: 10.1038/s41598-026-58467-1.
10. Angelotti, Giorgio et al. 2026. “Complete virtual unwrapping and reading of a rolled Herculaneum papyrus.” arXiv preprint 2606.29085. The paper should currently be cited as a preprint rather than as peer-reviewed research.

Suggested keywords

Oxyrhynchus; Homer; Iliad; ancient papyri; mummy cartonnage; Roman Egypt; Tebtunis; Derveni Papyrus; Herculaneum papyri; textual archaeology; manuscript preservation; archaeological taphonomy

Editorial summary

The 2026 Oxyrhynchus discovery is timely not simply because Homer has been found in an unexpected funerary setting, but because new archaeological and imaging research is revealing how much of ancient literature survives through processes—discard, recycling, burial and carbonisation—that were never intended to preserve books at all.