‘The Graz Mummy Book’: The Oldest Known Codex Fragment from 260 BC Discovered at Graz University Library, Austria

ABSTRACT

The oldest known fragment of a codex was discovered at Graz University Library and made public in June 2023. The so-called ‘Mummy Book’ shows features – a central fold, text layout, ink transfer and holes – that distinctly identify it as a bifolio from a codex-like structure. These characteristics are described here and their significance is summarized from our longer research article. By examining these features, we were able to reconstruct the history and use of the fragment. The results are significant because they identify the fragment as the oldest known bifolio, predating similar examples by 400 years. Although no comparable material from this period is known, the Mummy Book is evidence for the existence of pre-Christian codices, substantially changing the history of the book. We hope that collections will be systematically examined under this lens to unearth further examples providing a clearer chronology of book history before Christ.

ZUSAMMENFASSUNG

Das älteste bislang bekannte Fragment eines Codex wurde in der Universitätsbibliothek Graz entdeckt und im Juni 2023 der Öffentlichkeit zugänglich gemacht. Das sogenannte “Mumienbuch” weist Merkmale - eine zentrale Faltung, Textlayout, Tintenübertragung und (Heft-)Löcher - auf, die es eindeutig als Bifolio aus einer Codex-ähnlichen Struktur identifizieren. Diese Merkmale werden hier beschrieben und ihre Bedeutung wird in unserem längeren Forschungsartikel zusammengefasst. Durch die Untersuchung dieser Merkmale konnten wir die Geschichte und die Verwendung des Fragments rekonstruieren. Die Ergebnisse sind bedeutsam, denn sie weisen das Fragment als das älteste bekannte Bifolio aus, das ähnliche Beispiele um 400 Jahre übertrifft. Obwohl kein vergleichbares Material aus dieser Zeit bekannt ist, ist das Mumienbuch ein Beweis für die Existenz vorchristlicher Kodizes, was die Geschichte des Buches wesentlich verändert. Wir hoffen, dass nun unter diesem Aspekt (Papyrus-)Sammlungen systematisch untersucht werden, damit weitere Beispiele gefunden werden können, die eine klarere Chronologie der Buchgeschichte vor Christus möglich machen.

KEYWORDS:

• Book history
• codicology
• papyrus
• early codex
• mummy

SCHLÜSSELWÖRTER:

• Papyrus
• Früher Codex
• Mumie

The Special Collections of Graz University Library houses 52 Egyptian papyrus fragments from El Hibeh and Oxyrhynchus. Discovered during the excavations carried out by the British egyptologists Bernard P. Grenfell and Arthur S. Hunt, the fragments were sent to the university in return for financial support between 1904 and 1909 and have since been stored between glass. Over a century later, in 2023, the collection was assessed in the conservation department with the aim of improving the storage conditions of the fragments. In the context of these examinations, head conservator Theresa Zammit Lupi discovered evidence of fragment UBG Ms I 1946 – now nicknamed ‘The Mummy Book’ – having once been a bifolio of a book in codex format (Figures 1 and 2). Dated to 260 BC, this fragment predates the earliest known book fragments with similar characteristics by at least 400 years.

Figure 1. Recto of the fragment showing a central fold and Greek text within clearly defined margins.

Figure 2. Verso of the fragment showing plaster and pigments.

As it is seen flat and unfolded between two glass plates, the recto of the fragment contains Greek text in black ink originating from its first use as a document recording a banker’s account of taxes on beer and oil. The verso is decorated with painting on a plaster base, evidencing the fragment’s re-use as part of mummy cartonnage. The substrate is a papyrus sheet that seems to have been cut from a blank roll, as it contains a kollesis (overlap) of approximately 15 mm. The left, right, and lower edges of the fragment appear to correspond to the original edges of the bifolio in its first use. The upper edge is lost and appears to have been broken cleanly along a horizontal fold. The fragment has a vertical central fold. There is a number of holes, some of which contain thread remnants and appear to have been used in a form of stitching.

That a papyrus fragment from 260 BC might be a bifolio of a codex-like notebook initially seemed unthinkable, the implication being that the codex format did not originate with early Christianity as thought until now. To us, who focus on the material and structural aspects of books on a daily basis, it seemed so clear that the fragment is a bifolio that we initially questioned the dating of the fragment to 260 BC. After the discovery of the Mummy Book was announced in a press release, however, it became apparent that papyrologists and scholars of related disciplines were sceptical not of the dating of the fragment – which they considered reliable – but of our interpretation of the fragment as a bifolio, i.e., that the sheet was intentionally folded in half to create a material structure to contain subsequently written text. Encouraged by questions and comments and feeling the need to clarify certain aspects, we deepened our research and examinations of the fragment. This resulted in new conclusions, a reconstruction of the fragment’s history, and the publication of the document Observations on ‘The Graz Mummy Book, Graz University Library, Austria’.¹ What follows is a summarizing description of the major material characteristics which, in their particular combination, indicate to us that the document was designed as a bifolio and was once part of a codex-like notebook structure.

Central fold

The most salient feature is a perfectly vertical fold through the full height in the very centre of the fragment. Either side of it are soft kinks which, seen under raking light, resemble in shape a flying bird or a gentle ‘m’. There are also slight cracks along both sides of the central fold and disturbances in the plaster layer on the verso in these areas. The area around the central fold has a clean appearance compared to the rest of the papyrus. These features indicate that the opening of the bifolio was restricted in a way that the papyrus had to flex when the text was consulted, as would be the case in a stab-sewn book structure. A stab sewing would also have protected the fold from staining during the fragment’s first use as a document.

Layout

The writing is laid out within two clearly defined text areas with margins that are bordered by the edges of the bifolio on both sides and the central fold between them. On a scroll, the width of the columns and the distance between them could be determined somewhat arbitrarily by the scribe. In contrast, the text area on a bifolio is limited on all four sides by the sheet edges and central fold, which also act as a reference for creating even margins. In the case of our fragment, the text layout – the positioning of the text on the page, the alignment of the lines, and the proportions of the existing margins – is comparable to later manuscripts and even printed books in codex format. The placement of the text areas in relation to the fold appears so intentional that it seems that the papyrus was first folded to create a bifolio, and then written upon within the framework created by the fold and edges.

Ink transfer

During our examination of the fragment, we used multispectral imaging to increase the legibility of the text on the right side of the recto (Figure 3). The enhanced images also revealed multiple instances of ink transfer from the left side of the recto to the right and vice versa that are symmetrical in relation to the central fold. These are visible in the form of complete letters, duplicated in reverse, and individual spots of ink from the beginning and end of individual strokes – where the ink would have pooled – that transferred to the opposite side. As the ink has not bled, smudged, or faded in these areas, it seems unlikely that the transfer occurred due to moisture ingress at a later stage when the document was in a folded state. Rather, the appearance of the ink transfer is crisp, distinct, and unshifted, indicating that the document was closed along the pre-existing central fold while the ink was still wet and fresh from writing.

Figure 3. Enhanced MSI image with improved legibility of the text on the right side and ink transfer.

Holes

After a detailed examination of the holes found in the fragment, we divided these into three categories. The holes in categories 1 and 2 appear to have been pierced with a tool during the fragment’s first use; the remaining holes in category 3 are non-uniform and are more likely signs of damage than function. Focusing on pricked holes, we noticed that these formed sets of two or four and so were pierced simultaneously when the fragment was folded along the central and horizontal folds (see Figure 4). The holes in category 1 (X and Y) were pricked from the opposite direction to those in category 2 (A to F) and lack corresponding holes below the horizontal fold, indicating that the holes in category 1 were pierced independently of those in category 2, probably even before the horizontal fold was created. The holes in category 1 can be interpreted as evidence for a stab sewing because (a) they contain remnants of thread fibres, (b) they are positioned close to the central fold and at a distance from the lower edge where a stab sewing makes sense (a second pair of sewing holes likely existed in the upper portion of the fragment, which is now lost), and (c) the papyrus around the holes is distorted, suggesting that some force was required to pierce it, as would be the case when stabbing a needle through a stack of papyrus. Furthermore, the presence of the holes is an indication for the existence of a codex-like structure, as it suggests that the bifolio was attached to something, possibly to more bifolios like it.

Figure 4. Direction of pricking of the hole sets: X/Y (category 1) in green and A/B/C/D, E/F, G/H/I/J (category 2) in red.

While examining the fragment, our ideas of how the bifolio might have looked when it was used, developed as we discovered more details and tried to make sense of how individual features related to each other. By examining signs of damage and material evidence from the various stages of the document’s use, we were able to place individual features in a chronological order and so reconstruct the history of the fragment using our practical experience as conservators. Figure 5 illustrates how blank papyrus sheets were folded into bifolios, stacked on top of one another, and then stab-sewn together with individual loops of thread. The text was then written on the sides with horizontal fibres. Writing along the horizontal fibres of the papyrus was standard practice on scrolls, and this may have remained common in the transition period from scroll to codex. Eventually, the stab sewing was undone and the notebook dismantled into single bifolios. The preserved example was then sealed by folding over the upper and lower edge and securing these with thread tackets, using the holes in category 2, perhaps to protect sensitive information, archive the document, or send it to a relevant party. At this stage, significant damage occurred, and it was eventually applied, in a folded state, as the uppermost layer of mummy cartonnage, where it was plastered and painted.

Figure 5. Reconstruction of the stages in the ‘lives’ of The Graz Mummy Book: 1. A blank papyrus sheet (formed of two kollemata joined by a kollesis) – 2. The piece is folded to form a bifolio – 3. Multiple bifolios are stacked together – 4. The bifolios are stab sewn – 5. The notebook is written upon – 6. The binding is dismantled and the bifolios are separated – 7. The bifolio is secured with tackets – 8. Significant damage occurs to the bifolio – 9. The bifolio is reused as mummy cartonnage alongside further papyrus waste and textile – 10. The cartonnage surface is plastered – 11. The cartonnage is painted – 12. The bifolio is taken from the mummy and becomes a museum object.

While much information has been lost, the fragment’s current state as collection item allows its 2280-year-long history to be studied by examining the tiny clues that remain. This emphasis on material evidence is absolutely necessary, as there is little else to provide an insight into the early development of the codex. Beside a rumour around the ‘invention’ of the codex by Julius Caesar, the earliest textual evidence of a book in pugillaribus membraneis, interpreted as referring to the codex format we know today, stems from the Roman poet Martial around 85 AD. Even material evidence is sparse: the fragment of Rylands Library Papyrus P52, dated to 125 AD at the earliest, is so small (89 × 64 mm) that its description as a codex can only be based on textual distribution, not indicative physical features. The earliest codex fragments with preserved features that are comparable to those of the Mummy Book – a central fold, stitching holes, and codex-like text layout – are Demosthenes, De Falsa Legatione, dated 150–250 AD (parchment) and CBL BP 1 (P45), dated 200–250 AD (papyrus). Interestingly, a recent binding model of the latter² uses a multi-quire stab-sewn structure that closely resembles our interpretation of the Mummy Book’s notebook structure.

Until further examples of codex-like structures from the pre-Christian era emerge, the absence of comparative material makes it difficult to position The Mummy Book in its context, and leaves open questions regarding its use and development. Nevertheless, even if considered an early and perhaps experimental precursor to a notebook, its structure fits astonishingly well between that of a writing tablet and that used for the model of P45. At the very least, the resemblance of The Mummy Book to a bifolio from a third century codex is striking enough to make further investigations into the early history of the codex worthwhile. We hope that this discovery will trigger a systematic search for fragments of codices before Christ’s era to build a clearer chronology of the history of the book. Who knows what we might find if we re-examine our collections under this new lens!

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All information about The Mummy Book is available here: https://mummybook.uni-graz.at/en/. All illustrations were carried out by Lena Krämer.

Additional information

Notes on contributors

Theresa Zammit Lupi

Theresa Zammit Lupi studied art history at the University of Malta, after which she studied book and paper restoration in Florence and London. In 2009, she received her PhD in London in the conservation of manuscripts. In 2017, she was a research fellow at Harvard University where she focused on Renaissance illuminated manuscripts. Theresa worked in Malta, Italy, the UK, Switzerland, Egypt and Ethiopia and has taught conservation and codicology in several countries. Since 2021 she has been Head of Conservation for the Special Collections at Graz University Library. She is an accredited member of the Institute of Conservation (ICON).

Lena Krämer

Lena Krämer is a conservator of books and library materials. After graduating with a bachelor's degree in history and philosophy in Ireland, she studied conservation at West Dean College in England, where she graduated with honours for her master's degree in 2021. After that she worked at Green's Books, a private conservation workshop in England. In 2022, she returned to her hometown Graz, where she has since worked as a conservator in the Special Collections Department at the University Library. Lena Krämer is the author of several publications and volunteers to assist the editors of Bookbinder, the journal of the Society of Bookbinders.

Thomas Csanády

Thomas Csanády is the administrative director of the Special Collections at Graz University Library. He holds a PhD in medieval liturgical history and is researching the liturgical origins of the diocese of Seckau, Austria. His research focuses on the history of the liturgy, the liturgy of the medieval metropolitan seat of Salzburg, medieval liturgical manuscripts of the Western tradition (12th/13th century) and their codicology. Thomas was also trained as a scientific librarian in Graz and Vienna. He is a lecturer at the Institute of Liturgy at the Faculty of Theology at the University of Graz.

Erich Renhart

Erich Renhart studied theology in Graz and did his post-doctoral research in liturgical studies with a focus on manuscripts. In 2005/06 he founded ‘Vestigia’ – the Centre for the Study of the Book. Five years later, he took over as scientific director of the Department of Special Collections at Graz University Library, and Vestigia Research Centre was linked to the Special Collections. Erich Renhart works nationally and internationally in particular on Latin, Greek, Armenian and Syriac manuscript heritage. He is currently researching a palimpsested manuscript from the 8th century and other material from the first millennium.

Notes

1 Zammit Lupi et al. (2023).

2 Pictured in: Kristine Rose-Beers (2023) and discussed in: Brent Nongbri (2022).