Recovering an Erased Gospel

Two hundred years ago, a nobleman on the Greek island of Zakynthos presented a visiting British soldier with a handwritten copy of the readings from the Greek gospels used in church services. On his return to London, General Colin Macaulay gave this manuscript, Codex Zacynthius, to the British and Foreign Bible Society.

Although the text of this lectionary appeared to have been copied in the thirteenth century, scholars soon realized that the manuscript was a palimpsest: the gospel extracts had been written on the pages of a much older document whose text had been erased in order to re-use the parchment for another book. Pages where the remains of the earlier text could be made out with the naked eye enabled it to be identified as a copy of the Gospel according to Luke. Based on the style of the handwriting, it was estimated that it had originally been copied between the sixth and eighth century, at least five hundred years before its rewriting as a lectionary.

A source for early commentary

The text of Luke, however, only occupied the middle part of each page. In wide margins, another text had been added by the copyist in a smaller version of the same script. This showed that the original manuscript was a type of commentary known as a catena, in which extracts from early Christian authors had been joined together to form a chain of comments explaining each passage in the gospel text.

In many cases, the original writings from which these extracts were taken have been lost, and these commentaries are the only surviving source. As the only known manuscript in which both biblical text and commentary were written in majuscule script, Codex Zacynthius appeared to be the earliest surviving example of a New Testament catena by at least a century.

Recovering the text

Around the end of the nineteenth century, scholars began to experiment with using chemicals to make the underwriting stand out on palimpsest manuscripts. Although initial results were promising, subsequent deterioration made these pages even less legible than before. Fortunately, this was not attempted on Codex Zacynthius. Instead, after the manuscript was acquired by Cambridge University Library in 2014 following a public campaign which raised £1.1 million, it was examined through a non-invasive process known as multispectral imaging. Funded by the UK Arts and Humanities Research Council, the Codex Zacynthius Project engaged a specialist team which took fifty-one high resolution images of each page using different wavelengths of light, from infrared to ultraviolet.

Using advanced processing techniques, different sets of images were combined in order to produce a single composite photograph on which the undertext was as visible as possible. The final result was remarkable: on the majority of pages, the erased undertext could be clearly made out thanks to an artificial coloring of this type of ink in a dark blue color, while text written in red ink appeared in a different hue. The black writing of the overtext was transformed into a light cyan color in order to enable readers to make sense of the obscured portions of the letters below.

A digital edition

The new images were then used by scholars at the Institute for Textual Scholarship and Electronic Editing at the University of Birmingham to make a full transcription of every word of the manuscript. The only previous attempt to do this had been in 1950, when the American scholar J. Harold Greenlee spent a year working with the manuscript on a windowsill in Oxford’s Bodleian Library, trying to read the manuscript in direct sunlight.

The multispectral images made it possible for the Birmingham team to improve significantly on Greenlee’s unpublished results, including reading large passages which he had left blank. Once the transcription had been completed, the whole text was translated into English, preserving the same layout as the catena manuscript, to make this early form of commentary available to readers without Greek.

The multispectral images, the transcription and the translation have all been made openly available on the Cambridge Digital Library. This digital edition also includes a complete set of photographs of the lectionary overtext, along with a transcription of these gospel passages, showing the current appearance of the manuscript and the evidence it provides for Byzantine liturgical practices.

The struggles of a twelfth-century scribe

While examining the lectionary, members of the project team noticed an unusual feature. At the foot of many of the pages, there are a series of notes written by the copyist which have nothing to do with the biblical text. Several of these are appeals to God and to later users, such as “God be merciful to me, the sinner Neilos” or “Priests, remember Neilos in the all-night vigil.”

Notes like this found in other manuscripts featuring the same handwriting have enabled us to identify the scribe Neilos as a monk active on the island of Rhodes between 1170 and 1181, providing for the first time a precise date and location where the pages of Codex Zacynthius were reused. In fact, the copyist might even be the Neilos who became Abbot of the Monastery of St John in 1174.

Some of the notes, however, refer to the problems faced when copying the manuscript. After a page with many crossed-out words and erasures, the scribe has written at the bottom: “Very drowsy and foolish.” Elsewhere, he observes that “The one who writes tends towards errors,” and that a particular mistake is “The error of Theodore the squinter.” Most striking of all, at the foot of one page he exclaims “I am very tired, with a heavy head, and what I write I do not know!” These marginal asides offer an unexpectedly vivid portrait of the scribe struggling to copy a lengthy liturgical manuscript some 850 years ago.

New discoveries in the ancient text

The multispectral images have led to new discoveries in the original text of the manuscript. Codex Zacynthius has already been recognised as an important witness to the text of the Gospel of Luke. It preserves a series of chapter divisions which are only otherwise attested in the famous fourth-century Greek Bible known as Codex Vaticanus. At the beginning of the book, Codex Zacynthius features the earliest example of a table of capitula parallela, a means of cross-referencing the contents of the four canonical gospels. The project has identified three previously unknown places in which the manuscript attests the reading of the earliest form of the text of Luke, demonstrating the value of its biblical text.

The transcription of the catena has made this commentary available for the first time. Of the 343 extracts, no fewer than 300 preserve passages from early Christian writings which are not found in Greek outside this tradition. Half of these come from the Commentary on Luke by the fifth-century Patriarch Cyril of Alexandria. Almost fifty come from Titus, bishop of the town of Bostra in southern Syria which is now a World Heritage Site. His sermons on Luke, originally preached in the middle of the fourth century, are only preserved in catenae.

Most remarkable are passages from the Christian writer Severus of Antioch. In the year 536, just before his death, Severus was excommunicated and his books were banned by the Emperor Justinian: none survive in Greek. Codex Zacynthius, however, preserves thirty-eight extracts attributed to Severus, several of them quite extensive. Many of them give details of the sermon or the letter from which they were taken, and even describe the author as “Saint Severus,” suggesting that the compiler of this commentary did not subscribe to the condemnation of Severus. Nevertheless, in a later manuscript based on this catena, the extracts from Severus were omitted or ascribed to a different author. Codex Zacynthius therefore provides the only Greek text known to survive of certain writings by Severus. This discovery will enable scholars to look for other passages by Severus in catenae and examine portions of his works in their original form.

A study of the handwriting as revealed on the new images suggests that Codex Zacynthius was originally copied during the eighth century, confirming it as the oldest surviving catena manuscript. However, the project has also identified features which indicate that this is not the first instance of this compilation, but a copy of an even earlier catena manuscript. The research team published a volume of studies giving a full account of the manuscript and their findings. In addition, a printed edition of the catena with facing English translation has also been made available in open access. The University of Birmingham is now home to a major European-funded project which will undertake the first systematic examination of New Testament catena manuscripts and shed new light on the significance of this tradition of commentary.