Jubilees Palimpsest

The Jubilees Palimpsest Project

Pioneering the Recovery of Illegible Text from Ancient Manuscripts Through New Tools in Digital Archaeology

Quick Start

See images of manuscripts

Explore other ways to find a particular manuscript, page, or annotation

Learn more about Latin Moses (Jubilees and the Testament of Moses)

Learn more about technologies for capture, processing, and publication of advanced images of manuscripts


Overview of Project Contributions

Overview of Project Goals

Resources and Links


Older news is available in the News Archive (LINK).

Open Image Repository

Although the Jubilees Palimpsest provided the original impetus for the Jubilees Palimpsest Project, it subsequently became clear that the tools of digital archaeology could have a far greater impact if applied systematically to the tens of thousands of unreadable manuscripts already in libraries and museums. The following palimpsests at the Biblioteca Ambrosiana, particularly those acquired from the Bobbio, have been identified as the top priority for advanced imaging.

Early Judaism and Christianity

  1. Latin Jubilees (80 pages, LDAB 7663, Mirador), captured 2017, basic images published, advanced images in progress.
  2. Testament of Moses (16 pages, LDAB 7663, Mirador), captured 2017, basic images published, advanced images in progress.
  3. Arian Commentary on Luke (48 pages, LDAB 7662, Mirador), captured 2017, basic images published, advanced images in progress.
  4. Ascension of Isaiah (58 pages, LDAB 7375)
  5. Gothic Commentary on the Gospel of John / Skeireins (16 pages, LDAB 10984)
  6. Unidentified Greek Commentary on the Gospel of Luke (210 pages, LDAB 128487, Mirador), five sample pages captured and published 2017
  7. Juvencus, Evangeliorum Libri Quattuor (148 pages, E 57 Sup)

Biblical Literature

  1. Origen, Hexapla (112 pages, O 39 Sup., Mirador), five sample folios (ten pages) captured, processed, and published 2017
  2. Synoptic Gospels Fragments Palimpsest (32 pages, Q 6 Sup.)
  3. Ulfila, Gothic Translation of the Book of Nehemiah (6 pages, LDAB 108085)
  4. Ulfila, Gothic Translation of the Gospel of Matthew (4 pages, J 61 Sup.)
  5. Ulfila, Gothic Translation of the Letters of Paul including a Gothic Calendar (356 [212 + 144] pages, LDAB 10684 and LDAB 10685, Mirador), twenty sample pages captured, processed, and published 2017

Greek and Roman Literature and Reception

  1. Bobbio Commentary on Uncertain Work by Cicero (210 pages, LDAB 7372)
  2. Unidentified Palimpsest in Seventh Century Latin Script (6 pages, LDAB 7688)
  3. Fronto, Epistles (34 pages, LDAB 7371)
  4. Symmachus, Orations (62 pages, LDAB 7373)
  5. Pliny the Younger, Panegyric (0 pages [some portion of the 62 counted above], LDAB 7373)
  6. Seneca, Medea and Oedipus (10 pages, LDAB 7695)
  7. Cicero, Fragments (12 pages, LDAB 7713)
  8. Plautus, Fabulae or excerpt Vidularia (464 pages, LDAB 7693)

We also believe that some items in the collection that are not technically palimpsests could benefit from advanced imaging. Furthermore, we suspect there may be more palimpsests in the collection than are currently known. Catalogers may not have noted erased text if it could not be recognized or identified. In some cases, digital archaeology can recover text from manuscripts in which the human eye cannot even tell that the erased layer exists.

Standards-Compliant Viewers and Annotation Workspaces

The images captured and processed by the project are publically available (see section on Permissions below) and comply with open standards for viewing and annotation. There are two basic types of images (with plans for greater convergence on the horizon). The first type consists of static images in a repository following the International Image Interoperability Framework (IIIF) standards (Image and Presentation APIs, LINK). The second type consists of interactive relightable texture images (RTI). Additionally, transcription and translation information is provided in XML as it becomes available (TEI compliance pending). For advanced image processing scientists, the raw data archive is available at palimpsest.stmarytx.edu slash AmbrosianaArchive, but that is not recommended for the vast majority of users.

Mirador Viewer now includes a side panel that allows users to switch and and fade between “layers,” namely the enhanced color from spectral processing and raking light angles. For pages from Jubilees, the index tab shows the verse numbers preserved on a page, along with the reconstructed page number of the original fifth-century codex, and the page numbers written on the palimpsest codex in modern times. The annotations tab shows the resources available, including WebRTI images, transcriptions, translations, and line-by-line transcriptions over the images (in progress). The annotations toggle shows the annotations entered so far, and allows the user to add additional annotations.
Mirador Screenshot

The previous viewer, IIIF Navigator, lacks the ability to add annotations but remains available.

Feature lists are constantly changing. Please send updates and corrections to the project director, Todd Hanneken (LINK). For more information see the page dedicated to standards, viewers, and prototypes (LINK).

Reconstruction of the Jubilees Palimpsest

conventional image of page 111 of the Jubilees Palimpsest

It is possible to reconstruct the fifth-century codex containing Latin Jubilees and the Testament of Moses even though more than half of it is now lost (LINK). This is possible by extrapolating from quire numbers, the average number of words per page, and comparison with the Ethiopic recension of Jubilees. For each page we know the corresponding lines in Ethiopic Jubilees. For each preserved page we now have advanced spectral images, WebRTI (relightable texture) images, scanned microfilm images (useless for reading, but useful for conservation comparison), the transcription offered by Ceriani (1861), and the translation offered by VanderKam (1989). We also produced an English translation of Ceriani’s introduction in Latin describing the condition of the manuscript and Gryson’s description in French of the codicology of the Arian Commentary on Luke (see About the Jubilees Palimpsest LINK).

This information was used to compose the entry on Latin Jubilees for the Brill Textual History of the Bible. The pre-publisher copy is available here (LINK); the official publication should be consulted for citation.

IIIF Presentation manifests have been created to reconstruct the two older manuscripts erased and resorted to create the eighth century palimpsest, the preserved portion of which is known as Ambrosiana C73 inf. The two older manuscripts are Latin Moses (including Jubilees and the Testament/Assumption of Moses) and the Latin (Arian) Commentary on Luke. These manifests can be browsed in the Mirador Viewer. A IIIF Presentation manifest for eighth century palimpsest, containing portions of Eugippius’ anthology of Augustine, will be created after we identify the best way to represent the need for 180° rotation for some of the pages captured with orientation appropriate to the erased text.

A paleography chart for Latin Moses is also available (LINK).

Spectral RTI Technology

Spectral RTI Pseudocolor with Raking Light

Spectral RTI combines the advantages of spectral imaging (spatial resolution, color spectrum range and resolution, processed enhancements) with the advantages of Reflectance Transformation Imaging (pixel-level texture mapping, interactivity, enhancements). Spectral is helpful for manuscripts because it recovers traces of erased ink indistinguishable to the human eye. Texture mapping is helpful for manuscripts because it can show the outline of an erased letter in the corrosion on the surface of the parchment, as well as other features of accretion (wax, dirt, glue, other deposits) and depression (scores, punctures, dry-point notations).

In 2013–2014 the National Endowment for the Humanities (Digital Humanities Startup Grant) funded a collaboration of the Jubilees Palimpsest Project with the Early Manuscripts Electronic Library and the University of Southern California to test and develop methods to integrate spectral and RTI. The most efficient method uses the standard equipment for spectral and RTI capture and relies on processing to combine the chrominance features of spectral imaging with the luminance features of RTI.

That phase produced:

  1. A collection of RTI images viewable in WebRTI. The images use the most efficient method of Spectral RTI for objects with varying depths of texture. (LINK). Those images are also available as a IIIF Presentation manifest viewable in Mirador (LINK).
  2. A twenty-minute video demonstrating the capabilities of Spectral RTI (LINK).
  3. A summary table demonstrating the most efficient method compared with other experimental methods and each technology individually (LINK).
  4. For moderately advanced audiences, a detailed white paper describing the experimental methods, our findings, and information to replicate our procedure (LINK).
  5. For advanced audiences, the original ImageJ Macro used to process the experimental images (LINK), followed by a GitHub repository with a more user-friendly version of the software necessary for processing Spectral RTI (LINK).
  6. For advanced audiences, the complete data set of captured and processed images (LINK).

More information about the Integrating Spectral RTI phase of the Jubilees Palimpsest Project is available on the phase webpage (LINK). See also, “New Technology for Imaging Unreadable Manuscripts and Other Artifacts: Integrated Spectral Reflectance Transformation Imaging (Spectral RTI)” (LINK). For a shorter overview see the entry on “Spectral RTI” planned for the Brill Textual History of the Bible (LINK).

Extended Spectrum Processing

Extended Spectrum is a new processing technique for spectral imaging captures. It utilizes all the captures of spectral imaging (including ultraviolet and infrared) and principal component analysis within the three categories of color natural to the human eye. The result is more natural than pseudocolor but simulates what we would see if human color perception had a wider range and higher resolution. The technique divides the spectral captures into three categories: the shortest wavelengths (blueish), the middle wavelenghts (greenish), and the longest wavelengths (reddish). Principal Component Analysis finds the greatest contrasts in each of the three categories, which are then mapped to the three channels of an RGB image.

USC Mummy Mask in Accurate Color and Extended Spectrum

Roman-Period Egyptian Mummy Mask from the USC Archaeological Museum in Accurate Color (left) and Extended Spectrum (right).

Easy Access to Spectral RTI

The 2013–2014 Integrating Spectral RTI phase demonstrated the feasibility of Spectral RTI using open-source software. The ongoing effort is to put the technology in the hands of scholars, museums, and libraries. A three-year project (2016–2019) supported by the National Endowment for the Humanities is making significant progress in this direction.

ImageJ Spectral RTI Toolkit, Select Tasks

The Spectral RTI Toolkit oversees all the processing with a graphical user interface (without requiring use of command-line arguments or editing text files). Specifically, the Spectral RTI Toolkit for ImageJ processes captured images into RTI images (PTM, HSH, or WebRTI) with color enhancements for Accurate Color, Extended Spectrum, and PCA Pseudocolor. The Spectral RTI Toolkit is available on GitHub, initially as an ImageJ macro and eventually as a plugin.

Documentation for the complete process of Spectral RTI, including hardware options, capture procedures, processing, and publication, is available online (LINK). The documentation is under active development, and will be updated to when the plugin (not macro) version of the Toolkit becomes available.

The NEH grant will also provide opportunities for on-site training during 2018–2019. Contact the project director, Todd Hanneken (thanneken@stmarytx.edu) or follow the project on Twitter (@thanneken) to follow the latest development as easy access to Spectral RTI unfolds.

Vision for a Permanent Center for Digital Archaeology

We believe that a permanent center for digital archaeology will create significant added value compared to a series of isolated projects.

  1. First, stable commitments will reduce cost. The costs of digital archaeology are high, but not intrinsically more so than other international and high-tech programs. Traditional archaeology also required international travel, specialized tools, and scholarly expertise.
  2. Second, not only will a stable project reduce costs, it will produce a steady stream of discovery, adding to the “benefit” side of the cost/benefit ratio. All estimates indicate that we will not run out of palimpsests to image in the conceivable future.
  3. Third, a permanent center will create more opportunities to improve the tools of digital archaeology and attempt new techniques across many years. Challenges faced one year can create the agenda for the next year’s excavation.
  4. Fourth, a permanent center will be necessary for the vision described below for mentoring the next generation of digital archaeologists.

We are developing a multi-institution collaboration to achieve these goals. This phase has not yet been funded.

Vision for Advancing and Promoting the Tools of Discovery

In addition to the Spectral RTI Toolkit described above (LINK), the project is working steadily to improve the techniques of Spectral RTI capture, processing, and dissemination.

For more on the state of the art and future horizons see Todd R. Hanneken, “New Technology for Imaging Unreadable Manuscripts and Other Artifacts: Integrated Spectral Reflectance Transformation Imaging (Spectral RTI).” In Ancient Worlds in a Digital Culture. Edited by Claire Clivaz and David Hamidovic. Digital Biblical Studies 1. Leiden: Brill. The pre-publisher copy is available here (LINK); the published version should be consulted for citation.

Vision for Mentoring the Next Generation

All phases of the Jubilees Palimpsest Project include paid student researchers. The vision for the next level of educating the next generation of digital humanists encompasses three categories.

  1. Student researchers will have the option of receiving academic credit for internship by learning the academic context of their practical work.
  2. Student research will be joined to existing programs such as the Summer Undergraduate Research Fellowship and McNair Scholars program to increase opportunity for faculty mentoring.
  3. Regular courses during the academic year will cover related topics in the digital humanities, archaeology of the book, digital archaeology, ancient and computer languages, and so forth. These courses could lead to an undergraduate minor or graduate certificate in Digital Humanities. These courses could be shared between collaborating institutions using synchronous and hybrid distance learning technologies.

Vision for Scholarly Forums

The image repository is only the beginning of the process of discovery and interpretation in the humanities. We envision four channels in which the global community of scholars will be able to contribute to interpreting the images.

  1. First, Open Annotation will allow scholars anywhere on the web to contribute annotations to the repository. The annotations could be transcriptions, translations, links to scholarly resources, description of scribal features such as lines and columns per page, hair and flesh side, quire type, rubrication, marginalia, illumination, ligatures and paleographic features, etc. The annotations can be linked from elsewhere on the Internet or contributed to our repository.
  2. Second, we imagine the fastest and most democratic pipeline to sharing discovery will be on the Weblog of Digital Archaeology. Contributions to the weblog will be mildly curated, with an emphasis on quick publication and minimal vetting.
  3. Third, we imagine creating an online journal with high standards for peer review and vetting, The Journal of Digital Archaeology.
  4. Fourth, we imagine a digital venue for monographs, especially scholarly editions (see next).

Vision for Complete Scholarly Editions

We anticipate that the crowd-sourced annotations contributed to the repository will assist but not replace a principal scholarly edition overseen by a single editor. We envision a new generation of scholarly edition utilizing new standards (see MLA Guidelines at mla.org). In particular, we envision major improvements compared to past scholarly editions projects, such as the Discoveries in the Judaean Desert series.

About Digital Archaeology

Digital archaeology is the study of ancient things using primarily digital tools. Even as conventional digging-based archaeology increasingly utilizes digital tools such as GPS, satellite imagery, and laser scanners, we use the term for approaches that replace digging in dirt entirely. In that way the “digits” involved are binary code, not ten fingers with dirt under the fingernails. Other than the relative lack of dirt, digital archaeology is one with archaeology in uncovering and interpreting evidence of the past. Formerly, evidence was inaccessible because it was buried; for palimpsests and other illegible artifacts the inaccessibility is due to the limitations of the human eye. Digital technology overcomes these limitations.

Another intrinsic feature of digital information is that it can be duplicated and transmitted losslessly. Whereas the examination of artifacts in conventional archaeology is necessarily entrusted to a small number of archaeologists with first-hand access, digital archaeology is fertile ground for collaboration and crowd-sourcing. The project is committed to maximum open access (Creative Commons licensing) and interoperability (IIIF, W3C Open Annotation, etc.).

About the Jubilees Palimpsest

The Jubilees Palimpsest is named after the oldest of three texts contained in the erased layer of what now stands as a 144-page codex, unbound for conservation. In the fifth century a Latin translation of Jubilees (originally composed in Hebrew in the 150s BCE) was copied with another work associated with Moses, the Testament (or Assumption) of Moses. In the eighth century, presumably at the Bobbio abbey, that codex was unbound, erased, and recombined with another erased manuscript of the Arian Commentary on Luke. The erased parchment was then used to copy Eugippius’s anthology of the works of Augustine. In the seventeenth century that palimpsest was brought to the Biblioteca Ambrosiana in Milan, where it is preserved today. In 1861 A. M. Ceriani published the editio princeps. Additional scholarship on the manuscript was offered by H. Rönsch (1874). After the nineteenth century there were no successful attempts to recover additional text of Jubilees or the Testament of Moses.

See the full page About the Jubilees Palimpsest (LINK) for bibliography and links. Many of the definitive works on the palimpsest are out of copyright and can be downloaded in their entireties from this site or from Google Books. The full page includes the following sections:

  1. Descriptions of the manuscript – including an English translation of R. Gryson’s study of the Arian Commentary on Luke and an English translation of A. M. Ceriani’s introduction.
  2. Images of the manuscript – The most important images are the Spectral RTI images from 2017. Digital scans of the microfilm and a few additional images may be useful for comparison.
  3. Bibliography on Bobbio Manuscripts, Latin Palimpsests, Eugippius, and the Arian Commentary on Luke.
  4. Links pertaining to the Biblioteca Ambrosiana

About Web Dissemination and Accessibility

The Jubilees Palimpsest Project is committed not only to open access but compliance with standards for interoperability and discovery. This frees the data from the silo of a single website and allows it to be aggregated, viewed, studied, annotated in ways far beyond what can be presently imagined. A separate page on Standards and Tools (LINK) provides links and examples. That page can be outlined as follows:

  1. The International Interoperability Framework (IIIF) provides crucial standards for making the images available and organizing the image data and metadata for efficient viewing.
  2. Mirador is being actively developed for easy and powerful study of IIIF presentations. A number of alternatives are also worth considering.
  3. Open Annotation allows anyone to link annotations to our repository. The annotations can be stored anywhere on the Internet or contributed to our repository.
  4. WebRTI allows any modern web browser (even a smart phone) to provide the interactive visualizations of texture without downloading any plugins, installing software, or opening ports in firewalls. Specular Enhancement is now available for WebRTI.
  5. Aggregators and portals facilitate the discovery of digital resources.

About the People Involved in the Jubilees Palimpsest Project


The Jubilees Palimpsest Project was founded and is directed by Todd Hanneken at St. Mary’s University in San Antonio. See the complete list of people involved in various phases of the project (LINK).

Funding Partners

NEH Logo

The National Endowment for the Humanities: Celebrating 50 Years of Excellence

StMU Logo

St. Mary’s University, especially the Edward and Linda Speed Fund



Image of artifacts from the Biblioteca Ambrosiana can be used and reused under the CC BY-NC-SA license. That means the images can be used as long the sources are attributed, derivaties are shared with the same degree of openness, and no commercial profit is derived from the images. Commerical use of images of artifacts owned by the Biblioteca Ambrosiana must be licensed from the Ambrosiana (not the Jubilees Palimpsest Project). The contact for licensing permission from the Biblioteca Ambrosiana is direzione.biblioteca@ambrosiana.it.


Everything created solely by the Jubilees Palimpsest Project is licensed under the Creative Commons CC BY-SA license. This includes metadata, documentation, and software created by the project. Images of artifacts from U.S.C. and U.C.L.A. are also available under this license.

Attribution should include at minimum the Jubilees Palimpsest Project and the owner of the artifact (e.g., Biblioteca Ambrosiana, U.S.C., U.C.L.A.). If a named author is required use Todd Hanneken or contact the project director for specifics on persons who contrbuted to creating a particular image.

Some items linked from this site are governed by separate licenses. Examples include the digital versions of public domain books created by Google Books.