Final Report for Multispectral Imaging of Jefferson Manuscripts

Summary of the project to date:

Multispectral imaging has proven to be an innovative and revolutionary technology well suited to document imaging. Multispectral imaging is a digital imaging technique that scans documents at different wavelengths of light. By scanning manuscripts with specific wavelengths of light, previously illegible ink may be illuminated. This technology has been successfully applied to thousands of manuscripts, the Archimedes palimpsest being one of the most famous examples of successful multispectral imaging, and more recently, the diary of Dr. David Livingstone.

In its prototypical phases, multispectral document scanning proved to be both costly and complex, requiring specialized equipment and extensive training. However, under the direction of Drs. Dirk Obbinkand Alexander Kovalchuk at Oxford University, a new and more economical flatbed scanner was developed for multispectral imaging. After its initial development, the scanner was then commercialized and released for sale under the newly founded company, Oxford Multi Spectral Ltd.

With funds from the Jefferson Trust, the University of Virginia Library’s Digitization Services unit was able to procure a multispectral imaging scanner from Oxford Multi Spectral Ltd. The Oxford Multi Spectral scanner utilizes six distinct wavelengths of the electromagnetic spectrum, measured in nanometers (nm). When scanning a document, the multispectral scanner exposes the document to ultraviolet light at 395 nm, blue light at 450 nm, green light at 520 nm, amber light at 590 nm, red light at 650 nm, and infrared light at 850 nm, producing a total of six different images. By utilizing the entire spectrum of visible light, including infrared and ultraviolet, the scanner is oftenable to reveal details in documents that were previously imperceptible in simple white light.

After thoroughly training staff and several student employees in its proper use and operation, the multispectral scanner was put to work scanning some of the University of Virginia’s collection of Thomas Jefferson manuscripts. The Jefferson Collection proved to be a suitable candidate for multispectral imaging, as many of the documents contained within the collection possess crossed-out text, water-damaged text, ink spills, faded ink, and letterpress copies with ink that has bled substantially. With the use of multispectral imaging, more of the contents of these previously illegible documents were made legible.

Prior to the start of scanning, a student employee examined the entirety of the Jefferson Collection, analyzing 13,429 preexisting digital images. Of these 13,429 images, 468 manuscripts were chosen as candidates for scanning. After several preliminary scans, the initial list of 468 manuscripts was reexamined and reduced to 220 manuscripts, based on our expanded knowledge of the scanner’s capabilities, particularly with these kinds of materials. From these 220 manuscripts, a total of 1305 multispectral scans were completed and processed over the course of five months.

Total cost to date for the project:

A total of $24,931.66 ($4,903.83 for wages and benefits and $20,000 for equipment) has been spent leaving $68.34, which will not be spent.

Is the project completed?

Yes, we have completed the digitization of the Jefferson materials targeted for multispectral scanning.

What successes can you report coming out of the project?

After the preliminary list of 220 manuscripts had been identified, we contacted several prominent Jefferson scholars with information on the newly acquired multispectral scanner. Ms. Lisa Francavilla, the managing editor of The Papers of Thomas Jefferson: Retirement Series of Monticello’s International Center for Jefferson Studies, and Dr. Martha King, associate editor of The Papers of Thomas Jefferson of Princeton University, both replied with requests for multispectral scans.

Ms. Francavilla requested multispectral scans of several manuscripts, including Jefferson’s list of books at Monticello, Jefferson’s list of paintings and art at Monticello, and the deed for the sale of Monticello. Below is Ms. Francavilla’s analysis of the multispectral scans provided by Digitization Services:

“The multispectral manuscript imaging is remarkable for its ability, not only to illuminate documents that suffer from severe staining and water damage, but for literally bringing to light texts that had been heavily altered by Jefferson as he drafted and redrafted his writings. We are now able to take manuscripts like his lists of artworks and paintings at Monticello and his catalog of books loaned to friends – heavily marked by strikethroughs, blotches, and other emendations – and create more complete transcriptions and annotations. This will make the documents more accessible and useful when published in letterpress editions of Jefferson’s writings, which in turn allows for their inclusion in online collections like the University of Virginia Press’s Rotunda, and ultimately leads to their presentation in the National Archives’ Founders Online website.”

-Ms. Lisa A. Francavilla

Managing Editor

The Papers of Thomas Jefferson: Retirement Series

International Center for Jefferson Studies

Dr. King requested multispectral scans of several letters from the correspondence of Thomas Jefferson and Craven Peyton concerning Jefferson’s land transactions in Albemarle County and the estate of Bennett Henderson. Below is Dr. King’s analysis of the multispectral scans provided by Digitization Services:

“As it turns out, I have just been working on the Jefferson to Craven Peyton letter of 16 January 1804 this week for our Volume 42 of The Papers of Thomas Jefferson, and your very timely scan has vastly improved our reading of the letter, clarified dates and amounts mentioned, resolved questioned renderings of the text, and brought forward words that had been previously undecipherable to us. I also looked at the 30 April 1804 letter, which though torn and faint, now offers us a more certain transcription and will allow us to eliminate several [. . .] in our initial draft. The Henderson family land transaction case to which this correspondence all relates is still cloaked in mystery because of the incomplete extant paper trail and Jefferson’s own desire to conceal his involvement in it. But having your scans will help us provide a much-improved reading and a clearer context for the remaining historical evidence of Jefferson’s involvement in these important Albemarle County land purchases.”

-Dr. Martha J. King, Ph.D.

Associate Editor

Papers of Thomas Jefferson

Princeton University

The substantial impact on the work of Ms. Francavilla and Dr. King indicates that multispectral imaging has the extraordinary potential to both enhance and expand the academic research of scholars and students. The information revealed from multispectral imaging also has the potential to lead to entirely new discoveries, opening new topics of research that were previously inaccessible to the academic community.

In addition to its current and potential future impact on academic research, the funding provided by the Jefferson Trust to cover the cost of student wages for the multispectral scanning project has had a direct impact upon the student employees of Digitization Services. This digitization project has afforded student employees the opportunity to interact directly with the Jefferson Collection and to uncover previously hidden details within the manuscripts.

What problems, if any, have you faced as you rolled out the project?

The software for the scanner is still under development, though the manufacturer provided no updates during the course of the project. We learned how best to use the software and which documents would obtain the best results out of the scanner. Armed with this knowledge, we pared down our list of documents to scan what we felt would be the most promising ones.

Total number of people working on the project and number of students involved:

Over the course of the project, we had one staff member working on the project and three student employees.

How many people did the project impact:

I believe we will see the impact of this project for years to come, and it is unclear how many people will be affected over time. We directly impacted the research of two of the major Jefferson Papers projects, as explained above, and we anticipate making many more contributions to scholarship by utilizing this scanner for future projects on other materials.

Do you feel that the project has had an impact on the University and /or the University community?

We recently worked with the Office of University Communications to create a photo essay to highlight our work on this project. The photo essay received a lot of attention through various social media outlets for the University and the Library. I knew we had reached the student body in some way when a student offered to volunteer for the Library as a direct result of the photo essay.

I believe we will see future impacts as scholars learn about the scanner and its potential for use in their projects, which they may discover as a result of the photo essay or as a result of visiting Special Collections.

Are there plans for the project to continue once this funding cycle is complete? If so, what are those plans?

This particular project is done, so the next step is to apply the things we have learned about the scanner and its capabilities with 18th and early 19th century manuscripts to other materials and projects. The Library has many other types of materials we can explore using multispectral scanning, and once librarians and scholars identify likely candidates for attention, we hope to make many more discoveries.

Other comments:

The Library is very grateful for the generous support of the Jefferson Trust for this and other projects that have been funded by the Trust in the past, and we look forward to new collaborations in the future.

Report prepared by:

Taylor Krystkowiak

University of Virginia Class of 2016

Distinguished Major in Political Philosophy, Policy, and Law

B.A. in History

Christina Deane

Head of Digitization Services

University of Virginia Library