Charles Olson's Melville Project
Charles Olson's Melville Project Navigation Bar

About the Project

1. Description and goals of the project.

The University of Connecticut Archives & Special Collections at the Thomas J. Dodd Research Center was awarded $40,000 by the Gladys Krieble Delmas Foundation in December 2001 to clean and make accessible a series of hand-written but subsequently water-damaged cards produced by the poet Charles Olson during his effort to transcribe the marginalia in hundreds of books owned by Herman Melville. Access to CHARLES OLSON'S MELVILLE PROJECT, a new resource for Melville scholars, is through the World Wide Web. Under the direction of Rutherford Witthus, Curator of Literary Collections, the project was completed in 2004.

The renovated cards are scanned at a resolution of 600 dpi to allow presentation of each image at full size (5 x 8 inches) with the capacity for zoom enlargement. Originally, we intended to link the scanned images to a database of metadata about each image: title; date; description; content-specific subject headings; actual transcription of each card; digital format, resolution, and file size; and image identification number. The images, transcriptions, and metadata were to be presented on a database-enabled thematic web site from a server at the University of Connecticut Libraries. However, midway through the project we saw the advantage of using the new JPEG standard, JPEG2000. This new standard allowed us to embed all of the metadata and transcriptions in the image file itself, thereby eliminating the need for additional and separate databases. All supporting data for an image is therefore carried with the image file no matter where it resides, a definite advantage in our overloaded digital systems. As far as we know, we are the first academic library to use the JPEG2000 standard for creating digital surrogates of archival materials. Of course, scholars are welcome to use the cards themselves under supervision in the John P. McDonald Reading Room of the Thomas J. Dodd Research Center.

Charles Olson (1910-1970), the innovative poet, literary theorist, and essayist whose work had an international influence on young writers during the 1950s and 1960s, began a serious study of the life and works of Herman Melville while a graduate student at Wesleyan University in the early 1930s. Olson's Master’s thesis, The Growth of Herman Melville, Prose Writer and Poetic Thinker, was completed in 1933, but his work on Melville continued. Olson was one of the first scholars to consider the importance of Melville's reading and marginalia.

In the 1930s, Melville's surviving literary manuscripts, letters, personal papers and journals, and reading library were still, for the most part, in the possession of the family and a few institutional or private collectors. The most substantial collection of Melville materials unaccounted for at that point—and the materials that Olson pursued most vigorously—were the "lost five hundred," the approximate number of books Melville's widow had sold to a Brooklyn dealer in 1892. As a young scholar, Olson was indefatigable in his research; when he located a volume from Melville's library in a grand-daughter's home, in a private collector's hands, or on a public library's shelves, Olson carefully transcribed onto 5 x 7-inch note cards complete bibliographic information on the volume, as well as the content and location of Melville’s annotations and reading marks. Charles Olson’s note cards are, in a few important instances, the only account of Melville’s reading marks in books whose location is now unknown. Olson’s notes also provide scholars with Melville’s marginalia in volumes currently in private hands and not readily available to scholars.

In addition to the note cards on books from Melville's library, there are two other groups of cards at the University of Connecticut. On one group of cards Olson captured his notes of interviews and recorded his astonishingly thorough methods for tracking down relatives of those known or thought to have bought books from Melville’s library. Other note cards were used by Olson to record his reading and critical notes on Melville's published works. In all, nearly 1,100 note cards survive.

Unfortunately, when Olson moved away from Melville scholarship after the publication Call Me Ishmael (1947), he stored the results of his investigative work in a trunk in a friend's basement. Countless water leaks over the years damaged the note cards containing the transcriptions and research notes. Some cards were merely soiled; others were fused together in large blocks. After the University of Connecticut purchased the Olson papers in 1973, the note cards were stored separately while awaiting appropriate preservation measures.

Box of notecards
notecards damaged

The condition of the note cards before conservation,
showing the water and dirt damage.

Some were easily separated, while others are fused
together in a block.

A small pilot project to separate, minimally clean, and scan some of the "fused" cards was undertaken in Spring 2000. Of the 60 randomly chosen cards, three cards recorded Olson's notes on three previously unknown books from the Melville family library and others captured Olson's interview notes with Melville descendants.

typical note card
notecard magnified
A scan of a typical note card with water and dirt damage,
after separating and minimal cleaning.
With magnification on the Web, the defaced handwriting
can once again be read.

With the success of the pilot project and with the proven richness of the note cards as resource for scholars, the complete collection of note cards have been scanned and are available for remote viewing through our website to scholars interested in studying this important component of Charles Olson’s pioneering investigative work on Herman Melville.

The Melville cards are part of the Charles Olson Research Collection. This large and complex collection is currently being reorganized to provide better and complete access. An element of the reorganization is the construction of a new electronic finding aid, created using Encoded Archival Description (EAD). The images with their distinctive metadata will also be linked in the finding aid, thereby creating additional access within the context of the collection. All cataloged manuscript collections at the University of Connecticut appear in the OCLC (Online Computer Library Center) union catalog.

Two experts outside the University of Connecticut have been involved with the project. Debora Mayer, a paper conservator who has extensive experience working with primary materials, was contracted to separate and surface clean the approximately 1,100 note cards. Dennis Marnon, Administrative Officer at the Houghton Library, Harvard University, consented to provide the intellectual infrastructure necessary to understand Olson's work in the context of Melville scholarship. He also agreed to help with transcription of Olson's difficult script.

2. Preparing and cleaning the cards.

When the project began, we estimated that 1,100 cards needed cleaning and repair. We could only make an estimate because many of the cards were tightly fused together so that it was impossible to determine the exact number in each block. We sent approximately 70-80 cards per shipment to the conservator. At this point, we have received 814 cleaned and sleeved cards. In a few cases, small blocks of fused cards were left untreated because the possibility of information loss was great. We have segregated those cards, awaiting further decisions and treatment. The remaining cards will be sent in two batches for evaluation and treatment during the next few months. We had hoped to have all cards finished by now but the conservator ran into some extremely difficult situations along the way. At the end of the project, we will send the untreated blocks back to the conservator to determine how much loss will occur if we have the cards separated and, at the same time, determine the value of including those cards for the overall success of the project.

The cards were treated as follows:

  • A sample of cards was documented photographically before and after treatment.
  • Cards were separated from one another with a micro-spatula inserted between the cards.
  • Each card was dry surface cleaned, recto and verso, to the extent possible with a latex dry cleaning square.
  • Each card was humidified in a humidity chamber followed by drying between blotters and under weight and pressure. Crimps and creases were relaxed during the humidification process.
  • Vulnerable tears and very weak areas were reinforced with Japanese paper and wheat starch paste.
  • Each card was placed in a clear polyester (Mylar-3) 3-sided pocket enclosure with perma-life interleaving paper as a support.
  • At all times during the treatment, the order of the cards was kept as received.
  • Each treated card was photocopied prior to shipping back to the University of Connecticut. The conservator retained the photocopy until the shipment was received, at which time it was sent along to UConn.

3. Scanning the cards.

As each batch was returned to the University, a graduate student trained in the fine points of scanning rare and unique materials scanned each card as a 600 dpi TIFF file. At the same time, a smaller size file was created as a JPEG for use on the Web. Our original intention was to create three resolutions: the TIFF for archival purposes and for printing requests; a medium size JPEG file for use on the Web; and a thumbnail JPEG for use with the electronic catalog record and the EAD finding aid. Midway through the project, JPEG2000, a new international standard for JPEG files, was introduced. This new standard allows multiple resolutions and its concomitant metadata to reside in a single file. The files are considerably smaller than earlier TIFF files and are able to save the image in this smaller environment without losing any information. These JP2 files are produced in batches by compressing the TIFF files and their metadata. This new standard eliminates the possibility of the individual image and metadata files from being separated from each other, thereby maintaining the integrity of the data. The University of Connecticut Libraries agreed to purchase Aware, Inc.'s compression software that was necessary for CHARLES OLSON'S MELVILLE PROJECT to pilot this innovative standard. While a few changes to the original procedures were required, the overall impact on the workflow of the project was minimal. The major impact was the extra time and staff required to set up and implement the new standard.

Our interest in using JPEG2000 brought us together with Ronald Murray, Digital Conversion Specialist in the Preservation Reformatting Division of the Library of Congress. Mr. Murray has been following the progress of the JPEG2000 standard and has been engaged in testing the results of this new image compression standard. He has consulted with us during the implementation phase of our project, providing us with considerable image testing and enthusiastic support. To our and his knowledge, the Delmas-supported CHARLES OLSON'S MELVILLE PROJECT is the first project in an academic library or archives to use this new image standard.

As of July 1, 2003, 743 cards had been scanned and were awaiting conversion to the JPEG2000 format. The procedures and workflow for creating these new kinds of files have been worked out and will be available on the web site for other sites wishing to follow our work plan. We are currently processing the files as part of our regular workflow in the digital laboratory at the Dodd Research Center. We expect by first of next year to have all 1,014 scanned images available for research on the Web.

4. Preparing the metadata.

Melissa Watterworth, an applications specialist at the Homer Babbidge Library, has been instrumental in moving this project forward. She has created the databases required to gather the metadata for various aspects of CHARLES OLSON'S MELVILLE PROJECT as well as the larger Charles Olson Research Collection on the Web. Copies of entry forms for the database are available in Appendix 1. The following pieces of information are captured for each of the images:

  • Descriptive metadata: title, source, author, electronic publisher, transcriber, card caption, image name, transcription, and transcriber's notes.
  • Technical metadata: file name, source width, source height, scanning technician, file size, input dimensions, output and dimensions.

5. Transcribing the information on the cards.

Dennis Marnon from the Houghton Library at Harvard has spent many hours evaluating the cards as they were returned from the conservator. He has suggested the following format for transcription:

  • Card number
  • Text of card
  • Physical description of the card, including significant losses of text. "Index card, written in pencil with ink additions. Severe stains, losses at edges, resulting in undeciphered or lost text."
  • Textual notes, especially complicated ones. Brief corrections (not that many, actually) will be made in brackets in italics in the text.
  • Explanatory notes. Who, what, when, etc.
  • Biographical notes (slightly different format from the explanatory notes above). "Frank Jewett Mather (1868-1953), art historian and professor at Princeton."
  • Bibliographic citations (Sealts, Leyda, Parker, and NN eds. and not many more) and locations. "Sealts 1 (Harvard)."
  • Content notes as they relate to other cards: "Card 1 of 23 on HM's markings in D'Avenant. See Cards 90-111, marked by Olson "2.- 23. Davenant." Or "See Cards 300-302, 345, 346, and 388-395 for other prints owned by HM and seen by Olson."

6. EAD finding aid for collection

As with all of the manuscript collections in Archives & Special Collections at the Thomas J. Dodd Research Center, a detailed finding aid of the Charles Olson Research Collection is being recreated using Encoded Archival Description (EAD). The original printed and later HTML version of the finding aid is now inadequate to display the rich content that has been uncovered in the course of this project. CHARLES OLSON'S MELVILLE PROJECT will be expanded from a single line in the original finding aid ("Damaged material on Olson's Melville Project") to a full series with links to the images and transcriptions on the Web. This section will be used as a model for other areas of the Charles Olson Research Collection that we expect to undergo major revisions in the future.

7. Catalog record on OCLC.

All manuscript collections with a finding aid or other collection guide are cataloged into OCLC. The current record may be updated when the final reorganization of the collection is finished. A copy of the OCLC catalog record is available in Appendix 3.

8. Charles Olson Research Collection on the Web.

One of the major goals of this project is to make CHARLES OLSON'S MELVILLE PROJECT accessible to the scholarly research community on the Web. Because this project is one component of a larger collection, it was necessary to design a complete website to incorporate Olson's work on Herman Melville.


About Charles Olson's Melville Project Browse Charles Olson's Melville Project Search Charles Olson's Melville Project Contact Us Archives & Special Collections at the Thomas J. Dodd Research Center University of Connecticut Libraries The Charles Olson Research Collection