BJS has no permanent, paid staff.  It is run by the faculty members of the Program in Judaic Studies at Brown University, who are not compensated for their service.  The Managing Editor is currently Professor Michael Satlow, who served as the Principle Investigator (PI) for this project.  He hired and oversaw the Project Manager (PM), Laura Foster; generally coordinated the project and maintained its budget, and set the strategic direction; reviewed the proofreading and new prefaces; negotiated with relevant outside parties, including the vendor and partner organizations (e.g., SBL Press and the Association for Jewish Studies); and directed the rollout of the volumes.  The grant compensated him for these additional tasks.

The PM, who was hired on a one-year, full-time contract funded by the grant, was in charge of the project’s day-to-day operations.  These included organizing and keeping track of the files and manuscript preparation, developing a marketing plan, etc.  The PM worked closely with the PI on all aspects of the project and her expertise and previous experience in publishing was critical to the project’s success.

The administration of the finances is generally handled by the administrative coordinator in the Program in Judaic Studies as part of her normal job responsibilities.  For this project, however, most such matters were handled by a different part of the university specifically devoted to grant management.

We partnered with Newgen KnowledgeWorks (Newgen) to manage the digitization of the 52 back volumes. We worked with a highly skilled Project Manager, Angie Sheltz, at Newgen based in Austin, TX who worked with a team that integrated edits and created the final files. Newgen created PDF, ePUB, MOBI and TXT files for each volume in the project. Thanks to their work, the final files were easy to submit to each vendor and required no additional work from the PI or PM. Newgen charged BJS $1.20 -$1.40 per page to digitize the volumes. Additional fees were charged for pages that contained Hebrew and/or illustrations.

Project Organization

Our first step in the project was to set-up an organizational system to accurately track our progress and the details for each volume. The PM created a master spreadsheet that contained a project tracker, as well as numerous tabs with title and project specific information, such as ISBN’s, LCCN’s, DOI’s, contact information and payment amounts. The spreadsheet was a go-to resource during the project, and now at the conclusion of the project serves as an archive.

To share files between the PI and PM, we set up a Google Drive, so the PI and PM could access documents remotely and both had access to project information. The master spreadsheet for the project was also saved on the Google Drive.

BJS owns the copyright to all of our titles (although we did have unanticipated copyright issues, on which see below).  For about half of the titles, we already had a digital version (usually pdf); for the other half we had only the print version.  We had initially intended to commission about ten authors to write new prefaces in order to add value to, and hopefully increase interest in, the new open-access volumes.  We quickly decided, however, to provide all authors a chance to make minor changes (primarily corrections) to their books and to provide some additional proofreading.  This was our workflow:

1. We reached out to authors to explain the project and ask if they would like to make edits to their volume. All the authors that we contacted were excited about the project and honored that their volume had been selected. We offered to all of these authors the opportunity to submit a new preface or aftermatter.  Of the 52 volumes, we were able to get in touch with 40 of the authors; 12 authors either lacked contact information or were deceased. Three authors decided to submit revisions to their original text; and 11 authors decided to submit a new preface and/or bibliography for the digital version.  Authors that choose to submit a new preface were compensated $500 for their time.  For the authors accepting compensation, we did not offer additional proofreading.
2. To put this process in place, we needed an electronic version of the volumes. For 29 volumes, a PDF was available from SBL Press, as the title was available in print-on-demand (POD). For the volumes that did not have a PDF version, the volume was scanned at Brown.  The resulting electronic file was used only for this stage of the workflow.  In retrospect, this was probably less efficient than sending the physical volumes that we scanned directly to our digitization vendor, Newgen (they ultimately had to go to Newgen in any case to create higher-quality scans) for them to produce the OCR, scanned files that we could use as we did the PDF files provided by SBL Press.
3. For the remaining volumes (with the exception of edited volumes), we recruited graduate students to proofread the books and make typographical (and factual, if they catch them) changes. Proofreaders were found through social media and email contacts. Six proofreaders with subject knowledge expertise in Judaic studies proofed 33 volumes over five months. The proofreaders were compensated $100 per volume for their work. The proofreaders were very conscientious in returning manuscripts on time and taking a high level of care with the work.  Proofreaders made edits directly into the PDF version of the book, or by hand with the corrections being scanned and returned to Brown.  In hindsight, we would have requested that proofreaders make edits using the PDF edits tools instead of by hand, as it time consuming to review and integrate edits made by hand and re-scanned.
4. The Project Manager (PM) reviewed each of the corrections, standardizing them so that our digitizing vendor, Newgen, could incorporate them into the new files.  Newgen was flexible with what they could work with.   They accepted changes made to the PDF files (with the addition/strikethrough features); comments made to the electronic files; and scans of marked pages.  When the PM had questions relating to the content of the corrections, she referred them to the Principal Investigator (PI), who had subject expertise.  The PI would make the final determination about the changes.   All files were also reviewed by the PI before they went to Newgen.  This workflow - the files to the proofreaders; back to the PM; to the PI; back to the PM - was a bit cumbersome and took more time than we expected.  If we are able to add more books, in the future we might have proofreaders submit corrections directly to Newgen.
5. While we worked through the correction process, SBL Press created new covers for the volumes included in the project. The BJS covers had been recently redesigned and we updated the cover design for all 52 volumes. It cost $2,000 to redesign the covers.
6. Newgen prepared two types of digitized files, (1) a PDF and (2) a master digital version from which the other versions (e.g., EPUB, html) were generated.  It turned out, unexpectedly, that our ability to modify each of those versions differed.  As we tried to integrate the changes suggested by the proofreaders, we discovered that only in some cases could we modify the PDF as we wished.  In other cases, the changes would show up in the file in strange and unpredictable ways (e.g., different fonts; changing the spacing).  As a result, for files whose original PDF files we could not modify, we had to prepare two different versions of the book.  In such cases, the PDF version remained unchanged (but could incorporate new front and back matter, as long as the original text remained unchanged) while the changes were fully integrated into the master digital version.  
7. In order to be clear about the changes we made to the book in its new open-access format, and to ascribe proper acknowledgement for the support of the NEH/Mellon grant, we prepared a short Publishers’ Preface for each volume.  The Publishers’ Preface noted generally what changes were made (e.g., typographical changes; new author preface) and contains a few sentences about the book itself.  The Publishers’ Prefaces were all written by the PI and sent to Newgen with the corrected book files for merging. 
8. New title and copyright pages were prepared for each title.  See “Editions,” below. 
9. We submitted our corrected files in batches.  The first batch of 9 volumes was submitted in September, 2019, the second batch of 32 titles in October 2019 and the final batch of 11 titles in January 2020. The final batch of volumes that were submitted contained the volumes that had a new preface and/or bibliography that was submitted by the author. We originally asked each author to submit the new content by September, but most content was not received until December/January. This did delay our original timeline, but did not prevent us from meeting the overall project deadline.  We decided on batch submission (and roll-out; see below) in order to establish and debug our work-flow.
10. For each batch, Newgen sent back draft digitized files in PDF format.  The PM checked these files to assure that the corrections were properly included.  Any corrections were sent back to Newgen, who would incorporate them and resend the corrected draft files.  When the PM was satisfied that all the corrections had been properly made, she would authorize Newgen to produce the final sets of files.   Once we received the final files, we began submitting electronic files to each of the five online platforms, JStor, ProjectMuse (Muse), Brown Digital Repository (BDR), and ACLS Open Humanities (ACLS) Each platform hosts content in PDF format; JStor and Muse also host an HTML version of each volume; and PDF, ePUB, Text and Mobi files are all available through BDR.  We had planned to submit files to HathiTrust for hosting as well, but HthiTrust stopped receiving new publisher files in March 2020 due to Covid-19.

We debated about whether the new digital versions, especially those that had corrections and/or new material, should be registered as new editions and thus requiring new ISBN and Library of Congress Control (LCNN) numbers.  Both are required metadata that appear after the title page in the new versions.  We decided that we would register them all as new editions.

As each volume contained new content, a new ISBN was assigned for each format (PDF, EPUB and POD) and a Library of Congress Control Number (LCCN) was registered. The ISBN’s were obtained through Bowker with an account set-up by SBL Press. BJS was charged for the ISBNs through our existing arrangement with SBL Press. The LCCN was obtained through the Library of Congress, which is a free service. The new LCCN was included in the copyright page for each book, and the MODS Records that were created for each new LCCN were used to submit Metadata to BDR (detailed below in Metadata).

As mentioned, we had few issues with copyright clearance, as BJS customarily owns the rights to the books it publishes.  We did not, however, own the rights to the images included in those books, many of which were published decades ago. 

Very few of the BJS volumes contained illustrations--only 7. In those cases, we contacted each of the authors and asked them to clear permissions for us to reuse the images in the digitized version. In cases where we did not hear back from the author or the author was not able to clear permissions, we removed the image and placed a note in the text stating the image had been removed due to copyright.

In the future, BJS might ask all authors who include images in their volumes to secure copyright permission for the print and digital version, so that we might republish them in future digital editions.

Before submitting the final files to each online host, we prepared metadata for each provider that they used to ingest content into their system. We followed best practices, as recommended by our vendor partners.  We also investigated the practices of the open access publications at the Institute for the Study of the Ancient World, in order to develop metadata that would optimize discoverability of these electronic files. The process for creating the metadata was as follows:

1. A metadata spreadsheet was created for JStor, Muse and ACLS, using an excel spreadsheet template provided by each individual platform. Each online host required specific information in a specific order to easily ingest the final files.   The information that was required by each provider included the ISBN, author, title, and publication date.  In addition to the information they required, we also included the licensing information, and BJS website to ensure both items were included in the metadata. 
2. For the BDR, we provided a Metadata Object Description Schema (MODS) Record for each of the volumes instead of a metadata spreadsheet.  MODS records were developed by the Library of Congress to organize a book’s data. The MODS Records were created by the PM using the MODS Records from the Library of Congress catalog. With the help of the Center for Digital Scholarship, the PM added the licensing information and the book description to each record, so it would appear in BDR.  The MODS Record can be downloaded from BDR along with the other file types.

After preparing and submitting the metadata, we began submitting electronic files to each of the four online platforms, JStor, ProjectMuse (Muse), Brown Digital Repository (BDR), and ACLS Open Humanities (ACLS). Each platform hosts content in PDF format; JStor and Muse also host an HTML version of each volume; and PDF, ePUB, Text and Mobi files are all available through BDR.  There were nuances to submitting content to JStor, Muse, ACLS and BDR, but the general process is detailed below:   

1. A method of transferring files, file requirements and process for submission was established with each host. As final files were created they were submitted to JStor, Muse, BDR and ACLS.   JStor and Muse had systems in place to accept files (ftp site for JSTOR and web portal for Muse). JStor and Muse posted content within 5-10 business days. Shared Google drives were set-up to submit content to BDR and ACLS. BDR uploaded content in two batches and ACLS uploaded all content at once.
2. In addition to submitting content to be hosted on each platform, we also provided PDF and MOBI files to SBL Press to arrange for each volume to be available POD and as an eReader on Amazon. SBL Press will also work with EBSCO to make the volumes available electronically through their collection offerings. Files were uploaded to a Google Drive for SBL to access and they will make them available at a later date. 
3. As of June 30, 2020, we have not been able to submit content to HathiTrust for hosting. HathiTrust stopped accepting content from new publishers in March 2020 due to Covid-19, and as of that date BJS was not an established publisher. .
4. Once the volumes were available online, we coordinated with the University library to update the MARC records, so the digitized volumes would be included in search results within the library. University Library used data from WorldCat to update the MARC records with links to the digital version of the volumes. 

A DOI is a unique identifier that permanently identifies a specific volume. It was important for us to assign a DOI to each volume to increase discoverability and make them easy to cite.

JStor and Muse both assigned a DOI to each volume and deposited the DOI in CrossRef once the volume was hosted on their platform. ACLS and BDR do not automatically assign DOI’s to content hosted on their platform, but luckily, Brown University has the capability to assign DOI’s through the University Library.

 The process involves completing a DOI request that includes the title, author, URL, publisher and publication date for the volume. With the URL handle for each volume on ACLS and BDR, the PM completed the request process to assign DOI’s for each platform. As a DOI can only direct to one URL handle, each volume has four DOI’s, one that directs to content hosted on each platform, JStor, Muse, ACLS and BDR.  Each individual platform lists the DOI associated with the volume hosted on that platform, and all four DOI’s are cataloged in the MODS record on BDR. HathiTrust does not assign or catalog DOI’s.

We made text files available on BDR to enable researchers to conduct their own topic modeling. Text fields for all 52 volumes were zipped and saved on BDR in the BJS collection, which makes them easy to use. With the help of the Head of Digital Scholarship Project Planning, we drafted instructions and provided additional resources to help academics conduct their own topic modeling. 

When content was available online, we announced the project through various channels. Our activities focused on announcing that BJS volumes are available online, open access, and driving usage to each volume. We conducted the following activities in May to share the good news with the Judaic studies community

1. We created a webpage on the BJS monograph series website that explains the project and lists each volume that is included (https://www.brown.edu/academics/judaic-studies/monograph/bjs-open-access-books). Each volume listing includes a link to download a free version, and to purchase a POD version. We used the webpage as a landing page that we can direct people to when marketing the project.
2. Each author whose title is included in the project was sent an email when their volume was made available online. We encouraged each author to share links to their volumes via email, their personal website and social media. We have also had some authors make their volume available through Academia.edu.
3. We sent a press release to the Executive Director/President of related societies including American Academy of Religion, British Association of Jewish Studies, European Association of Jewish Studies, World Union of Jewish Studies, Modern Language Association, American Historical Association.
4. A graduate student compiled a list of Judaic Studies scholars globally that was sent to SBL Press to be added to the marketing contacts database they use when marketing BJS volumes.  SBL Press sent an email to their list of contacts announcing the project and directing people to the online version of each volume.
5. The project launch was announced through two list-servs, H-Judaic and the Association for Jewish Libraries.
6. We informed the Association for Jewish Studies, who is preparing their own press release.
7. Through our relationship with Muse, BJS volumes were also indexed in the Directory of Open Access Books (DOAB). DOAB serves as another route to BJS volumes and helps to reach people outside of Judaic Studies.

We also planned to discuss the project in person to share our experience with the Brown community, and create a vodcast that could be hosted on the BJS webpage. However, due COVID-19, Brown University, like all colleges and University, closed to in-person classes in March 2020, and as of June 2020 has not yet resumed on campus activities. We hope to conduct these two activities once we are able to meet in person on campus.

The first volumes launched on Muse in March, JSTOR in April, BDR in May, and ACLS in June. Each online host provides usage statistics for the BJS volumes that are hosted on their platform. The statistics are provided in different formats, and each platform counts downloads differently. The number of downloads on JStor and Muse have increased each month, and can be seen below:

Overall there has been more usage to volumes hosted on JStor than content hosted on Muse. This can be contributed to the marketing activities that were discussed in the previous section. The marketing campaigns included links to each volume hosted on JStor. 

It is also interesting to review the geographic breakdown of usage on Muse. In the first two months, 28% of usage to BJS was from the US and 23% was from China.  A full geographical breakdown of usage on Muse from March through May 2020 can be found below:

Overall, we would categorize the project as a success. 52 BJS volumes are now fully accessible open access on four platforms. BJS’s missions is to disseminate content as widely as possible and we already feel we are reaching a wider audience. For comparison with the usage figures above, BJS typically sells a total of 500-600 volumes per year.

Authors were very supportive of having their volume included in the project. Many were delighted to contribute new pieces and everyone was excited at the prospect of their books reaching a new generation of scholars. We are hopeful that the project will also help to attract new authors and encourage authors to work with BJS again. 

As more usage data is available we will be able to compare and contrast the figures. It will be valuable to analyze usage by book and country. This data will help us to evaluate future projects, as well as books that could potentially be part of a second project. Reviewing the data by country and comparing this with our previous sales data will let us know if we are reaching a larger audience. We are also interested in reviewing the POD sales data with the previous sales data for each volume to evaluate the impact to revenue of making a volume available open access and determine if it is a sustainable model for the future.

The biggest challenge we faced was integrating edits into the text of each volume. In the future, it may be worth considering only integrating new material and therefore handling edits to the text as corrigendum in the back of the book.  The proofreading and edit integration was the most time consuming portion of the project. It was also a piece that we had to adapt to after assigning volumes to proofreaders. We did not realize edits could not be integrated into the text of volumes until our first batch of content was submitted to Newgen.

We also faced a challenge with the volumes that were only available as hardcopies. The project could have been more efficient if we sent the hardcopies to Newgen for scanning. We scanned the books in house in order to send them to proofreaders and authors, and thought it would save time during the digitization process. However, the scans were too low quality and the hardcopies had to be sent to Newgen for scanning. This resulted in duplicate tasks that could have been completed more efficiently.

As expected, we experienced time delays due to the COVID-19 pandemic. In mid-March 2020, academic institutions and most companies were closed and transitioned to online learning and remote working overnight. The transition to remote working for some of our vendors was more involved and delayed stages of the project. For example, it took approximately 8 months for content to be digitized with Newgen, which we did not predict.  The delays were due to Newgen transitioning their workforce to remote working and reassigning staff. Newgen, however, did do a very effective job in managing the changes and working with us to complete the project on time.

While COVID-19 has presented a challenge, it also highlights the need for BJS content to be available digitally. We are hopeful to be able to find a sustainable way to continue to make BJS volumes available open access. The volumes serve as a tremendous resource to the Judaic studies community and we want to ensure they are easily accessible to everyone. BJS is investigating options to make each volume available online and discussing sustainable ways in which we can continue to experiment with open access.