The 1978 volumes have been included because they contain entries for pre-1978 registrations that were not complete at the time of publication of the 1977 CCE’s.  These entries appear in a separate section at the end of each volume.  All registrations made under the Copyright Law that went into effect on January 1, 1978 are available online at the Copyright Office website:  http://www.copyright.gov/records/.

Scanning of the cards in the Copyright Card Catalog is continuing with completion of cards back to1955 expected in a few months.  Based on the positive feedback to the recent posts about a virtual card catalog, we are also researching how to best construct such a catalog using derivative images from these scans.

Study and testing of the use of optical character recognition (OCR) to capture data from the card images is proceeding as well as the feasibility of using crowd sourcing.  Some of the records will lend themselves to these processes and some probably won’t.  We’ve also recently engaged several staff from all parts of the Copyright Office to assist on a volunteer basis with analysis of the card formats to determine ways to parse out the titles and the author and claimant names in order to produce index terms for a full online search capability.  It remains our ultimate goal to provide a search capability that spans both pre-1978 and post-1977 records and that supports searching by title and name with means to narrow the results to the particular item of interest.  Work is proceeding on a demonstration model of such a database and we hope to provide you with access to that model in the not too distant future to obtain your feedback and comments.

Progress is being made on both the preservation and the access fronts and we’ll keep you posted on new developments.  In the meantime your input is most welcome and most appreciated.

Because of the positive response, we intend to pursue this option.  It seems to be a good interim step while we figure out how to muster the resources, maybe through crowd sourcing, to achieve the eventual goal of robust word searching of titles and names.  Some of the comments included specific suggestions such as the ability to skip some preset number of cards and to display only the top half inch of the cards in the scrollable search panel with a full card display in an adjoining panel.  These suggestions are most helpful and most welcome.

While the overall card catalog is large, it’s divided into six time periods and a seventh set for assignments and transfers.  Two of these seven sets have already been digitized and the completion of a third one is near.  The sets could be made available as they are scanned; no need to wait until they’re all done.  Good performance will be fundamental to the efficacy of a virtual card catalog and that will be a key factor in selecting the type and size of derivative images to be displayed as well as how they are organized.

We’ve already begun market research to find out who has done this before and to benefit from their experience.  There are probably pitfalls to be avoided and we don’t want to reinvent the wheel.  We’ll reach out to organizations and particularly libraries that have similar online catalogs.  The Internet Archive was mentioned in a couple of comments.  We’ve been working with them for the past two years on the scanning of the published Catalog of Copyright Entries and will consult with them on this initiative as well.  The Princeton University Library has a supplementary online catalog that is somewhat like what we have in mind and we’ll seek input on their experience.  If you are aware of others that have put virtual card catalogs online, please let us know.

I’ll report on plans and progress on this initiative in future posts.  Thank you again for your feedback and I look forward to receiving more of your comments.

As an interim step, the Copyright Office is considering making the images of the cards in the catalog available online through a hierarchical structure that would mimic the way a researcher would approach and use the physical card catalog. We’re calling this a virtual card catalog.  While it would not provide the full record level indexing that remains a principal goal, it would make information available as we’re doing the scanning and as searchable as the actual cards.

The card images have been organized by drawer, each in its own folder, and the image file names contain the time period, the drawer label, a sequential four digit number starting with 0001, and occasionally an alphabetic suffix when information exists on a verso or there are multiple card images for a single entry.  So there’s already a hierarchical organization of the images that could enable a virtual card catalog.

But how would this virtual card catalog look and operate?  A search would probably begin at the top of the hierarchy with the selection of a catalog segment  (e.g., Registrations from 1971 to 1977) perhaps from a drop down list.  This would be followed by the entry of a search term (i.e., a name or a title).  The software would step down to the next level of the hierarchy within the selected catalog segment and find the “virtual drawer” folder that alphabetically within the segment should contain the term and then display that drawer label along with labels for some number of drawers immediately preceding it and some number of drawers immediately following it.  The researcher could select any one of the drawers displayed or return to the initial search screen.  For a selected drawer the software could display small scrollable images in one panel and a full card image in another panel.   One could scroll through the smaller images to different points in the virtual drawer, select and display a specific card, navigate to the next card, the previous card, the beginning of the drawer, and to the end of the drawer.  The software could support a return to the list of drawers, forward and backward navigation at the drawer level, and a return to the initial search screen.   The following is a mock-up of how the card images might be displayed.  Click on the image to see a larger display.

Mock-up of a virtual card catalog display

 We are exploring multiple ways of making the Copyright records available online sooner rather than later.  The notion of a virtual card catalog is an example and one that could probably be done at a modest cost.  It sounds good to us but we want to hear what you think of it.  While not the optimal solution, would it nevertheless be useful to you as an interim step?  Do you know of other organizations that have done something similar and done it well?  Please take a moment to consider this option and let us know what you think.

Last December, a post on the Library’s Performing Arts blog In the Muse told the story of a niece who found among the Copyright records 42 songs registered in the 1920s and 1930s by a great uncle, most still unpublished and hidden away for over 70 years (http://blogs.loc.gov/music/2011/12/pic-of-the-week-uncle-bennie-edition/). 

These are touching stories about creative accomplishments that bring out the personal side of Copyright records, records that have meaning not only to the creators but also to their families.  The pride of a grandmother showing her grandchildren the records of songs that she wrote and the thrill for a niece finding records of a great uncle’s long lost songs stoke the fire of a passion that I and my colleagues share to make these records available online sooner rather than later.

Beyond being a source of family pride, the records also show where copyright may still persist for specific works. The amendment of the Copyright Law in 1992 made renewal automatic for works still in their first term of protection and made renewal registration optional for works originally copyrighted between January 1, 1964 and December 31, 1977.  In the case of the song registered in the early 1970s, the copyright very likely has not expired and would persist under the present law until at least 2065.  For the songs registered in the 20s and 30s, a search of the records for renewals would tell about their status.  So the song, book, or other work that has lain silent perhaps for decades could be the subject for the next hit tune or blockbuster motion picture, carrying with it all the benefits that can accrue to a copyright owner.

Copyright records reflect ownership of intellectual property that can have significant commercial value. They are a treasure trove for those doing research on the cultural development of our great nation.  But they also tell the personal stories of creative accomplishments of everyday folks that can be the inspiration for us and others in the future.

So Barbara, that song of so many years ago is still under Copyright protection and your grandchildren could one day inherit those rights.  And we remain committed to the task of making the records available and easily searchable online.  In the meantime, you might want to look in the online Catalog of Copyright Entries that we recently had scanned and that are now available at http://www.archive.org/details/copyrightrecords/ .   The 489 volumes that have been scanned so far contain about 12.6 million registrations dating from 1924 to 1977 in all classes including music, motion pictures, works of art, prints, photos, pamphlets, periodicals, maps, books and renewal registrations.  The CCE volumes are by year, cataloging period, and class of material.  Word searching of the online volumes is available and there are indexes included in each volume or in an accompanying volume.

Page from the District Court 1855 Record Books

The image to the left shows a page from the 1855 record books of the U.S. District Court, Southern District of New York containing the original registration for Walt Whitman’s Leaves of Grass.  Whitman provided the title to the court on May 15, 1855.  The record shows his claim as author and proprietor and it was signed by the Clerk of the Southern District, George F. Betts.

Until 1870 when copyright functions were centralized in the Library of Congress, claims in copyright were recorded by the Clerks of the U.S. district courts.  The district court record books are now located in the Rare Book and Special Collections Division of the Library of Congress.  Microfilm copies of the books are available to the public in the Copyright Office.

Page from the Library of Congress 1883 Record Books

Under the Copyright Law of 1831, Whitman’s original registration of Leaves of Grass had a term of 28 years. On March 15, 1883, two months before the end of the term, Whitman renewed the registration for an additional 14 years in accordance with the statute. The image to the right shows the page of the Library of Congress record book containing the renewal record signed by the Librarian of Congress at the time Ainsworth Rand Spofford.  The record shows Whitman’s street address in Camden, New Jersey.

Whitman continually updated Leaves of Grass until his death in 1892 and registered other editions.  Original registrations for two of these are found in the 1876 and 1881 record books.  Images of those registration records are shown below.

Page from the Library of Congress 1876 Record Books

Page from the Library of Congress 1881 Record Books

These four records are just a glimpse at the interesting stories found within the leaves of the Copyright record books.  The principal purpose of the Copyright records is to identify ownership of intellectual property, but collectively they tell the much larger story of creativity in literature, music and art since the founding of our nation.

Over the past 2 months I’ve shared information with you about the Copyright records and the plans, challenges, and visions for preserving them in a digital form and making them widely available online.  Today’s post is a brief update on recent progress.

Copyright Card Catalog

First, I’m happy to report that digitization of the records is continuing.  In the two months since this blog was started, nearly 2.2 million more catalog cards have been scanned, and the images  inspected and placed in archival storage.  This brings the total to more than 16.6 million cards completed, which is more than a third of the entire catalog.

1928 Catalog of Copyright Entries

Progress has also been made in scanning and making available the Catalog of Copyright Entries with 36 more volumes processed since late December bringing the total now to 456 out of the 660 volumes.  We’re nearing 70% completion with some registrations from as far back as 1928 and including all classes of works.

The volumes scanned so far are available at the link below the CCE page shown to the left.

http://www.archive.org/details/copyrightrecords/

We also have a small group of subject matter experts and information technologists studying how the records can best be indexed and made available to users.  A test database is being used to assess the efficacy of different approaches to indexing and displaying these older records.  Integration with the post 1977 records is a goal, but accomplishing that is not without its challenges.  Records since 1978 have been collected in a database containing multiple indexes for titles, authors, claimants, and other related index terms.  This granularity has enabled flexibility in searching by title only, by name only, or by combinations of different indexes.  But the card catalog consists largely of a single alphabetical index within each time period with names, titles, and other index terms interfiled.  The plan is to capture all of the text on a card and to programmatically parse as much of the data as possible, placing it in the appropriate fields in the data record.  We are studying the data patterns found in the cards to see if specific types of index terms can be distinguished based on position on the card or on unique characters such as the presence of the copyright notice symbol ©.  But the data in the registration cards for the most part is not labeled making it difficult and perhaps impossible to programmatically distinguish all names and titles.

As we work through these challenges I’ll keep you posted on findings and continue to seek your input.

The records are of two types: copies of the original documents and finding aids in the form of catalog cards.  After being recorded the original document was returned to the party who requested recordation, but the Copyright Office retained a copy of the document.  From 1870 to 1927 the copy was a handwritten or typewritten transcription in a record book.  From 1927 to 1953, photostat copies were made and bound into record books.  After 1953 the Office made microfilm copies.  The following are examples of recorded documents.  Click on any of the thumbnails below to see a larger image.

Assignment/Transfer document

Assignment/Transfer document

Assignment/Transfer document

The documents had no prescribed format but frequently took the form of a contract.  The Copyright Office has microfilm copies of all recorded documents as well as PDF derivative copies.

The official copies of the recorded documents are filed by the assigned number which is made up of the record book number or microfilm number followed by the page number.  Specific documents are located through a search of the catalog cards.  There are about 2.5 million cards that make up the indexes to the documents.  A card was created for every title included in a document (about 1.7 million) and a card for each party (assignors and assignees) to a document (about 800 thousand).  Names are filed separately from titles and until 1941 assignors were filed separately from assignees.  Titles were not indexed prior to 1928 although they frequently appear on the assignor and assignee cards.  While there were some changes made in the formats over the years, the data fields are labeled and it is easy to distinguish the type of card.  The following are examples of the cards.

Assignment/Transfer title cards:

Assignor cards:

Assignee Cards:

The labeling of the data fields on the cards and the separate filing of names and titles will facilitate the data capture and enable indexing that’s consistent with current online post-1977 assignment records.  For those looking for ownership information online, these records would be a useful complement to the Catalog of Copyright Entries that we are currently digitizing and making available via the Internet Archive website at http://www.archive.org/details/copyrightrecords.

Please let us know how you would prefer to see these records made available online and whether you’d want to see us do it sooner or later relative to the registration records.  Your input is important and most appreciated.

The following is a list of the search terms that one might include in a search of Copyright records. Are there others?

• Title
• Name (author/claimant)
• Name (assignor/party 1)
• Name (assignee/party 2)
• Type of record (Registration or transfer/assignment)
• Registration number
• Document number (for transfers and assignments)
• Date or date range (effective date of registration, date of recordation, date of execution)

How would you prefer for the search results to be generated? One option is a sortable list of records based on keywords found in specific fields or in any field in a record.  An alternative is a list of all records containing words in the search argument occurring anywhere in a record with the functionality to filter the results by facets reflective of the search terms above.  Would either suit your needs better or is there another alternative that might provide more useful results?

What we can eventually provide will be determined by cost and what the Copyright Office can afford. Nevertheless, we want to know your thoughts so whatever is built comes as close as possible to your needs. Your input is important and most welcome.

Optical character recognition (OCR) and crowd sourcing have been mentioned and both are on the table as ways to facilitate data capture and overcome resource constraints.  Through testing, discussions with other organizations, and input from experts we are aware of the benefits and limitations, but nevertheless see opportunities to apply them to some Copyright records.

As to which records we should focus on first, there were a couple of mentions of assignment and transfer records.  These are a separate set of about 2.5 million cards pointing to about 350,000 assignment and transfers from 1891 to 1977.  They would form a complement to the Catalog of Copyright Entries gradually being made available through the Internet Archive website (http://www.archive.org/details/copyrightrecords).  Also mentioned were records for copyrights first registered between 1923 and 1963.  If these works were never renewed, they may well be in the public domain.  A third set mentioned was the 1971 to 1977 registration records because of the recency of the records.  Let us know which of these three sets or what other set would be most valuable to you.

Some comments referred to how we might construct the data record and how it would appear online.  One option is to include in the data record all of the information in the catalog cards for a registration or assignment/transfer and enable word searching across the content.  Search results could be refined by category and class of material, and limits set by date.  The records would contain pointers to images of the respective paper records.  We are building a demonstration model using a small set of interrelated records (i.e., registration, renewal, assignment for several works) that we’ll share through the blog.  Look for a future post with the URL.

Rather than reinvent wheels, we have conferred with other organizations including the National Archives, Ancestry.com, Internet Archive and others and learned from their experience.  Among the comments were references to some others including the Patent and Trademark Office, Stanford University and the University of Pennsylvania where they are working with copyright or similar records.  We’ll take a fresh look at how they have approached the challenge.

Your input is important.  Thank you for your comments and I encourage you to use this blog as an opportunity to influence how the pre-1978 records are made available online.  The more we hear from you, the better informed will be our decisions.

The following are examples of cards in the Copyright Card Catalog.  These are the principal finding aids for titles, authors, claimants and other names associated with the ownership of copyrights.  Our vision is to create a data record from the content of these cards and make that available online for searching.  Linked to each data record would be images of the cards as well as the respective pages from the record books (see below) showing the full information about each registration and each assignment or transfer of rights.

Serial registrations

Registration card

Registration card

Requests and applications for copyright registration are found in the record books.  The books are organized by year, class  and the assigned number.  Once a registration number is found using the card catalog, the record books can be consulted for the full set of copyright facts about the work.  These are a few examples.

1907 record book

1961 record book

1973 record book

An alternate finding aid is the published Catalog of Copyright Entries.  These were periodic compilations of registration records in catalog card format distributed through the Federal Depository Library Program.

1960 CCE page

1973 CCE page

I hope you found this glimpse at some of the records interesting.  As always your comments are most welcome.  Last week’s post was about our short term strategies.  If you haven’t seen it, you might want to take a look (Copyright Records: Short term strategies for making them more accessible).