ANSS Subject and Bibliographic Access Committee
Question/Answer on Cataloging Issues – January 2020
Question: What is traditional library name authority work, and how might ISNI affect it?
By Isabel Quintana, Harvard Library, Harvard University
Librarians have always worked with some aspects of identity management – that is, the act of figuring out if the person who wrote your book is the same entity as the person who wrote other books. In traditional library work, when a cataloger receives a book, he/she usually ascertains if the library has other materials by the entity responsible for the work, and if this entity already exists in an authority file. In other words, I would want to see if I had any other works by the same author. I would also want to see how the author’s name had been set up in the past, so that I could collocate the materials.
For example, I receive a book titled “How to love a pug puppy” by Olivia Quintana. I search in my catalog and realize that I have another book titled “How to love a lab puppy” by Olivia Quintana. I may assume these are the same person. Then I want to know how to enter this person in my catalog. I notice that the book on labs has this access point:
Quintana, Olivia, 1989-
I also notice that there is a name authority record (i.e. a catalog record that records the form of name to use to collocate all materials by one person) for this person that is set up as “Quintana, Olivia, 1989-“ So I use that same form of name on the catalog record for my book. In some libraries, if there is no name authority record, then the cataloger will create a name authority record.
Continuing the same example, I then receive a book titled “The beauty of Scotland” by Olivia Quintana, published in 1988. In this case, I know my author has to be a different person from the person who writes about puppies, since the Scotland book was published before the other Olivia was born. I search in the authority file but find no authority record for my entity. In some libraries, I would then create an authority record to distinguish the two Olivias. In some libraries, I would just enter the name on my book record differently from the other Olivia Quintana to differentiate them. For example, I might use:
Quintana, Olivia (Writer on Scotland)
This is how librarians, performing traditional library work, collocate works from the same author, and differentiate works by different authors. Our name authority files are carefully maintained to ensure that each name string (i.e. official form of name) represents a unique entity. This work can be very time consuming, because we have to first identify the different entities, and then also construct a unique name string to identify each entity.
Also, even if we do this work for all authors in our local library catalog, the authors may also be listed differently on their wiki pages, or have name authority records in other databases. For example, an author from Spain may have an authority record in the catalog of the Biblioteca Nacional de España, and his form of name may be set up differently. Furthermore, there are e-journals with articles with many authors. Most of these authors names are not under authority control. In short, once we are looking at information on the web, that is coming from many sources, it can be quite difficult to tell if a name string refers to the same entity or not.
This is where an ISNI can be helpful. ISNI stands for International Standard Name Identifier, and it is ISO standard 27729. According to the home page of the ISNI website (http://www.isni.org/ ), “ISNI is the ISO certified global standard number for identifying the millions of contributors to creative works and those active in their distribution, including researchers, inventors, writers, artists, visual creators, performers, producers, publishers, aggregators, and more.” An ISNI works as a standard numerical identifier for an entity, just as the ISBN is a standard numerical identifier for a book. Usually entities described in ISNI are persons, but ISNIs can also be assigned to corporate bodies.
The purpose of ISNI is not to create yet another identifier for the same person, but to create a bridge identifier for that person. ISNIs serve as a link between different sources, because ISNIs are used by numerous libraries, publishers, databases, and rights management organizations around the world. The ISNI record for an entity contains many different identifiers for the same person. The ISNI record can then be used to link the correct entity across multiple library catalogs, even if the name string is slightly different.
Also, where library authority files are added to manually, ISNI is primarily expanded by batch. Once a batch of data is loaded into ISNI, complex matching algorithms add data to existing entities, or create new ones. Of course, there is always manual clean-up work to do. So ISNI has a quality team located at the Bibliothèque nationale de France and the British Library. According to the ISNI website, “This team evaluates new data sources, identifies strategic databases to add, reviews statistical samples, and corrects records flagged for manual review, as well as responding to end user input.”
There was a pilot program by the Program for Cooperative Cataloging (PCC) on how libraries can best use ISNI created in 2017. According to the program’s wiki site https://wiki.lyrasis.org/display/PCCISNI/PCC+ISNI+Pilot+Home the program was set up to “explore the mutual benefits of PCC contributors engaging in identity management activities within the ISNI database.” Here are some examples of what the program explored:
- Incorporating ISNIs within library bibliographic and authority records
- Using ISNIs within our institution repositories
- Exploring how ISNIs can complement and enhance faculty research platforms, such as VIVO
My library, Harvard University, participated in this pilot, and one takeaway I had was that creating a unique name string, although a vital part of authority work, was not a crucial part of identity management. Librarians will always be faced with the challenge of differentiating and collocating entities and their works; but we may not need to worry as much about how we set up the names. In fact, our current rules are sometimes not that helpful in finding the correct entity, because our current rules give a preference to birth dates (and many times the patron doesn’t know the birth year of the author), and to field of study, which can be described many different ways.
For example, if I’m looking for the person who recorded Thriller and I’m in a library catalog index, I would see this list of names like this:
Jackson, Michael, 1948-
Jackson, Michael, 1958-2009
Jackson, Michael, 1958 August 13-
Jackson, Michael, 1958 February 11-
Jackson, Michael, 1962-
Only one name has a death date, so that would help me to identify the correct person, but it’s easy to see that birth year is not the best identifier for most patrons to choose from.
Likewise, if the cataloger doesn’t know the birth year, we are allowed to qualify a name by any term that can differentiate that person, if there is already another entity with the same name in our catalog. So we end up with name strings like “Quintana, Olivia (Writer on Scotland).” This seems more straightforward, except that what if it turns out afterwards that Olivia is a travel writer and her later books end up being about other places. These are just two of the issues that librarians and patrons face when trying to differentiate entities by using a unique from of name.
ISNIs have helped those of us in the PCC Pilot to realize that the identity management tasks are critical, but the final form of name may not be. As long as the patron has a way to find the entity he wants, and then find other materials by that same entity, the catalog will serve his/her needs.
Furthermore, other cultural areas are also seeking to collaborate with ISNI. For example, rights management industries have a strong commitment to identity management, since it’s so important for them to differentiate entities correctly in order to distribute royalties accurately. The music industry also has a deep interest in identity management because they want folks to be able to correctly identify the artist when we hear a song, so that we will purchase more materials from that artist. These folks have less of a need to collaborate with libraries. However, if they are using ISNIs to identify entities, and we can use the same ISNI, our efforts will match up with theirs, and we can have better identity management on the web.
The PCC (Program for Cooperative Cataloging) is interested in exploring the use of identifiers in MARC cataloging. In the near future, catalogers will begin to experiment with the use of ISNIs and other identifiers in MARC cataloging. Our efforts will be documented, and analyzed, so that the PCC can come up with guidelines for catalogers on how to use identifiers effectively. We will never be able to provide full control over all entities on the web; but hopefully, in the future, our catalogs will better integrate with the web, and we will be able to provide better identity management for our patrons.