The complexity of metadata

Since I graduated from my MS in information science, I have heard some pretty outrageous things about metadata. One of the worst ones was this sentence: “monkeys can do cataloging.” The context for this remark was that anyone even a monkey could do cataloging. Frankly I didn’t ask the speaker to elaborate that time because I had heard the comment before unfortunately. The context is that metadata and cataloging are seen to be simple tasks that do not require advanced degrees or knowledge. The comment was meant to illustrate the total simplistic nature of metadata and cataloging. But I should elaborate. In this context and when I’ve heard the pronouncement of simply task in reference to metadata, this task is often equated to data entry. In other words, there are some (far too many in librarianship and related fields) that definite metadata and cataloging as data entry. So there is only one task, namely entering data into fields.

This is so far from the truth. But before looking at that, it would be beneficial to evaluate data entry and whether or not it is simplistic. I don’t believe data entry is simplistic. It is difficult to enter consistent and accurate data day in and day out. In fact, companies pay good money for competent data entry staff. What does it mean to be consistent and accurate? That determination is made by shared interoperable standards. These standards change with technology and current trends. There are also questions of how to apply these standards in a variety of systems that might not be as friendly as we want with these standards. Already, the “simple” nature of data entry is no longer that simple.

For metadata/cataloging, there are certainly tasks that require data entry. It is important to understand the shared standards and local best practices. It is not just one standard that we follow. There are a number of different standards that are created and managed. It is typically not just 1 system that we work with but often several. For example, I work with Alma, Primo, our data harvester (repox), BePress, Fedora/Islandora, Omeka, ArchivesSpace. But data is not just managed in our systems. Data come from a variety of external sources such as vendors, OCLC, spreadsheets, etc. I’ve had to go fetch data from Flickr and random MySQL databases.

The complexity of our jobs is first and foremost knowledge and implementation of a variety of standards. It is also being able to work in a host of systems. It is also dealing with data. The following question comes up frequently in listservs. Do people create separate records for print and electronic? Of course, some might say that we’ve long ago answered this question with provider neutral records. However, in Alma, there is the possibility of implementing a more linked data approach to records. In this universe, it is possible to have a work record and then manifestations that are print, electronic, and digital. Yet, Alma is not quite linked data ready. The bibliographic record still has information that spans the WEMI spectrum. To prepare for linked data, should we consider taking advantage of Alma’s ability to attache print, electronic, and digital records to the bibliographic? This question alone is far beyond data entry. It goes beyond standards. It involves specific systems and how these systems are evolving.

So no, cataloging and metadata is not so simplistic that you can do it with your eyes closed. If it was, this would have all been automated years ago.