I attended the NISO webinar entitled, “Metadata for Managing Scientific Research Data” presented by Jane Greenberg. I was pleasantly surprised by the presentation. The slides are freely available and are helpful for getting a hold on metadata in general and more specifically for research data. I especially enjoyed Jane’s focus on the function of metadata. It wasn’t so long ago that I was still in library school taking courses on cataloging, metadata and digital librarianship. The focus of these courses was to investigate the type of national standards available to catalogers and metadata librarians. My cataloging course focused on AACR2 and MARC21 for bibliographic records. My metadata course emphasized learning about the 15 core elements of Simple Dublin Core, a little bit of MODS, VRA, MIX, and the rest of the metadata standards ecology. I would have to summarize this approach as learning the different types of metadata. Metadata standards were broadly and generally organized into 4 or 5 types: administrative, descriptive, rights, technical, preservation. Sometimes, structural metadata is added to this list, sometimes not. This type of organization can be easily seen on many catalog/metadata library department websites such as Harvard’s LTS page, MIT’s Metadata Services page, Getty’s “Introduction to Metadata” or NISO’s “Understanding Metadata“. A few take this organization to a more detailed level to describe the types of data such as data structure, data value, data format, data content or data preservation standards. A great reference that includes this presentation is Steven Miller’s Metadata Reference page. What I noticed about the material in my courses at school and these resources is that metadata is often referred to in the singular and the focus is on type and not function of metadata standards.
As my institution began looking into Fedora for our digital repository, I realized that understanding metadata in the singular and by type was insufficient. The FGSC CSGDM Workbook helped me towards this understanding that metadata are data that can be expressed and understood in the variety of functions they fulfill.
Coming back to Jane’s presentation, she also began with a quote from the FGDC workbook. She added a quote from Greenberg (2002, 2003, 2009) that is worth replicating here: “[Metadata defined:]Structure information about an object (data) that facilitates functions associated with the object.” Jane’s summarizes this emphasis on functions with her chart on slide 10. Typical functions of metadata are: discover, manage, control rights, identify versions, certify authenticity, indicate status, mark content structure, situate geospatially, describe processes. Notice that Jane is combining types and functions in this list.
Going back to the FGDC workbook, metadata or the structured data used to describe the how, when, why, what, if, and who of a resources have “uses” or functions in regards to the structured data: organization and maintenance of information, providing information for a variety of purposes such as clearing houses or harvesting. These terms are from the FGDC workbook.
Expanding on this, I have a number of directions that help me think about metadata, of which I’ll consider two here.
1. Data: Increasingly I have found it easier in certain circumstances to refer to metadata as data. One, with the push to respond to scientific data and the acknowledgement of the deluge of data, referring to metadata as data reminds people that metadata “are”. Using the plural helps to think of the number of ways information about a resource can be organized depending on the function or functions that this information has to serve. Two, striking meta from metadata has helped removed some of the negative associations at least in my institution that metadata are AACR2, ISBD, or MARC21 standards. I’m not saying that these standards have nothing to do with metadata. However, referring to just data has allowed to conceptualize more than just textual data. This slight change has helped me better communicate with our digital library folks and programmers. This might seem a minor concern. But when you need to talk about data structure it is extremely helpful to talk about the variety of structures used such as real numbers, integers, text, or dates.
2. Functions of Data: I have found it helpful to place a stronger emphasis on the functions of data. A good starting point is to think about the how, what, when, why, who, and if of data. Like Jane, data serve the functions of discovery and management. I would add searching, identification, distribution, or reference. NISO’s “Understanding Metadata” also talks about functions in terms of resource discovery, interoperability, or digital identification. A comment from the recent ALCTS e-forum on gathering statistics in technical services helps to understand functions of data. Paraphrasing this comment was that for statistics to have a value, there must have been a prior planning process where you ask why and determine how the data are being used. This process is also necessary for data. It is essentially to ask the questions of how the data are being used and by who (which can include users and computer systems) before even determining the types of data standards needed for the project. Otherwise it doesn’t matter what standards you use in the best digital platforms.
What do you think are the functions of (meta)data? Is it useful to talk about data and not metadata?