Metadata can be a sticky subject; when I bring it up in conversations about digital scholarship, I’m often met with mistrustful looks. I think it’s because the word is used to describe so many concepts in so many contexts, and that can be confusing and frustrating. The Oxford English Dictionary defines metadata as “data that describes and gives information about other data”. In the field of library science, the Library of Congress defines three types of metadata: administrative, descriptive, and structural. It’s ironic that the word itself causes confusion, because the purpose of metadata is to make information more usable. In fact, it’s the underlying structure of digital scholarship; the bones that support digital access and archiving. Here, I’ll be focusing primarily on descriptive metadata – contextual information like names, places, times, etc.; I’ll discuss its role in oral history and explain how to easily and simply integrate it into your project from the second you start interviewing. By doing this, you’ll save yourself a lot of unnecessary work, and as a bonus you’ll make your librarians and IT collaborators very happy.
Central to oral history is the descriptive metadata created during the interview process, which often happens without anyone thinking about it. An interview might begin:
“This is Jenna Nolt, interviewing Ric Sheffield about his Community Within oral history project on the 4th of April, 2016, in her office in Olin Library.”
This basic opening is packed with valuable metadata, which will make the interview accessible, searchable, and preservable. Let’s borrow the metadata-tinted glasses of a librarian and look at the sentence again:
“This is Jenna Nolt [name of interviewer], interviewing Ric Sheffield [name of interviewee] about his Community Within oral history project [topic of interview] on the 4th of April, 2016 [date of interview], in her office in Olin Library [location of interview].”
The bracketed information can be used to create points of access to the interview through names, topics, geolocation, and chronolocation. Software can make your transcripts full-text searchable (so you can Google a phrase from an interview and get the transcript in the search results), and create maps and timelines to contextualize the content. Even if all you plan to do with your interviews is store them on a hard drive (not recommended without backup), it will make your interviews easier to find as you process them, and more retrievable in the future. I’ll discuss software choices and storage options in-depth in later posts; for now I’ll focus on the interview itself.
Here’s another way to think about beginning an interview; try to include what journalists refer to as the five Ws:
- Who the interviewee(s) and the interviewer(s) are
- What the name of the project is
- When the interview is taking place
- Where the interview is taking place
- Why the interview is taking place (i.e., the purpose of it)
This simple step ensures that the basic context of the interview is never lost. However, you can take it one step further with the OHLA Metadata Form which I developed for OHLA as a result of working on prior oral history projects that would have benefited from it. It is intended to be filled out right after the interview, and once the A/V files are transferred to a computer, this form should be kept with them throughout the archiving process. It serves as a quick reference for the files, insures that proper nouns are spelled correctly, and remains accessible in the worst-case scenario that files become corrupted, so that there is at least a record of what they were. Asking people to spell out names, place names, etc. during an interview can also be helpful, but might be awkward and break the flow of conversation. Jotting down this information in this form after the recording is easy enough, and helps prevent the possibility of misspelled terms, which can significantly decrease the accessibility of an interview.
Descriptive metadata is helpful regardless of the goals of a project, but it is essential if the material is intended for long-term use; for example, if it will be included in an institutional repository. I will discuss institutional repositories in more detail in a later post; in short, they are online platforms supported by an institution to promote and preserve the scholarship of that institution in digital form. Different institutions have different requirements to include scholarship; not all repositories accept oral histories, and not all institutions have a repository. Basic descriptive metadata is a fundamental requirement of all institutional repositories; it is necessary to make the material usable. Check with your liaison librarian to learn more about what is available from your college or university, and what requirements they have. At Kenyon College for example, we have Digital Kenyon, which is built and maintained by the library.
Metadata also comes into existence the moment someone hits the record button, whether or not they realize it. Recording devices automatically generated metadata such as file and folder names, time stamps, and information about file format and size. Later, information like modifications to the files and date of access is also automatically generated. A basic awareness of this metadata will help prevent technical questions from slowing or derailing a project. Below are some common examples of questions I’ve come across, which good metadata practices help with:
Q: I have multiple versions of a file, and I don’t know which is the original (or which is the most recent).
A: Check the metadata encoded in the file; right-click on a file and select “Properties” for a PC, control- click and select “Get Info” on a Mac to see file size, date created, date modified, and last date accessed.
Q: What software do I need to play/edit this file?
A: Record in standard formats, such as AIFF or WAV formats for audio files, and AVI or Quicktime Movie formats for video files.
Q: How do I know which files are what?
A: Rename files to be human as well as machine-readable as soon as you transfer them from your recording device. See Stanford’s guide to best practices for more guidance on file naming.
Metadata isn’t an absolute; any is better than none, and more is generally better. If you forget to include the five Ws at the beginning of an interview, or use the Metadata Form, you can always go back in and add information later. However, building simple practices into your process will streamline the technical side, leaving you free to focus on your content.