This week, I’ve been thinking about how MARC records deal with non-book formats. It’s important for a library catalog user to be able to see what kind of resource they’re looking at before they actually search for it in the stacks: any given title could be for a book, an audiobook, a CD, a DVD, some microfiche, an atlas… (maybe not any given title…) Regardless, MARC records have three fields to describe an item’s content, media, and carrier (fields 336, 337, and 338, respectively). Books are described as containing text (“336 $a text $2 rdacontent”), requiring no special equipment to use (“337 $a unmediated $2 rdamedia”), and having a book-like physical structure (“338 $a volume $2 rdacarrier”). Each field uses controlled vocabulary from an official RDA list, which is cited in the $2 subfield, making it possible to use other Thesauri to describe materials if there are better vocabulary choices.

And indeed, e-literature might need its own official Thesaurus to more accurately describe e-lit works. The UCLA Library catalog has an entry for the first two volumes of ELO’s Electronic Literature Collection (as well as a CD-ROM version of volume one), but it doesn’t do a sufficient job in describing the resource. The collection as a whole has a “text” content type, a “computer” media type, and an “online resource” carrier type – which essentially just describes a website. A user can only see what works and authors are included in each volume by exploring the information provided on the resource website itself.

While I couldn’t find any examples, the MARC 347 field allows catalogers to input Digital File Characteristics, such as File Type (text file, audio file, video file), Encoding Format (PDF, CD audio, Blu-Ray), File Size, and more. (This information could also be put in a Field 300 Physical Description Note under subfield $b (Other Physical Details). RDA’s File Type vocabulary can be used for audio, text, program, image, data, and video files; however, this might become complicated when it comes to describing multimedia files.

Persistent Identifiers

On a slightly different note, I’m surprised that I haven’t been able to find Digital Object Identifiers (DOIs) attached to many works of e-literature. DOIs would allow specific works to be findable despite changing formats, migrating to different websites, or having multiple copies, which is important as so much of the original software used to access these works has become obsolete. Users would be able to click on the DOI and be taken to whereever they can access the work currently, regardless of the URL.

Eastgate systems, which sells the hypertext software Storyspace, uses ISBNs as identifers for the earlier, groundbreaking works of e-lit that is sold on its website. These works are sold on physical USB sticks, which either include the software needed to access the work, or recommend the user find a legacy computer that still has software like HyperCard on it. However, the ISBN isn’t useful as a persistent identifier for a digital resource, since clicking on it does not forward the user to the work’s webpage.

The Interactive Fiction Database offers two types of identifers for both internal and external use. Each work has a randomly-generated 16-digit string called a TUID (Tads.org Unique Identifier) that is used to organize entries within the database, as well as a IFID (Interactive Fiction Identifier). The latter is a community standard for IF bibliography established in 2006 by The Treaty of Babel - “an agreement between active design systems, the IF-archive, and other interested parties.” (Source) IFIDs can be generated as new works are published, or applied to legacy works, making it useful for finding works that might not otherwise be a part of a system.

I will be looking into IFIDs, as well as the IFDB’s organization structure, in another blog post soon.