Some of you may already have seen the announcement from the Cornell Legal Information Institute about our new project for the Library of Congress, where we propose to build some new ways to access legislative information. If you stumbled upon the original announcement (I’m betting few of you did, except for the odd law librarian in the bunch), you’ve perhaps been waiting with bated breath for me to spill more of the details of this, which Tom promised I would in his announcement. I’ve been distracted by a few other things in the meantime (like ALA Midwinter) but figured I’d better fulfill Tom’s promise before I get too far into the project and forget what we were thinking about when we wrote the proposal.
As is usual for those kinds of things, we looked around at what other people were building in other jurisdictions and noticed that a lot of people were using FRBR to model legislative information, including the UK. (For more about that project see this blog post). This decision made no sense to me, in particular (I’m not sure how much my LII colleagues know about FRBR, but I’ve been immersed in it for a while now) and I was pretty adamant that we shouldn’t go down that road. It’s unclear to me what the reasons would be to adopt the FRBR model in a legislative context, but I could speculate that part of it is that there’s been a lot of buzz around FRBR in the bibliographic community, and if you’re trying to do something sooner rather than later, reusing something that’s already there seems attractive. The Library of Congress, in its solicitation, was strongly in favor of reusing not only FRBR, but also the standards they have developed over the years for bibliographic data. In the normal course of things, that all makes a lot of sense, but for us it made much more sense not to adopt an explicitly bibliographic model which worked reasonably well for literary works but not so well for the kind of shape-shifting that goes on in the life of legislation.
From the proposal:
“Traditionally, libraries have approached the question of incorporating specialized kinds of materials into their descriptive workflow by focusing on the similarities between the new materials and the materials for which they normally provide descriptive metadata. In the past, this worked well–materials in newer formats and those for use in special communities were able to be incorporated into existing tools with a minimum of fuss. In the area of legal materials, the treatises, standard monographic materials, and standard serial titles were in general easily incorporated, while loose-leaf services and other materials with updating services were not. Primary legal materials were treated either as collections, as serials, or, in the case of most legislative materials, as standard monographs. Now that the digital revolution is well upon us, with full text more available and users’ experience with search engines generating more pressure to look beyond the simple access to printed materials, we’re starting to see more clearly how limiting our traditional approaches have been.
There are several areas where the traditional bibliographic approaches fail:
* the model of ‘stand-alone’ monographs with few (if any) relationships are
insufficient to provide the functionality desired for specialized legal materials
* the new bibliographic approaches, such as RDA, are based on a FRBR model of published works, which, while rich in relationships, provide neither element sets nor
relationships particularly useful for primary legal materials, legislation in particular
* primary legal materials have traditionally been entered under jurisdiction with collective
uniform titles which are often meaningless to users
* insufficient distinction is made between jurisdiction and place”
Once we decided that we needed to start from the beginning on a model that worked specifically with legislation, and we thought about what kinds of materials we needed to cover, it seemed fairly clear that we had to think about an events-based model. Well sure, you say, but isn’t FRBR about events, too? Definitely, but those events have to do with traditional publications, not legislation, where events like ‘House vote’ constitute the kinds of events we need to think about.
One question that came up was whether this approach would end up building a silo from which legislative materials would never emerge to play well with related legal materials. One way we hope to forestall that possibility is to build the descriptions around these legislative events in ways that they can be reused in other environments, even bibliographic environments. Not too surprisingly for those of you who follow this blog, we’re talking about using some of the strategies for building the legislative data that we used to build the RDA Vocabularies.
We’re going to use the process defined by the DCMI Singapore Framework, which expects us to build use cases to figure out what people want to do with the data, functional requirements from those use cases with which to test our model, and a model that grows from those solid foundations. From there we will define description set profiles for our data, and, we hope, have something useful to talk about. I’m guessing we’ll be talking about it all along the road, if for no other reason that we’re very excited by this project. We think we’ll learn a lot and enjoy the process, so do wish us well with it!