Recently I attended the Boston Library Consortium Networking event. This is a one day event which brings together members of the BLC. It’s a fun day to see colleagues, network, and participate in the number of working groups or interest communities of the BLC. As part of the event, there are the lighting talks which are always exciting and a way to learn something new. This year’s can be found here.
I would like to highlight the one on Data Architecture presented by Greg Colati and Patrick L. Carr. This was a shorter presentation of the one they did at CNI earlier this year. This is really an important endeavor. In short, it melds curatorial thinking based on data architecture to library acquired digital content. The question is how we manage strategically and systematically our digital collections in terms both of information and data management. I encourage to read through the presentations to get a better sense of their work. I’d like to jump forward to their six principles.
This gets to the heart of how do we systematically and strategically think about digital content. This is important in terms of how budgets are spent or what resources are dedicated towards digital content. In other words, this could help to more efficiently and effectively plan and manage library resources now and in the future.
What’s missing from these principles and event both presentations is metadata. I think there’s a reason for that and let me try to explain why. The focus of these principles is digital content that has been acquired by the library. The library has said yes to managing this digital content. Once you have said yes, then that meaning of yes brings with it a curatorial analysis of the digital content. What are we managing (is this a research data set and/or publication for instance)? Is the digital content something the library wants to invest in long or short-term? Based on the results of this decision tree, the appropriate technology to store and manage the digital content is chosen along with the methods of discovery and appropriate discovery systems. Again, the missing elephant in the room is metadata.
How is the decision tree possible? Just like in research data, the curatorial process asks for a README text file or a data management plan. This process provides contextualization for the digital object (or metadata). From there, it sets the stage for storing, managing, discovering, and accessing the digital content. These are services provided by the appropriate systems and its features. In others words, these are services provided by the system as set up by the decision of saying yes to the digital content and the curatorial process of what yes means for the library. To use a phrase from my former supervisor, Roger Brisson, these are metadata services. What is missing from these 6 principles are metadata services.
However, are the metadata services really missing? In this context, the digital content is a static entity (final version) that is to be managed by the library in some shape or form. The metadata used to make storage, management, discovery, and access possible is far from static. Metadata are dynamic and have a lifecycle. Metadata form an ecosystem that is evolving. In this context of the 6 principles, I see metadata as that which form the neural networks joining these principles and others together. These neural networks are changing and growing. A good example is what people create through the use of APIs when they remix and re-use data from DPLA for instance.
What are the implications of this line of thinking? For one, metadata is dependent on technology. I’m not saying that we need computers and fancy programs to have metadata. That technology can be a pencil or a stone tablet. What I’m saying is it is necessary to have the tools to create these metadata and those tools are continuing evolving. So Yes! We need to be smart or scope our metadata to be adaptive and reactive to technology. If we want any metadata services, it is crucial to ensure that our metadata are still alive and doing the job they’re supposed to be doing of helping us store, manage, discovery, and access content efficiently.
As a result, metadata needs to be part of this curatorial conversation. Unfortunately, that discussion on metadata is curiously absent in these presentation. The problem with it being absent is that we overlook the complexity and delicate nature of the “neural network” aspect of metadata. We do this at our own peril. Because once we let metadata be forgotten or we take advantage of the fact that our systems will just create metadata that is meaningful for us now and in the future we risk losing our digital content because that metadata can no longer support the services needed.