My dissertation investigates the use behaviors of musicians interacting with their primary information source, the musical score. Although there have been a number of recent studies exploring the information seeking and management behaviors of people interacting with music retrieval systems, there have been very few published studies of interaction and use behaviors of musicians themselves. The absence of this research is unfortunate, because the musical score is both semantically rich and highly formalized, and therefore has special properties that make its study profitable for general theories of information behavior, interaction and use.

These properties include:

1) Musical scores are symbolic and notational. This type of information, which also characterizes dance scores and architectural drawings, often supports teamwork and collaboration. Analysis of musicians’ interactive behaviors might augment work in the CSCW community, and would help in advancing to the next generation of collaborative Internet technology

2) Musical scores are highly structured and formalized. Studying the users who naturally and regularly interact with this kind of data, albeit in a non-digital form, may prove valuable for refinement of other, digital systems that deal with structural data, like GIS and statistical programs.

3) Preliminary findings show that the musician’s interaction with the score, as characterized by their annotations of that score, is itself relatively symbolic and structured. While symbolic and formal annotations are narrowly applied in the analog realm, exploration of musician behavior from this point of view might provide a viable model for digital annotation technologies.

Additional research interests include representation and meaningful organization of complex objects in digital repositories for the purpose of retrieval and comprehension. These objects can be singular, like images and sound files, or multiple, like archival collections. There are four main areas of interest, all interrelated:

1. Digital Preservation: Digital materials are inherently unstable: How do we preserve them? Because of the variable nature of digital materials, do we need to decide which characteristics are important for preservation? How do we make those sorts of decisions? How do we know which characteristics to save? Can archival concepts like reliability and authenticity support meaningful digital preservation? How do we apply those concepts, which have a rich history related to physical materials, in the digital sphere? What intellectual traditions is current research drawing upon? Are there any other intellectual traditions that might augment or improve current practice?

2. Archival Concepts of Context and Provenance: How do traditional archival principles like context and provenance transfer to the digital sphere? What are the specific differences between the concepts of context and provenance? Can archivists and librarians leverage these principles to represent primarily non-textual collections for more reliable and comprehensive retrieval?

3. Intellectual Access to Visual Information: How are images currently described in digital systems? What role does subject analysis play in the retrieval of images? According to the extant user studies, how do users think about images when performing retrieval tasks? Is subject analysis, as it is currently practiced, the most effective way to represent non-textual objects? What traditions are current researchers drawing upon? Are there other representation techniques or intellectual paradigms that might be more effective for access and retrieval of non-textual objects?

These three topics are profoundly inter-related. It’s impossible to think about digital preservation without thinking about adequate representation, especially for non-textual objects. Likewise, it’s impossible to think about digital archiving and archival representation without thinking about the archival concepts of context and provenance.

There’s a fourth area of interest, which is the method by which I intend to answer some of my questions:

4. Annotation Studies: Why do people annotate any given “text?” Do annotations represent a meaningful interaction between annotator [user] and annotated object? If so, what can be inferred from an annotation? If not, what are the circumstances under which an annotation could be considered meaningful? What role does the document’s primary use or purpose play in inferring meaning from a given annotation? Given a set of annotated objects, is it possible to characterize important or difficult sections of those objects through annotation patterns? If so, what can we do with that information? What value does an annotation add to the annotated artifact?

---