This qualitative research study was an attempt to gain a deeper understanding of the interaction and use behaviors of performing musicians. Through qualitative data analysis of 25 musician interviews, and 193 musical scores (or parts), representing over 25,000 separate annotations, this project uncovered the motivations, necessary knowledge, and methods by which musicians annotate and thereby internalize the instructions set out in the score. The aggregate data from the interviews and data analysis provide the basis for understanding annotation's utility for future users, and the development of more robust and useful music digital library tools and systems. This research also has the more general and theoretical applicability for those interested in how humans interact with interpretable, structured, largely symbolic and formalized information. Bringing together practical applications and theoretical concerns, the findings generated by this research will have wide effect in the Information Science community.