During the course of our readings and discussion on metadata, one of the biggest challenges I have had is with keeping the distinctions between cataloging guidelines, schema, and vocabularies distinct. Part of the difficulty stems from the fact that terms which in broader domains are largely interchangeable like “expression” and “manifestation” become narrowly defined in information science and we are approaching these narrow definitions within a theoretical rather than an application-level context. I think another part comes from the phenomenon that Thomas Baker alludes to in his Library Hi-Tech article, that two or more of the above concepts are usually implemented in tandem and over time become intertwined. The separation of language from the specific data formats that RDF represents means that local frameworks for data can be unbound and knowledge can be more readily extracted .
The coherence of linked data does raise questions about how terms across ontologies are aligned. What is the careful review process that Baker refers to? As RDF vocabularies evolve with new data relationships, how will this affect the definitions to which the URIs resolve?