This paper presents a differential usage study of a web-based resource database that provides both search and associative browsing functionality. The associative browsing is based on emergent meta-data: meta-data that is derived from the terms that users associate with resources they have contributed to the system. We argue that this approach provides a low cognitive load information seeking mechanism, and can also reduce the effort required by the user to enter meta-data when contributing resources. In this paper we concentrate on a three-month study of student librarians using the system, with analysis of their activities and other data collected by questionnaire. The results suggest that associative browsing was at least as popular as search, and that providing perspectives on emerging meta-data during the contribution process may have helped the community self-organize a vocabulary.