Subject access: from card catalogs to online catalogs, with a look at Artificial Intelligence

Abstract

Building on Karen Coyle's observations about the role of semantic metadata in online catalogs and the loss of their organizing function, the paper examines the intellectual ferment that characterized the transition from card catalogs to online catalogs in the field of subject access. It explores the challenges, expectations, and risks associated with the adoption of new technologies, their implications for modernizing indexing tools, and their impact on catalog design.

The historical context from four decades ago bears striking resemblance to contemporary developments in the field, as institutions now explore the potential application of artificial intelligence technologies to enhance metadata generation processes.

The paper concludes with an examination of artificial intelligence, considering the conditions and boundaries within which it can operate effectively, and the constraints that must be addressed when designing AI applications.

Between the initial reference to Coyle and the final examination of AI, the paper traces significant developments in catalog evolution and design regarding subject access. It discusses Pauline Cochrane's 1984 lectures on Modern subject access in the online age, including expert proposals she solicited on this issue; the excitement surrounding this transition in Italy, as reflected in Luigi Crocetti's presentations and the 1985 Trieste Conference on Subject indexing and automation; and the design principles proposed by Marcia Bates during this period.

Rather than presenting a mere historical overview, the paper offers a critical reflection on key issues relevant to modern catalog design: the complexity of subject access; the effectiveness of concepts such as ‘uncertainty’, ‘variety’, and ‘redundancy’; their relationship to traditional notions of ‘pre-coordination’, ‘specificity’, and ‘summarization’; and the interaction between vocabulary and technology.