Abstract

In On the Internet, Hubert Dreyfus notes that by moving documents from libraries to the Internet we make ourselves dependent on search engines to locate the information we need. Because search engines are incapable of understanding the semantic content of documents, he suggests that we risk losing access to the information we archive online. I examine the strengths and weaknesses of the strictly hierarchical libraries that Dreyfus prefers and conclude that there are lines of inquiry that such rigorously-structured hierarchies actively resist; namely, they resist questions dealing with relationships between objects and questions dealing with aspects of objects that are secondary to the hierarchy's branching principle. In other words, there are good reasons to move documents from hierarchical libraries to the unstructured Internet. I then discuss the salient features of a search engine that could make relevance judgments of the sort Dreyfus claims are impossible.