The formal hierarchical structures of organizations can be expected to affect the organizations' behavior in many different ways. One kind of impact is that their structures may affect how their knowledge is organized. This is important because the different ways in which knowledge can be organized may affect what the organizations' decision makers learn from the information that their organizations have gathered, stored, and processed. However, rigorous testing of this general argument would not be easy, and so it would be useful if a preliminary means of assessment could be found so that we can better judge whether a full-scale test is warranted. Hammond (1993) noted that a library catalogue is also a formal structure for hierarchically organizing knowledge, which means that different kinds of cataloguing systems represent different ways of organizing knowledge. Two different kinds of cataloguing systems — the Library of Congress classification and the Dewey Decimal classification — are widely used in US libraries, and Hammond conjectured that these two different cataloguing systems will tend to bring different sets of books to the attention of library users; this in turn should be expected to have consequences for what the library users can most easily learn from the library. This article tests Hammond's conjecture in two university libraries, focusing on 40 classic books in political science. The results provide strong empirical support for the conjecture: although the two cataloguing systems do not appear to organize knowledge in completely different ways, what differences they do have nonetheless appear to have a striking impact on what users can most easily learn. Support is thus provided, albeit indirectly, for the general argument that how organizations are structured hierarchically should be expected to affect what organizational decision makers are able to learn.
