Georeferencing: The Geographical Associations of Information (review)

In 2005, to the consternation of certain members of the international security community, Google launched its Google Maps and Google Earth applications. Both are examples of what Linda L. Hill calls georeferencing, or relating information to geographic location. In Google's case, the associations are straightforward: text-based queries specifying place names, physical addresses, or geographic coordinates link users to maps or aerial images. (Hill calls such representations "information object types.") Google's interface is forgiving. Its geocoding protocol is flexible and it can interpret nonstandard street addresses and even misspellings.

As nifty as Google Earth may be, Hill is far more ambitious. A leading figure in library science and the development of "knowledge organization systems," she sees georeferenced information everywhere (on the masthead of this journal, for example, which indicates an impressive degree of geographic dispersal, or in the note at the end of this review, which suggests that I am writing in my office at the University of Pennsylvania), but she laments the primitive ways in which users can presently sort and utilize it. For the past decade, she and her colleagues have been working to standardize the way such information is recorded and stored. Their goal is to produce a "unified georeferencing system" that will allow users to work across dissimilar collections and platforms and to go beyond text-based queries: "users should be able to start with what they know, whether a place-name or a geospatial footprint, and create a query to find georeferenced information from a variety of library catalogs, data centers, museums, archives, directories and web-based resources" (p. 8).

As may already be clear, this admirable book is not a work in the history of technology or STS. Instead, Hill has drawn on more than three decades of professional experience to write a primer for information workers—practitioners of library science, information science, museum informatics, and geographic information science. Nevertheless, Hill's volume will be useful to historians specializing in computing, information technology, cartography, or any of the increasing numbers of fields that rely on geographic information systems.

Hill begins with an introduction to "spatial cognition" in which she borrows models from cognitive psychology in order to suggest the theoretical attributes of an ideal geographic information system. A basic dilemma emerges. Human beings regularly operate with a high degree of geographic "vagueness," while computer information systems rely on "crisp" boundaries. As a consequence, coding geographic references, whether textual or visual, is fraught with challenges and compromises. Sociologists and historians of technology will find the meat of the book in chapters 5 and 6, which detail the operation of "gazetteer services" and the brief history of competing "metadata standards." Although the relative merit of the Dublin Core Metadata Initiative versus the Darwin Core Metadata Georeferencing Elements may appear arcane, the material here offers insight into as-yet unsettled controversies within the field of geographic information storage and retrieval. Together, these chapters begin to open up the black box of computer-based georeferencing and may guide active researchers to key actors and conferences.

Standards and categories matter. Historians of technology have explored their construction and consequences with regard to subjects ranging from the thread of the American screw (Sinclair) to the American system of manufacturing (Hounshell). Historians and philosophers of science have investigated issues as varied as cartography (Alder, Graham), the history of statistics (Hacking, MacKenzie, Desrosières), and life in the laboratory (Latour and Woolgar, Shapin and Schaffer, Kohler). Standards in computing and information technology have earned less attention, in part because they comprise recent history, but perhaps also because of the collective, corporate nature of computer engineering. Janet Abbate's Inventing the Internet (2000) demonstrates the far-reaching importance of protocols such as TCP/IP, however, as does the excellent work that Geoffrey C. Bowker and Paul Edwards have done on cyber-infrastructure. Historians would do well to think about the nuts and bolts of such technology, since the decisions that Hill and her colleagues make will determine—visually...