A Question of Semantics
In “The Semantic Web in Action,” Lee Feigenbaum, Ivan Herman, Tonya Hongsermeier, Eric Neumann and Susie Stephens describe the development of the Semantic Web, a set of formats and languages to find and analyze data on the World Wide Web easily. The problem with this system is that different people will not agree on exactly how to define all concepts. Any computer application that tries to standardize its ontology will necessarily distort what at least some people are trying to express.
I am also concerned that in conventional formal logic, if even one inconsistency exists it will be possible to draw all possible conclusions and their contradictions!
Robert W. Jones
Emporia State University
FEIGENBAUM AND HERMAN REPLY: Discrepancies among ontologies do occur, but the Semantic Web does not rely on having one, all-encompassing ontology. Instead it is built up from small, like-minded communities that can find agreement on terms among themselves. Applications can therefore interact without attempting to achieve global consensus: a system that presents a retailer’s wares to customers will harvest information from suppliers’ databases (themselves likely to use heterogeneous formats) and map it onto the retailer’s preferred ontology before being displayed to customers. Automated tax-return software takes bank data, conforming to individual banks’ ontologies, and maps them onto the tax form. There is no requirement for global ontologies: instead an application need only map the terms relevant for a particular transaction into a common vocabulary. Of course, although agreement need only be local, adoption of existing vocabulary still facilitates data sharing and integration.
As to the dangers of using conventional logic, “inference” in the Semantic Web can be characterized as discovering new relations. But the inferences on this network are done within a restricted, “guarded” subset of first-order logic. Ontological reasoning on the Semantic Web does not use the full power of first-order (or higher-order) logic and therefore avoids some of the dangerous conclusions that can come from an inferred inconsistency.
180 Degrees of Aberration?
In his profile of Hugh Everett and his many-worlds interpretation of quantum mechanics, “The Many Worlds of Hugh Everett,” Peter Byrne states that Bryce S. DeWitt “swung around 180 degrees [on Everett’s theory] and became its most devoted champion,” thus implying that previously DeWitt had a negative opinion of the theory. In fact, DeWitt promoted Everett’s work from the beginning. DeWitt was the acting editor for a section of the July 1957 issue of Reviews of Modern Physics containing proceedings from a physics conference that he and I had organized in January of that year (he was not the regular editor of the journal). DeWitt decided to include Everett’s thesis in the proceedings even though Everett had not attended the conference. DeWitt had read the thesis at John Archibald Wheeler’s request, but Wheeler had doubts about a work questioning Niels Bohr’s understanding of quantum mechanics. DeWitt found it “new and refreshing.”
A few years later DeWitt discovered that Everett’s paper had “slipped into instant obscurity” and resolved to start a publicity campaign. This campaign was not a 180-degree swing in his position but a continuation of his 1957 interest in Everett’s work.
I am not writing as Bryce’s widow but as Bryce S. DeWitt’s colleague—a colleague who knew how to disagree with him when necessary. I wish to correct a misleading presentation of a historical fact.
Jane and Roland Blumberg Centennial Professor, Emerita
University of Texas at Austin