Our main motivation for formalizing contexts is to deal with the problem of generality in AI. We want to be able to make AI systems which are never permanently stuck with the concepts they use at a given time because they can always transcend the context they are in. Such a capability would allow the designer of a reasoning system to include only such phenomena as are required for the system's immediate purpose, while retaining the assurance that if a broader system is required later, ``lifting axioms'' can be devised to restate the facts from the narrower context to the broader one, with qualifications added as necessary. Thus, a necessary step in the direction of addressing the problem of generality in AI is providing a language which enables representing and reasoning with multiple contexts and expressing lifting axioms. In this paper we provide such a language.
The goal is that no matter what corners the specialists paint themselves into, what they do can be lifted out and used in a more general context.
For an overview of the AI research on formalizing context see [1]. For technical papers on context in AI and Linguistics see the following special issues of journals: [48, 32, 14].