Date of Award
12-1990
Document Type
Dissertation - Restricted
Degree Name
Doctor of Philosophy (PhD)
Department
Philosophy
First Advisor
Lee C. Rue
Second Advisor
John E. Naus
Third Advisor
Michael Wreen
Fourth Advisor
A. Tallon
Abstract
This dissertation computerizes Aristotle's categories in the form of an object-oriented database to be used as a semantic substructure in more comprehensive cognitive science knowledge representation models. It is hypothesized that mind is produced through an organization of physical components, and so it is possible for machines to become mindful. Thus Aristotle's categories will be computerized in the hope of developing a more efficient way for machines to identify concrete physical objects. Knowledge representation studies originally concentrated upon logic and theorem proving. An overview of our currently most powerful logic, the predicate calculus, shows that theorem proving systems result in a purely logical taxonomy, which is necessary for we do reason. But it is concluded that logic alone is incapable of explaining how mind represents knowledge. More recent attempts like productions, semantic networks, and frames have added semantic procedures in the hope of explaining how mind represents knowledge. These endeavors have also proven to be insufficient to explain how mind works. From the above analysis it is concluded that all previous knowledge representational models have problems, and it is conjectured that these problems result from concentrating almost solely upon higher-level mental operations like discursive reasoning. Therefore this study proceeds to an investigation of lower-level, pre-discursive ways in which mind might organize knowledge, that is, Aristotle's categories are investigated and computerized in order to see whether they can be used to form a primitive semantic ordering for tracking the meanings of higher level terms. The practical job of computerizing Aristotle's categories begins with the computer science issues of language choice and object-oriented programming. Next it is shown how category theory applies to knowledge representation. Thirdly Aristotle's categories are normalized. Finally Aristotle's categories are computerized in the form of a database management system. This system, called ACOOD (Aristotelian Categorical Object-Oriented Database), is generalized thus allowing for the insertion and investigation of particularized object categorizations. Being general, ACOOD does not offer conclusive evidence that Aristotle's categories have furthered the attempt to computerize knowledge representation. But along with other evidence, ACOOD does support the hypothesis that effective categorization will eventually help us solve the conundrum of mind.