Information categorization approach to literary authorship disputes

Scientific analysis of the linguistic styles of different authors has generated considerable interest. We present a generic approach to measuring the similarity of two symbolic sequences that requires minimal background knowledge about a given human language. Our analysis is based on word rank order frequency statistics and phylogenetic tree construction. We demonstrate the applicability of this method to historic authorship questions related to the classic Chinese novel "The Dream of the Red Chamber," to the plays of William Shakespeare, and to the Federalist papers. This method may also provide a simple approach to other large databases based on their information content. (C) 2003 Elsevier B.V. All rights reserved.