On Deciding Cases of Disputed Authorship

Summary The subject of this paper is the problem of authorship decision based on vocabulary studies. Because of its importance to statistical as well as to linguistic methodology, the basic work by Mosteller and Wallace (1964) on the Federalist Papers is discussed and compared with three Swedish contributions. The problem of authorship decision was developed by Mosteller and Wallace as one of Bayesian hypothesis testing. We shall relate their approach, and certain extensions of it, to the concept of errors of the first and second kinds as conceived by the classical school. The list of references includes works on authorship decision based on other criteria than vocabulary characteristics.