The distinction between inductive and deductive forms of argument, as proposed by G. W. Ziegelmueller and C. A. Dause, is the focus of this paper. The first part of the paper reviews several grounds for distinguishing induction from deduction and explains that neither the traditional distinctions--those based on particular or universal premises, or probable versus certain conclusions--nor a more recent effort that depends on semantic clues unequivocally distinguishes induction from deduction. The paper then summarizes the theory of Ziegelmueller and Dause, criticizes it, and concludes that their definitions are unsatisfactory. The last part of the paper considers the following approaches open to the study of argument if it is not possible to distinguish induction from deduction: (1) to eliminate the supposed distinction and treat all arguments as deductive; (2) to abandon the terms induction and deduction and proceed to some theory of argument that does not depend on them; (3) to move away from normative models and use empirical descriptions of arguments; and (4) to try to rescue the induction/deduction distinction. (EL)