Rethinking the Fallacy of Hasty Generalization

Abstract

This paper makes a case for a refined look at the so- called ‘fallacy of hasty generalization’ by arguing that this expression is an umbrella term for two fallacies already distinguished by Aristotle. One is the fallacy of generalizing in an inappropriate way from a particular instance to a universal generalization containing a ‘for all x’ quantification. The other is the secundum quid (‘in a certain respect’) fallacy of moving to a conclusion that is supposed to be a universal generalization containing a ‘for all x‘ quantification while overlooking qualifications that have to be added to the more limited kind of generalization expressed in the premise. It is shown that these two fallacies relate to two different kinds of generalization.

The classification of fallacious generalizations is based on a new theory of generalization that distinguishes three kinds of generalizations – the universal generalization of the ‘for all x’ type, used in classical deductive logic, the inductive generalization, based on probability, and the presumptive generalization, which is defeasible, and allows for exceptions to a general rule. The resulting classification goes beyond a logic-oriented analysis by taking into account how a respondent may oppose a potentially fallacious generalizing move by falsifying it. Using a dialectical interpretation of premise-conclusion complexes, the paper outline a richer concept of generalizing argument moves embedded in a communicational reconstruction of the strategic uses of such moves in which two parties take part in an orderly dialectical exchange of viewpoints.