Mark Twain once quipped that “All generalizations are false, including this one.” Twain's statement - probably never meant to be analysed logically - is an example of the Cretan or Epimenides Paradox where a statement is explicitly self-referentially refuting. If it is true, then it must be false.
Twain’s words introduce the subject that I want to talk about in this post, which is generalizations.
One of the criticisms I receive against my work on the conjunction of ideology and sociology (in particularly my articles on the Enlightenment HERE and HERE) is that I employ too many generalizations.
For starters, we all would have to agree that some generalizations are legitimate (when, for example, a term is distributed universally to all members of a class) while other generalizations, such as those underpinning prejudice and discrimination, are not. This being the case, it will not do to dismiss an argument simply because it employs too many generalizations without first doing business with the content of the generalizations in question. In this regard, it should not be overlooked (as I mentioned a few posts back) that a generalization does not have to describe all the members of a class distributively in order to be valid. As Doug Wilson points out in his book on courtship, generalizations are legitimate to the degree that they honestly describe an overall pattern and provide evidence to support their claims. Generalizations are consequently not refuted through particular and individual counter examples. When Jesus generalized about the sins of the Pharisees, he did not find it necessary to qualify his remarks to cover instances such as Joseph of Arimathea or Nicodemus. What we should ask from any given generalization is whether it is honest, fair and supported by evidence, not whether it is true in every given instances.
To join my mailing list, send a blank email to phillips7440 (at sign) roadrunner.com with “Blog Me” in the subject heading