Page:Aristotle (Grant).djvu/85

 will still remain, and is still always incident to human reasoning. And this it is which Aristotle undertakes to classify. It might be thought that errors in reasoning were infinite in number, and incapable of being reduced to definite species; but this is not the case, because every unsound reasoning is the counterfeit of some sound reasoning, and only gains credence as such. But the forms of sound reasoning are strictly limited in number, and therefore the forms of fallacy must be limited also. Ambiguity in language is, of course, one main source of fallacy; and fallacy arises whenever either the major, the minor, or the middle term of a syllogism is used with a double meaning. It will be seen above that the quibblers in ‘Euthydemus’ employ the terms “wise,” “learn,” and “know” in double senses so as to cause confusion.

Aristotle’s account of the fallacies attaching to syllogistic or deductive reasoning is complete and exhaustive, and has been the source of all that has subsequently been written on the subject. The fallacies of amphibolia, accidens, a dicto secundum quid ad dictum simpliciter, ignoratio elenchi, petitio principii, consequens, non causa pro causa, and plures interrogationes have become the property of modern times, with names Latinised from those by which Aristotle first distinguished them; and in Whately’s, and other compendiums, they may be found duly explained. It is true that Aristotle does not investigate the sources of error attaching to the inductive process; the “idols of the tribe” and “of the den” he left for Bacon to denounce; and the fallacies of “inspection,” “colliga-