totalities. The third class is due to the influence of language upon thought (Idola fori). The formulation of words is governed by the needs of practical life, but exact thought frequently requires distinctions and combinations which differ widely from those of common speech. In certain cases there is a superabundance of words, in others there are too few. The fourth class (Idola theatri) is ascribed to the influence of traditional theories.
We must get rid of all these illusions. Bacon makes no attempt to show how this may be accomplished. The conception of the idola tribus contains a profound problem which Bacon failed to see, a problem however which acquired vast importance at a later period; we are obliged in every case to interpret reality from the human standpoint (ex analogia hominis); but in that case the question arises as to how our knowledge of the world can possess objective validity.
Bacon takes exception to the prevalent method of induction on the ground of its being limited to positive cases (as an induction per enumerationem simplicem). He insists that we must likewise take note of results in cases where the phenomenon under consideration is absent. He demands furthermore that we investigate the modifications of phenomena under varying conditions. After sufficient material has been gathered by these methods—and in order to avoid being overwhelmed by the confused mass of facts (for, citius emergit veritas ex errore quam ex confusione)—it is necessary to formulate a tentative hypothesis and examine the cases which seem to establish or refute the hypothesis. Bacon’s method is therefore not a pure induction. He has a presentiment of the profound mutual dependence of induction and deduction.