260 ADVANCEMENT OF LEARNING 



a neglect of many useful things relating to it. We, there 

 fore, think proper to advance a substantial and capital doc 

 trine of method, under the general name of traditive pru 

 dence. But as the kinds of method are various, we shall 

 rather enumerate than divide them; but for one only 

 method, and perpetually splitting and subdividing, it 

 scarce need be mentioned, as being no more than a light 

 cloud of doctrine that soon blows over, though it also 

 proves destructive to the sciences, because the observers 

 tliereof, when they wrest things by the laws of their method, 

 and either omit all that do not justly fall under their divi 

 sions, or bend them contrary to their own nature, squeeze, 

 as it were, the grain out of the sciences, and grasp nothing 

 but the chaff whence this kind of method produces empty 

 compendiums, and loses the solid substance of the sciences. 1 

 Let the first difference of method be, therefore, between 

 the doctrinal and initiative. By this we do not mean that 

 the initiative method should treat only of the entrance into 



1 The design of Ramus, whose method of Dichotomies is here censured, was 

 to reduce all divisions and subdivisions to two members, with a view to obtain 

 a basis for the construction of dilemmas and disjunctive syllogisms. We are 

 never certain that these species of reasoning are legitimate, except when the 

 divisions out of which they rise are exact; and the only test of this accuracy 

 is to be sought in a dichotomous contradictory division, where the supposition 

 of one member necessarily leads to the exclusion of the other. This method of 

 exhausting a subject by an analytic exhaustion of its parts, which he mainly 

 derived from Plato, has its proper sphere in logic; and though condemned in 

 the text, was employed by Bacon in many of his prerogative instances. The 

 error of Ramus consisted in taking only a part for the whole of logic, and ap 

 plying what is strictly applicable to subjects of a peculiar nature, to the whole 

 range of inference. It is evident, however, that the dichotomous process can 

 only be employed in the investigation of subjects which admit of a twofold con 

 tradictory division, and that where the primitive elements are composed of four 

 or five distinct members, the method is totally inapplicable. Its use, therefore, 

 ought to be attended with the greatest caution, as the Ramist can hardly be 

 certain that the twofold division, in many cases, is not more apparent than real, 

 and that a further analysis would not necessitate a multiform classilication. 

 For want of this foresight, Ramus, with all his subtilty, falls into inconceivable 

 errors, and a great many of Bacon s exemplifications of his method in the cru 

 cial instance are direct paralogisms. Milton framed a logic on the model of 

 Ranms s method, seduced rather by the bold antagonism of the latter against 

 Aristotle, than by its philosophic justness. Both the original and the copy are 

 now forgotten, and Ramus is committed to the judgment of posterity rather on 

 his absurdities than his merits. See Hooker, i. 6, with Keble s note. Ed. 



