200 On the Analytical and Synthetical Modes of Reasoning 



notions of another kind, they often find themselves deceived, 

 notwithstanding all their exertions to preserve rigour in the 

 form of the demonstrations; a thing which no one will dis- 

 pute their knowledge of. 



It is therefore less in the method than in'the simplicity 

 and evidence of first notions, that the certainty of reasoning 

 consists; and with regard to general principles, which Con- 

 dillac always speaks of with merited contempt, they have 

 no place in geometry. That of the least action, which 

 Maupcrtuis made so much noise about, has only been re- 

 garded by mathematicians as an analytical result, arising 

 from the central laws of mechanics; and it has never been 

 cxactlv dellned but by the assistance of mathematics ; for it 

 formerly assumed very different forms in the hands of me- 

 taphysicians. 



The second and third articles of the Pensees de Pascal ap- 

 pear to me to contain what is most luminous upon the man- 

 ner of reasonintr; and I do not perceive that Condillac has 

 made any improvements lo them. Already Pascal had per- 

 ceived the abuse of definitions, and reduced them to their 

 just value, that is, to descriptions and impositions of names; 

 "but far from proscribing any method of reasoning, as has 

 been done latterly, by calling that synthesis which was but 

 the abuse of rcast)ning, he classed the different methods of 

 treating the sciences in such a manner as to show the as- 

 sistance which might be derived from each of them. 



It is possible, says he, to have three principal objects in 

 the discovery of truth : the first, to discover it when we seek 

 it; the second, to demonstrate it when we possess it; and 

 the last, to distinguish it from false when we examine it. 



In fact, these three cases are able to present themsel,ves ; 

 the first evidently almost always takes place; it likewise 

 sometimes happens that tb.c analogy of circumstances makes 

 a proposition suspected, and then we endeavour to assure 

 ourselves of its existence by a formal demonstration. Lastly, 

 if wc wish to submit a proposition to examination, in order 

 to discover its truth or falsity, it is useful to know the ge- 

 neral means for fulfilling this end. 



With reuard to the exposition of accjuircd notions or 

 known truths, tlic only rule to be observed as often as pos- 

 sible consists in comparing them in those parts in which 

 they have the greatest comiicetiou, and where the fewest in- 

 termediates are required. 



Bv calling, as is most proper after the etymology of the 

 ■wor(ls, that svnthcsis, by means of which we proceed fron)^ 

 simple to coujpositc ;and that analysis, which returns from 



composite 



