The Rhciorica of Philodemus. 297 



[Obscurity also arises] from ignorance of the proper mean- h 159. col. 



ings of words, their connotation, and the principles on which 



one word is to be preferred to another. 



In addition to these there is a fault treated separately bv the I. 160, col. 



XVIII 

 theorists, namely the too frequent use of hyperbata, and failure 



to make the gap between the separated words short enough when 



it is necessary to use this figure ; and the separation of correlated 



conjunctions by too large an interval.*' 



One should use ordinary expressions appropriately, and not I^. 161. col. 

 express onesself inaccurately, nor vaguely, nor use expressions 

 with double meaning.^ 



They (the sophists) have not explained the' intricacies of 

 subject matter. [This belongs to the philosophers.] 



(A discussion on the choice of words has preceded.) The ^' ^62. col. I. 

 most important of the rhetorical sophists err in their too great 

 devotion to homoioteleuton and similar figures, and pay little 

 heed to the use of words. 



Collision of vowels is rather frigid, but sometimes not inop- I, 163, 

 portune. However they (the sophists) do not define each case ^^ ' 

 ( i. e. when it is to be avoided and when permitted), but they 

 depend entirely on subjective tests. ... If , then, the observa- j j^^^ ^^^ 

 tion of the principles laid down by them involves anything HI. 

 extraordinary, and there are present TrdOr) and T]6ri and the other 

 characteristics of artificial speech, I wonder if a satisfactory 

 form of expression has not been moulded from the vulgar speech. 

 This artificial speech they divide into three parts : rpoTros, a-xw"^^ 

 7rXd(TfjM • TpoTTos includcs metaphor, allegory, etc. ; crx^/xa, periods, 

 cola, commata and the combinations of these; irXda-fjia refers to 



donee perveniatur eo nt res cum sono verbi aliqua similitudine concinat, 

 ut, cum dicimus, aeris tinnitum etc. . . . Sed quia sunt res, quae non 

 sonant, in his similitudinem tactus valere, ut si leniter vel aspere sensum 

 tangunt. lenitas vel asperitas litterarum ut tangit auditum, sic eis nomina 

 peperit. . . . Lene est auribus, cum dicimus, voluptas, asperum est, cum 

 dicimus, crux. . . . Haec quasi cunabula verborum esse crediderunt, ut 

 sensus rerum cum sonorum sensu concordarent. Hinc ad . . . simili- 

 tudinem . . . processisse licentiam nominandi. . . . Inde ad 

 abusionem ventum est. . . . Innumerabilia enim sunt verba, quorum 

 ratio reddi non possit ; aut non est, ut ego arbitror, aut latet, ut Stoici 

 contendunt. 



'Cf. Arist. Rhet. Ill, 5, 2. 



' Cf . Arist. Rhet. II, 22, 12. 



