166 GREEK LANGUAGE 



effect of placing the accusative at the head of a sentence is not due 

 to any vigor of the case itself, but to the fact that its position an- 

 nounces a departure from the every-day arrangement of the parts of 

 the sentence. As "omne ignotum pro rnirifico," so " omne inusita- 

 tum pro grandiloquo." The subject accusative with the infinitive in 

 oratio obliqua is to me nothing more than a nominative. Analog- 

 ously, to my mind the difference between the genitive and the dative 

 is purely grammatical. In 6 Tran/p p.oi TC^VT/KCV, /ioi is not possessive : 

 it means "I have lost my father," just as in Cicero's single mention 

 of his father, " Pater nobis decessit a. d. mi. Kal. Decembres," " nobis " 

 instead of " noster " is the one note of feeling. The Pindaric flvyoY^p 

 ol (if ot is dative) is due to the predication involved in apposition, 

 a latent predication which may become active. The possessive 

 dative used attributively is a solecism, or rather a Colophonism. 

 When the case is a predicate, the distinction between ownership 

 and possession is purely grammatical. 'Ei/ra0a Kvpa> /SacnAeia fa . . . 



tori 8e /cat //.eyaAov /^atriAe'cos /3a<rt'Aeia ev KeAaivats : here the predicate 



dative and attributive genitive do not imply different kinds of posses- 

 sion. So TO. v KeA.au/aZs /JacriAeia eon Kvpov would not allow Kupo>. 



The rule, however, that if the subject has the article the genitive is 

 used, if not, the dative, is inadequate and does not get at the root of 

 the matter. A noun with the article may have the predicate dative 

 (Dem. 43, 52), and the genitive may be used when there is no article 

 with the subject. As this paper does not offer solutions of problems, 

 no attempt is made to state what seems to be the correct rule. The 

 ordinary distinction between " possession " and " ownership " is prob- 

 ably due to the fact that Ion Ku'p<j> may be rendered "Cyrus has." 



The problems of the cases have not all been solved. As yet the 

 cases have usually been treated separately, and for individual 

 authors or works, whereas they need to be treated conjointly and 

 comprehensively. To one point attention is directed. The prevailing 

 distinction between the accusative and the dative with the infinitive 

 after feo-Ti and Trpocn/Kei, though sadly muddled in some of our 

 standards, is theoretically plain enough; but what are we to make 

 of examples like Isocr. Paneg. 28, where it is said of a Aoyos that 

 has become /j-vOwS-r)? : O/AW? aur<3 Kai vvv prjOTJvai Trpoo-rJKei t Is this semi- 

 personification : "it deserves to be told"? A complete collection of 

 examples would be useful. 



With the cases the prepositions are intimately associated. Not 

 to mention the more general problems, the simple question of differ- 

 ent cases with the same preposition is often misunderstood, and we 

 find efforts to force the idea of motion into all examples of Trapa 

 with the accusative, or the view that Trapa of rest must take the 

 dative at least of a person. Here, by the way, style has its effect to 

 the extent that poetry has the greater privilege of being quaint. 



