168 Walter Petersen, 



use within itself or more probably by semantic syncretism with 

 I. E. -ko-, which has this function e. g. in Skt. usrika-s ' miserable 

 bullock,' rajaka-s 'worthless kingling,' or in Lith. kuinukas 'nag,' 

 Slov. kravclca 'wretched cow,' travuca 'poor grass,' extended by 

 -lo- in Lat., e. g. sermunculus ' wretched talk, babbling.' 



45. Though the fact that the three genders of -itxo- are a qual- 

 ity which it has in common with I. E. -ko- makes it seem prob- 

 able to me that syncretism of the two suffixes caused the deteri- 

 orative use of the Greek, yet there are a number of words which 

 are on the border line between this meaning and its use to des- 

 ignate similarity. This, however, does not necessarily argue an 

 independent development from the latter, but these transition 

 types themselves may have been transmitted from -ko- by the 

 same assimilation of meanings, and they would consequently be 

 rather an indication of how the simple suffix became deteriora- 

 tive. ^ 



46. These transition types were words in which -igywO- was added 

 with the idea of mere similarity as opposed to complete identity, 

 and with the consequent notion that the primitive would really 

 be inapplicable, but the point of comparison involved or implied 

 was itself some point of inferiority of the derivative to the prim- 

 itive. Thus aiyiaxoq : caE, was an animal ' like a goat, but not a 

 real one,' because castrated. Cf. Hes. aiyto^ov • alya s/vT-ojjiav, 

 where there was no doubt as to the judgment of inferiority in- 

 volved. In the same way (iaaiUaxog (§ 35) : ^(xc>ikz6g as ' king of 

 a small country,' therefore ' not a real king,' i. e. not worth the 

 name, was on the border line of the deteriorative use. Similarly 

 ^ocrnXiayioc, is used of one who was too homely to deserve the name 

 ' king ' in Ath. 566 A yuvaixos auTw xcckfic, (patvojxsvY]?, kzi^xc, Be 

 atoypa? xai xXoucria?, 6)C, aTCxT^ivsv iizi ty]v x>;OLKriav, ^r;[j.i5iaoct -zout; 

 s(p6pou? auTov, ExdsyovTa? oti [^ao-dicTxou? avTt [3a(7i>.£wv tS Ixapira 

 yswav Tcpoaipslxat, Also dqansTiaxog (Luc. Fugit. 38) : Bpa7;£Ty](;, 

 ' one who is hke a fugitive, but not a real one,' because unsuccess- 

 ful, therefore ' a poor excuse for a fugitive.' Perhaps also Tqayl- 

 axog : Tpayo?, applied to a lame goat in Hes. (s^ayco jyikb^^ rpayi- 

 oxov • TuaiBiag slBo? Tcapa TapavTivoi?), belongs here,^ in as much as 

 its lameness caused the conception ' not a complete goat, but only 

 like one.' In all these cases it is to be noticed that the deteri- 



^ The fact that all of the apparent pattern types were post-Classical makes 

 it improbable that they were the real patterns. 



^ On the other hand it may be hypocoristic. cf. § 83. 



