146 Walter Petersen, 



adjectives in which it designated an approximation to a state or 

 quaHty designated by a primitive adjective. 



10. The extension of the use of the suffix -isko- to form deriva- 

 tives from substantives could take place after it was felt as a 

 single suffix.i On the one hand there might be an adjective and 

 a substantive coexisting from one and the same stem as possible 

 primitives, e. g. *v£av6<; and veavia?, and veavicTxo?, though formed 

 from the adjective could be referred to the latter ; or perhaps some 

 word like the proper name ''h-^^iay.oi., an epithet of Athena, origi- 

 nally derived from an adjective, in this case aypto? 'field-like' 

 ' wild,' could be referred to a noun, e. g. aypo? ' field,' and xu- 

 v.\i(yY.ij<;, originally a synonym to xu/,>.tO(;, could have been referred 

 to vxivXoc,, etc. There are, however, very few if any cases where 

 the derivation from such an adjective is more natural than from 

 a substantive, and yet both are possible, and we may assume that 

 the chief cause of the transfer of the suffix to the substantive 

 primitives was semantic syncretism. When e. g. an I. E. adjective 

 like O. H. G. altisc meant ' rather old ' (§ 7), -isko- was identical 

 in meaning with -ko- in words like Skt. babhru-ka-s ' brownish ' 

 i. e. ' rather brown, ' or vsavicTKO? ' he who is rather young ' would 

 have an -isko- exactly like the -ko- of Skt. rohitaka- (: rohita-s 

 ' red ') ' that which is rather red,' the name of a certain tree.^ 

 Since -ko- was also widely used to form derivatives from substan- 

 tives, -isko- followed with ease. 



11. Partially by the same process of syncretism -isko- took on 

 to itself widely divergent meanings in the different languages, but 

 at the same time it could on its own accord pass through a se- 

 mantic development which was much Hke that of -ko-, which itself 

 started from the meaning ' belonging to the category of ' ' being 

 like,' and in some languages, e. g. Skt., ended by having a large 

 variety of meanings.^ Consequently we may often not be sure 

 whether a certain use was influenced by simple -ko- or not, nor is 

 a common usage in different languages by itself a guarantee of 

 common origin. Thus -isko- may be deteriorative in Greek, Slavic, 

 and Germanic, but the Greek deterioratives, like the other 'di- 

 minutives,' seem to be due to syncretism with -ko-, since their gender, 

 like that of the -ko- diminutives,* follows that of the primitive, 

 while the Slavic neuters must be due to a secondary development 



^ Cf. Brugmann, op. cit. 503. 



2 Cf. Edgerton, JAOS. 31. 118. 



3 Cf. Brugmann, op. cit. 503 f.; Edgerton, JAOS. 31. 96 f. 

 * Cf. Brugmann, op. cit. 669. 



