323 
again, having in their course described two semicircles, which 
combined give us — nothing. 
6. The Platonic Vieiv of Language . 
Between Hermogenes and Kratylos stands the Sokrates of 
the dialogue, who freely communicates his “ view,” or perhaps 
“ views,” rather than any definite theory. Language has un- 
doubtedly a natural element ; names are appropriate and not 
arbitrary. But there is also a conventional element; thus, e.g ., 
shall (Old Eng. sceal , “ belongs to ”) is now generally employed 
merely as a tense-symbol. There is, moreover, the element of 
accident, the element of the effect of time, the element of the 
desire of euphony, and (as regards Hellenic considered alone) 
the foreign element. Granting that language is the imitation 
of something ; imitation, like convention or any other human 
effect or arrangement, is comparatively feeble and imperfect. 
And here it may be further observed that human imitation is 
second-hand, i.e,, we reproduce the impression which the parti- 
cular circumstance has made upon us. This line of idea is 
most true, and excellent so far as it goes ; but having nothing 
except a priori brilliance to support him, Sokrates starts the 
singular theory that “ the work of the legislator is to give 
names, and the dialectician must be his director if the names 
are to be rightly given ” ; so that “ this giving of names can be 
no such light matter.” We who know that all languages are 
more or less related, and that simple primitive man was the 
great maker-of-terms ( poietes), are, of course, aware that the 
appellation kwan was not the product of the joint reflective 
wisdom of dialectician and legislator ; unless, indeed, the 
simplest thinker be dignified with the former name, and the 
simplest ruler with the latter. But Plato sees with luminous 
clearness the real crux in language, i.e,, “ What is the natural 
fitness of names?” “Names rightly given are [in some way 
or other] the likenesses and images of the things which they 
name.” We see, of course, how direct onomatopoetic (“ bow- 
wow ”) names answer this description. We do not, however, 
call a hound bow-wow, but dog ; granting that dog means 
“ runner,” we see how this name answers the description, but why 
does the root dhav mean “ to flow ” ? If language is “ imita- 
tive sound,” and if this fact be “ the greatest and deepest truth of 
philology,”* what and how does dhav (allowing it for argument’s 
sake to have been once used as a word) imitate ? Names could 
never “ resemble any actually existing thing, unless the original 
* Jowett, The Dialogues of Plato , ii. 192. 
