322 
Hermogenes agrees, but would fain know what is 44 the natural 
fitness of names.” Why does dog mean 44 runner ” ? But this 
is the actual mystery of language, and Sokrates cannot help 
him except by a few ingenious general suggestions. 
Kratylos, in the opposite extreme, not only holds that names 
44 are natural and not conventional,” and that 44 there is a truth 
or correctness in them,” which I do not doubt ; but also that 
one name is no better than another, that all names are rightly 
imposed, that if a man addresses you by a name not your own 
44 the motion of his lips would be an unmeaning sound, like the 
noise of hammering at a brazen pot,” and 44 that he who knows 
names, knows also the things that are expressed by them ” ; for, 
44 as is the name so is also the thing ” ; and on being pressed 
with the argument that 44 if things are only to be known through 
names how can we suppose that the givers of names had know- 
ledge before there were names at all,” takes refuge in the 
supposition that 44 a power more than human gave things their 
first names, and that the names which are thus given are 
necessarily their true names.” This latter position we have 
already found reason to reject on its own merits; and the view 
that 44 a word is either the perfect expression of a thing, or 
a mere inarticulate sound,” is, as Prof. Jowett observes, 44 a 
fallacy which is still prevalent among theorizers about the 
origin of language.” So far from a name being perfect, it is 
obviously imperfect ; inasmuch as it gives an incomplete view, 
which itself naturally corresponds with an experience only 
partial and a defective apprehension. 44 Bunner” is a good name 
for a dog, so far as it goes ; but evidently not an absolutely 
good name. But there being thus an element of imperfection 
in names, there are therefore degrees of imperfection, so that 
one name is better or worse than another ; and therefore all 
names have not an equal degree of truth or appropriateness. 
It is, then, absurd to regard names as god-bestowed. 
Again, we do not, by knowing names, know the things that 
are expressed by them. Thus tesem raises no idea of the dog 
in our minds. Yet we are willing to admit with Kratylos that 
he who bestowed this name had a reason for so doing. But if 
all names are equally valuable, and indeed divine, so that Javan , 
kalb, tesem , lik, and x are perfect names for dog (if, indeed, 
there can be more than one perfect name, as, says the adage, 
44 Mortals have many tongues, immortals one ”), then the result 
is precisely the same as if these terms were merely valueless in 
themselves, i. e ., had only a conventional value. In either case 
we could understand nothing about them, except that we had 
them. Thus these two opposite systems, starting from the 
same point in different directions, traverse the world and meet 
