334 
tional cries (of which more anon), they may or may not be 
words, but how much of the dictionary comes from them? 
Have they not, as far as we know, been almost always stationary 
in number ? Theirs must at most be a minimum of influence ; 
and this fact Prof. Sayce fully admits, since, as he says, they 
represent emotions, not ideas. But if the other two and the 
chief elements in the formation of language be gesture and 
onomatopoeia, how can it be said of words generally, at all events 
in their origin, that there is no necessary connexion between an 
idea and the word that represents it ? Noticing that natural 
sounds strike different ears very differently, he excellently 
remarks : — 
“ Of course, it is not necessary that the imitation of natural 
sounds should be an exact one ; indeed, that it never can be • 
all that is wanted is, that the imitation should be recognisable 
by those addressed. Many of our modern interjections, like 
alas! [ = a (ah !) -f Zcts (wretched, Lat . lassus)\ lo, are words 
that once possessed a full conceptual meaning, but have lost 
their original signification, and been degraded to the level of 
mere emotional cries. So hard is it for language to admit 
anything ivhich was not from the first significant in thoughts* 
Therefore, the mind, we notice, positively demands signifi- 
cance in word-making ; but significance excludes the arbitrary 
element, and if men formed of language on gesture and 
onomatopoeia, we again find that the means practically employed 
negatives mere arbitrariness. Thus from any and every point 
of view we arrive at the conclusion that language in origin is 
not arbitrary and conventional. But although these arguments 
may be fully admitted in the abstract, yet we falter when con- 
fronted with the crowd of utterly variant words ( e.g ., Icwan , etc.). 
The fact that people attempting to imitate the notes of the 
nightingale, produced forms as various as bulbul , jugjug , 
whitwhit , tiuu-tiuu , etc.,t shows at least that there are cases in 
which highly different results may attend the attempt to express 
verbally the same thing ; but I make no suggestion that the 
four dog-names are variant phases of a prior and original name, 
in the same way as numerous Aryan dog-names are variant 
phases of hwan . 
As our method is strictly comparative, let us in the attempt 
* The italics are mine. 
+ The cock is referred to in the Avesta as “the bird named Parodarsh, 
which evil-speaking men call by the name Kahrkatds ( V endidad , xviii.). 
Cf. our cock-a-doodle-doo , the Tahitian aaoa , the Yoruban koklo, the Zulu 
kuku, the Finnish kukko, the Sanscrit kukkuta , the Spanish quiquiriqui , the 
Chinese kiao-kiao, the Mantchu dchor-dchor, etc. 
