Since these animal cries have purposive significance similar to that of language, in what respect do they differ from language? I shall try to give a definite answer to this question.

First, they are inarticulate while language is articulate. And what precisely does ‘inarticulate’ mean? It means (1) that these cries are not explicitly articulated sounds with clear-edged beginnings, middles, and endings, as are the word-sounds ‘h-a-p’ and ‘h-o-p,’ or ‘d-i-g’ and ‘d-i-p,’ so that one sound can be clearly and definitely differentiated from another; and (2), as a corollary to this, the sounds have no conventional meanings; that is, they are not invested each with an arbitrary and definite connptation quite apart from any natural sound-suggestion which they may have, so that one sound stands exclusively for one thing, another sound for another thing, as do the words of language like ‘wolf and ‘bear.’ In other words, the animals’ cries are natural cries as distinguished from the conventionalized sounds which we call words, and they have the characteristic vagueness or indefiniteness of significance which all natural sounds have. In this respect of natural vagueness of significance the cries of animals resemble the mechanical sounds in nature, though differing from these in expressing purpose.