§ 8. Certain Other Theories of the Origin of Speech. — Attempts have been made to explain the origin of language without emphasising the importance of the visible gesture as the startingpoint. There are three main theories of this kind, which have been nicknamed by Max Muller the poohpooh theory, the bowwow theory, and the dingdong theory. Their more pretentious titles are the Interjectional, the Onomatopoeic, and the Pathognomic theories. The principle involved in all these theories is essentially the same. They all attempt to trace back conventional signs to natural signs; but they exclude from consideration visible gestures, and confine attention only to vocal signs. It is evident that to mimic the mewing of a cat, in order to convey the idea of that animal, is as much an imitative gesture as going on all fours and humping the back for the same purpose. It is mimicry of this kind on which the bowwow theory relies for explanation. The same holds good of imitating the cry of fear, in order either to convey the idea of the emotion or of the approach of a dangerous object. This is the sort of expressive sign which is most primitive according to the poohpooh theory.

The dingdong theory is more subtle, and it has the distinction of being advocated by Professor Steinthal. According to it specific kinds of objects so affected primitive man as to elicit from him, or to use Max Muller's metaphor, to ring out of him, correspondingly specific utterances. The most primitive words would therefore be phonetic types rung out from the organism of the first man or men when struck with an idea. There is a harmony of sound and sense which does not depend on the imitation of one sound by another. The charm of literary style and especially of poetry consists largely in the subtle affinity between vocal expression and the objects or activities expressed, which may exist apart from any resemblance of sounds to one another. The word zigzag is a good illustration. The zig goes this way, and the zag goes that way, thus describing a zigzag course. Again, take the line

"The vorpal blade went snicker snack."

The sound is expressive of the gleaming and rapid motion of the blade, rather than of the sound of it. What philologists call reduplication has often this intrinsic expressiveness, e.g. a "big big man"; a "wide wide sea"; "far far away." Among the Botocudos of Brazil ouatou stands for stream, ouatouououou is the sea.

In this metaphorical expressiveness of vocal utterance we may detect under a somewhat deceptive disguise the essential principle of the imitative gesture.

Even the disguise is not present in the case of reduplication; here more of the same kind of sound represents more of the same kind of thing. Other instances may look more mysterious. But the mystery to a large extent disappears when we consider that articulate speech consists not merely in articulate sounds, but also and as well in the motor process of articulation. The tongue actually does go zigzag in uttering the word zigzag. Ticktack imitates not only the sounds of the clock, but the rhythmic movement of the pendulum by a corresponding movement of the tongue. Even born deafmutes use the organs of articulation in this imitative way. Heinicke, as quoted by Tylor, mentions a "deafmute, nineteen years old, who had invented many writeable words for things." Some of these were arbitrary; but at least two, mumm for eating, and schupt for drinking, were, as Tylor remarks, an imitation of the movements of the mouth in eating and drinking. In like manner njan means to eat in the NegroEnglish dialect of Surinan, and njan njan means food.* Thus the dingdong theory is in its more obvious applications reducible to the general principle of the imitative gesture. That part of it which is not so reducible is of little value as an explanation of the origin of language. Vague and recondite affinities between sound and sense cannot in the first instance constitute a natural and spontaneous language, because they are not sufficient to make the vocal utterance selfsignificant or selfinterpreting. For this it is not enough that a word should be dimly felt to be appropriate when its application is already known. It is necessary that the sign should be so stamped with the character of the thing signified as to make clear its application in a given context and under given circumstances. On the other hand, it must be admitted that when once the meaning of a word has become a matter of convention a general feeling of affinity between sound and sense may operate powerfully in determining the creation and selection of new words.

* Tylor, Early History of Mankind, p. 73.

These and similar theories must all be regarded as part of the general doctrine that natural signs psychologically precede conventional signs. They are true and useful inasmuch as they emphasise the part played by phonetic elements in imitative expression. The imitative use of vocal utterance paves the way for the development of conventional speech. Why conventional language has come to consist almost entirely of phonetic elements we have attempted to explain in the last section. The reason why natural signs have to so large an extent been displaced by conventional signs lies in their superior convenience and power.