From a review of languages we may see that various methods of producing starters are used in different degrees. In some languages the phonetic methods receive greater application than the recombinations of existing starters; in some other languages the formation of accessory starters receives prevailing importance; and in some other languages the order of the starters appropriates the importance of the preferential method. But most of the existing languages use all methods; e. g., in the Chinese spoken language there are many starters which are used in accessory functions, like «affixes»; in the Mongol language A. D. Rudnev («Material for the Dialects of Eastern Mongolia'') has shown that the change of phonetic complex («flexion») is practised; the method of the «word order» is used by nearly all languages, etc. Since we know from the history of some languages that in the course of time various methods may receive greater or smaller practical application, we cannot say what the «original» language was as to its preferential method. In fact, there are some languages which are changing their preferential method under the observation of linguists; yet there are some other languages which preserve the original method for centuries with no essential change. The condition which is implied by the study of facts is that it is impossible to say that there existed in former times [37] a single-sided adaptation in certain groups of languages unknown in other languages.
Indeed, the present character of existing languages is a result of the long adaptation lasting for thousands of years, and in most cases we have no evidences for suggesting what they originally were. The preference for phonetic changes, or the using of accessory sounding starters, may originate, develop, and die out within short or long periods. Yet it may be preserved forever as a prevailing character. The same is true of the existing sounding starters. If there is no impulse of change, they may be preserved for hundreds or thousands of years; as, for instance, certain «gestures» which are known to big apes, monkeys [38], and all human groups in the same «meaning» as typical starters. Since the appearance of the physical possibility of producing elementary varied sounds, certain combinations might receive general recognition and use, just the same as coups de poing, or fire making, and they may persist up to our day as well. It is impossible to have any hope of restoring them, for we know nothing as to the physical possibilities of sound production in the early ancestors of man. In the course of time they might change thousands of times. These reconstructions are not less artificial and dangerous than the reconstruction of the linguistical types.
There are starters preserved from «prehuman» ancestors, there are starters invented on various occasions and for various needs, there are starters borrowed from various sources: from the primary milieu — various sounds of «nature» and animals — and from the interethnical milieu. They are modified according to the easiest manner of their reproduction, or they are adopted as they are perceived. This complex is again modified under the influence of changing phonetic systems. The latter sometimes spread their elements over certain territory as any other ethnographical element which can be easily adopted when needed or desired. Sometimes the whole phonetic system is gained, little by little, by new-spreading fashions. Yet the methods of producing complex starters («phrases,» «sentences») may also change under the influence of neighbours who show a change of prevailing method. So the languages may also be composed of various phonetic and constructive methods, the origin of which may be traced back to different periods and sources. As an entity, it may persist for a very long time with no change at all, and it may also change any day. It may be borrowed as an entity and in its particular elements, and it may disappear altogether.
The fact that there are different ways of adaptation and preferential methods for the formation of starters is in agreement with the phenomenon of internal equilibrium of a once well-adapted ethnographical complex. Such a complex, owing to its utility, has a certain definite function in the whole cultural complex, so that, together with other complexes, it is transmitted to the succeeding generations again as a complex of elements, i. e., element by element, till the complex is ready for its function, as it is with other complexes of the secondary milieu. During the transmission, various conditions making the transmission of the complex more or less difficult are involved. Practical ways of learning and of teaching a language is one of the important features of the process of transmission of the complex. On the other hand, imitation by children and their ability of observation and reproduction is another important feature of the same function.
However, during the process of transmission, the complex may be slightly modified. The modifications constitute the history of language. But it is different when the complex is adapted by an ethnical unit which has previously lost the adopted complex. In such a case, the whole entity may be modified. The language is not thus the same, for it is modified; and it is not of indigenous origin, for it has been borrowed from another ethnical unit. When it is transmitted through a series of generations and a new generation comes into contact with the ethnical unit from whom the language has been borrowed, it may produce its influence, which will be an «alien» one, on the continuing language, and so on. Indeed, we may imagine an ideal case where a language has spread over a certain territory, together with the bearers; as, for instance, where a language spreads over territory with an ethnical unit which multiplies itself with great rapidity and occupies new territories, at the same time destroying other ethnical groups and gradually changing the phonetic system, structure, and basic elements of the language. But practically such cases are extremely rare. Yet the first question will be about the bearers of this language.
37. To assert or even to suppose it is a dangerous hypothesis, for it implies other logical inferences without being itself an established fact. Yet, all the above-outlined methods are so simple that their existence in the languages of human predecessors is admissible as well. Then, the problem of their origin and sequence has to be brought from the field of facts to that of hypotheses. We do not know the needs of the ancestors of the early «unhuman» man, and we do not know what were their physical conditions for the production of sounds. However, we know that the man of the middle quaternary, who was not «man,» did possess the complex idea of a soul (the practice of burial) and a well-developed brain. The lack of facts regarding his predecessors is not a reason for denying the possibility of a still earlier appearance of the need of a complex language. The attempts at the restoration of this complex in its «primitive form» is a mere speculation, satisfying a mind worried about the »'unknown.» Yet the reconstruction of the gradual discovery and prevailing methods in the formation of starters is also artificial and dangerous in its nature.
38. Recent investigations on gorillas (Yerkes) and chimpanzees. I have my own observations on monkeys.