The methods of producing starters as stated are numerous. In man, at certain moments, human or prehuman, the sounding starter has received special application. However, so far as the operation with the sounds is concerned, it presents rather limited possibilities; namely, production of sounds by the lips, the tongue, the nose, the teeth, the soft and hard palate, the glottal complex, and rarely the cheeks, also all physically possible combinations of these organs. The combinations are not very numerous. For producing new varieties of fundamental sounds these may be increased by length, by tones, by sequence of tones, and by distinct degrees of stress; and lastly, by the combination of various sounds into new complex sounds. The subsequent addition of sounding starters permits a variation of complex reactions (in hearer and speaker) and gradually directs a series of subsequent reactions. The direction may be produced by various means; as, for instance, special starters, the order of different starters, etc.
The formation of stable, conditioned reflexes responding to the starters and their observation is the way to find a practical value (meaning) of various sounding starters and their practical use, as «engrams»[29]. It is naturally a long process and it is not particularly «human.» So the sounding starters, by the side of other methods, have become a powerful means of social intercourse, owing to which sounding starters might be greatly increased up to the point of forming stable complexes. The differentiation of sounding starters into that used for the phenomena directly percepted, that directing the process of connecting conditioned reflexes in new combinations, and that producing starters only in the complex combinations is a long process, in which the sounding starters (also their written symbols and, generally, their optical starters) went through the variations which at different moments might correspond to what is called «sentence,» «word,» and «particle»[30]. However, these variations are not one-sided, i. e., the sounding starter may first correspond to a «sentence,» and afterwards to a «word»; and a simple conditioned reflex perceived as «thought» may be produced by a simple starter first, but afterwards it may require a complex starter. Yet the starters may be preserved for thousands of years without any changes.
In so far as the reactions of hearers are concerned, the increase of combinations of starters produces new effects. They may become very complex. Beginning from a simple sound and going through a gradual increase and changes, the complex starter may attain the length of a process lasting for several seconds. A simple complex sounding starter isolated from a compound starter sometimes loses its functional value altogether. Thus the discrimination of nouns, verbs, words, sentences, etc., is, when made, sometimes absolutely impossible and artificial. Such a proceeding with the languages is a, mere adaptation of one ethnographical complex for another. In the choice of ways for finding new starters and their combination, individuals are often led in the line of least resistance, as it is characteristic of all other phenomena of human adaptation. If a new starter appropriates its function as a useful starter, it is adopted and comes into use. If the method invented for rearranging starters for producing complex starters is practicable, and a directing starter is effective, they may receive general recognition. Their nature is, indeed, a functional one. It may thus be stated that all methods are good in so far as they have a practical functional value and the methods are different [31].
29. This is the term used by Semon, defined by C. K. Ogden and I. A. Richards (op. cit., p. 140),—«to call upon excitation similar to that caused by the original stimulus.» The engrams may be dissected into elements.
30. The distinction of «sentences,» «words,» and «particles» is conventional, indeed. It is sometimes impossible to distinguish these formal elements. E. Sapir asserts that words exist as an entity somewhat independent of the sentence, for proving which he brings some facts of his experience with the American natives. I can support this statement by similar instances of the Tungus, whom I taught to use a phonetic transcription for recording their own language. However, these facts must not be overestimated, — in some languages, the words are entirely fused together to form the «sentence,» which becomes a complex starter beyond which words do not exist; yet in some languages, the «meaningless» particles used only fcr directing the connexion of conditioned reflexes are consciously used separately as «words.» The starter may be as small as a single sound and it may be as long as a long sentence. Children often perceive sentences (J. Piaget's observations in «The Language and Thought of the Child») as single starters which may be easily «explained» by starters of various length. But it does not mean that the unit is a «sentence,» and not a «word» or a «particle.» The sentence may be a single starter in which words cannot be abstracted, and a compound sentence may form a compound starter in which single words may be separated. It all depends upon the character of the starters.
31. The history of the Indo-European languages shows instances of loss of tones as an ancient element of the formation of starters, while the Chinese languages show a phenomenon of their probable increase. The old special starters (flexions and suffixes) of the Indo- European languages were lost, after which new methods were introduced (prefixes), while «flexion» and «suffixes» continue to survive. As these facts are well known, there is no need to repeat them here.