Page:Mind (New Series) Volume 12.djvu/535

FEEDEBIC W. H. MYEBS, Human Personality. 521 of the unwary to his assumption of their subliminal character. It is therefore necessary to criticise Myers* classification of hallu- cinations (veridical or not) as "sensory automatisms". In the chapter on " sensory automatisms," automatisms are defined as " messages from the subliminal to the supraliminal self ". Now the meaning of the term " motor automatism " is clear enough ; it denotes such more or less intelligent activities as automatic writing and speech, of which the characteristic feature is that the ordinary self remains unaware of the movements, and of the ideas expressed by them, unless it sees the writing or hears the speech. How then does Myers manage to bring the motor automatisms under this definition (as he does on p. 88, vol. ii.) ? For by so doing he makes it appear that hallucinations are due to the same class of process .as. motor automatisms, which are truly subliminal in the sense of being wholly outside ordinary consciousness. It is effected in a second definition of automatisms as " manifestations of submerged mental processes, which do not enter into ordinary consciousness " .(vol. ii., p. 82). On making this discovery I was overcome, not for the first time, by a feeling of admiration for Myers' literary skill. But my duty as critic compels me to point out that this seemingly successful assimilation of motor and " sensory " automatisms con- sists in an illegitimate manipulation of terms, and depends upon the ambiguity of the relative pronoun in the sentence quoted. If the word ' which ' relates to the noun, mental processes, then the definition includes not only motor automatisms and hallucinations, .but all other contents of ordinary consciousness, for, as we have seen, all states of consciousness (with the possible exception of states resulting from voluntary effort) are manifestations or pro- ducts of processes " which (the processes themselves) do not enter into ordinary consciousness ". The definition, in this case, fails to mark off automatisms from other kinds of mental state or process. If, on the other hand, the ' which ' relates to the noun, manifesta- tions, then the definition ceases to be true of hallucinations and most of Myers' " sensory automatisms," for they are manifestations that do enter into ordinary consciousness. There is much virtue in a name, and Myers' exposition exhibits an extremely skilful application of that truth, nowhere more strikingly than in this instance, but perhaps equally so in his definition of suggestion in the opening of the chapter on hypnotism as " a successful appeal to the subliminal self " ; and again in the pronouncement that telepathy is an impact of the " subliminal self " of one man upon another's, and in a hundred instances of his use of the word subliminal. Let us turn now to the third class of evidence for the reality of the " subliminal self," namely, supernormal phenomena. Here if anywhere must lie the strength of the case, for we have seen that the phenomena of other kinds that seem to call for the hypothesis are so few and call with so uncertain voices, that we can only be justified in applying it to their explanation if we find other and far better grounds for its adoption.