affective suggestion. This is indubitably the stage that precedes any other process, and on its existence the later processes depend. Secondly, there is the acceptance of the idea suggested, the process termed by Durand1 ideoplasty and by myself2 verbal suggestion. Thirdly, there is the ultimate effect realised by this idea after it has been incorporated into the personality.

I will now quote four of the most notable definitions that have been given of suggestion, and it will be seen that they fall into two groups, according as the main importance is attached to the second or third of these processes respectively. Bernheim3 gave the broad definition of suggestion as "l'acte par lequel une idée est introduite dans le cerveau et acceptée par lui." McDougall, with evidently the same point of view, has rendered this more precise in the statement that "Suggestion is a process of communication resulting in the acceptance with conviction of the communicated proposition independently of the subject's appreciation of any logically adequate grounds for its acceptance." In contrast with this attitude stands Janet's conception of suggestion as the "développements complets et automatiques d'une idée qui se font en dehors de la volonté et de la perception personnelle du sujet." Similarly Th. Lipps regards suggestion as "die Hervorrufung einer psychischen Wirkung, die normaler Weise nicht aus der Weckung einer Vorstellung sich ergibt, durch Weckung dieser Vorstellung" ("the evocation, by arousing an idea, of a psychical effect which normally would not result from the arousing of such an idea"), and he further insists that “nicht die Weckung der Vorstellungen, sondern diese weitergehende psychische Wirkung ist das Charakteristische der Suggestion. Diese psychische Wirkung ist das eigentlich Suggerirte" ("it is not the arousing of the ideas, but this further psychical effect, that is the characteristic of suggestion. This psychical effect is what is really 'suggested""). There can be little doubt that the emphasis laid here by Janet and Lipps on

1 Philips, op. cit. p. 44.

2 Loc. cit. The only exception to this is with Moll's Stumme Hypnose in which not a word is spoken, and this affords one of the many interesting transitions between heteroand auto-suggestion.

3 Hypnotisme, Suggestion, Psychothérapie, 1903 édition, p. 24.

4 Op. cit. p. 10.

5 État mental des Hystériques; Les Accidents mentaux, 1894, p. 30.

6 "Suggestion und Hypnose," Sitzungsbericht der bayerischen Akademie der Wissenschaft, 1897 (1898), S. 394. It is a matter for regret that this essay, doubtless because of its relative inaccessibility, is not more widely known, for it contains the most searching discussion of the subject yet provided by any psychologist.

Idem, op. cit. S. 392.

the further effects or action (Wirkung) of the suggested idea represents a definite advance on the intellectualistic conceptions of Bernheim and McDougall. Even if the latter would maintain that they too have in mind a psychical effect of the idea introduced, it is plain that their definition refers chiefly to one effect only, namely, disturbed judgement, and does not take into sufficient account the other abnormal effects, such as hallucinatory sensations, influence on bodily processes, etc.

Lipps made two further steps in the nearer definition of the psychical action or effect (Wirkung) in question. In the first place, he points out1 that what is remarkable in connection with suggestion is not the actual nature of the effects, which can all be produced by other means, but the way in which they are produced. It is the conditions under which the effects follow an idea that are peculiar to suggestion, for these effects would not follow the idea under other conditions. The characteristic of these conditions he sees in a special combination of intact psychical energy with diminished psychical excitability2. By the latter phrase he means an inhibition of the counter-ideas which normally would oppose the action of the suggested ones. This inhibition is, of course, related to the contrasting freedom with which ideas are accepted from the operator, and is thus the secondary result of the state of rapport mentioned above. He therefore includes these two additional conclusions in his final definition of suggestion, which is3: "Die Hervorrufung einer über das blosse Dasein einer Vorstellung hinausgehenden psychischen Wirkung in einem Individuum, durch Weckung einer Vorstellung seitens einer Person oder eines von dem Individuum verschiedenen Objectes, sofern diese psychische Wirkung durch eine in ausserordentlichem Masse stattfindende Hemmung oder Lähmung der über die nächste reproducirende Wirkung der Suggestion hinausgehenden Vorstellungsbewegung bedingt ist." ("The evocation in an individual, through an idea being aroused by another person or an object distinct from the individual, of a psychical effect that goes beyond the mere existence of this idea, provided always that this psychical effect is conditioned by an extraordinary inhibition or paralysis of the ideational movement which passes beyond the proximate reproductive effect of the suggestion.") He explicitly included autosuggestion in this definition in a way which will presently be noted.

The actual phenomenology of the effects of suggestion are too well known to need recounting here. Concerning their nature Lipps has shown

1 "Zur Psychologie der Suggestion," Zeitschr. f. Hypnotismus, 1897, Band VII. S. 95. 2 Idem, "Suggestion und Hypnose," op. cit. S. 520.

* Idem, "Zur Psychologie der Suggestion," op. cit. S. 117.

in detail that all of them, even the eliciting of hallucinatory sensations, represent the normal logical consequences of the suggested ideas, differing only from the usual consequences of the same idea in that, through the inhibition of the criticising ideas customarily operative, they are allowed to proceed to their logical termination without hindrance. We may therefore conclude that the characteristic of suggestion lies in the free development of the effects of communicated ideas, the forces usually hindering this development being neutralised by the presence of the rapport, or concentration on the idea of the operator. It is generally agreed that this rapport consists of an emotional bond; as is well known, psycho-analysts consider the bond to be sexual in nature and due to the re-animation of an infantile attachment to a parent.

Our formulation of the three processes thus runs in order: rapport; inhibition of all mental processes except those suggested; free development of the latter. We are now able to reduce the difference of opinion noted above to differences in the view held of the way in which the rapport operates; all are agreed that it is in this that the operative force resides. From this point of view the two schools of thought may be contrasted somewhat as follows: according to one, the main thing is the remarkable influence exerted by the operator, or hypnotist; granted this and the rest follows, the ideas developing to their logical conclusion by the sheer force imparted to them. According to the other school, the main thing is the subject's peculiar attitude towards the operator; it is this which neutralises any critical ideas inimical to his. Psycho-analysts may certainly be classed as belonging to the latter school. Some thirteen years ago, for instance, I wrote1: "We can no longer regard the subject as a helpless automaton in the hands of a strong-willed operator; it is nearer the truth to regard the operator as allowing himself to play a part, and by no means an indispensable one, in a drama constructed and acted in the depths of the subject's mind."

From what has been said, it is not astonishing that the two views just described lead to contrasting attitudes towards the subject of autosuggestion. Those who expound the former of the two views tend to decry the importance of auto-suggestion or else to deny its existence altogether, to depreciate its practical value, and to attribute most of its phenomena, whether therapeutic or pathogenic, to some more or less disguised form of hetero-suggestion. In this group of authors may be mentioned Baragnon2, Camus and Pagniez3, McDougall1, and Grasset5;

1 Op. cit. p. 220. 4 Loc. cit.

2 Loc. cit.

3 Isolement et Psychothérapie, 1904, p. 57. 5 L'Hypnotisme et la Suggestion, 1904, p. 131.

the last-named of these goes so far as to hint that auto-suggestion is in most cases the result of previous hypnotism. Janet1 would appear to take up an intermediate position; he ascribes at least a great many pathological processes to auto-suggestion, apart from the intervention of an idea from without. Forel2 also holds that "Jede Suggestion wird durch Autosuggestion des Hypnotisierten ergänzt und modifiziert" ("every suggestion is added to and modified by auto-suggestion on the part of the hypnotised person"). At the other extreme there is Baudouin3, the leading exponent of auto-suggestion, who holds the diametrically opposite view that "hetero-suggestion, even during induced sleep (i.e. hypnosis), is still an auto-suggestion." Similarly Levy-Suhl maintains: "Jede Suggestionswirkung beruht letzthin in einer Autosuggestion" ("every effect of suggestion rests ultimately on an auto-suggestion”).

We thus return to the problem of what phenomena, if any, are to be classed as belonging to auto-suggestion. The matter is certainly not to be settled by simply asking whether the operative ideas have originated from within or from without. In the first place, this is often very hard to determine, and in a certain sense it might even be maintained that all ideas take their ultimate source from the outer world. Secondly, the question does not touch the essential part of the problem, for clinical psychology no longer regards ideas as active agents in themselves; any activity they may exhibit is due only to their being representatives of some impulse or other. We must therefore concentrate our attention on the nature of the dynamic factors at work, and in this way seek to determine whether two classes of them can be detected, corresponding with hetero-suggestion and auto-suggestion respectively. Several writers, e.g. Baudouin, insist that the ideas produce their effect only through acting outside the field of consciousness, but being unfamiliar with what goes on in this unconscious layer of the mind they were unable to throw any light on the nature of the forces operative in the transformation of the 'idea' into its effect, i.e. the 'realisation' of the idea. Lipps holds that in auto-suggestion, just as in hetero-suggestion, there is a general inhibition of mental excitability, particularly of ideas antagonistic to the ones being 'suggested.' In hetero-suggestion this is brought about through a high degree of psychical investment of the idea of the operator; 1 Op. cit. p. 71.

2 Der Hypnotismus, 11 Auflage, S. 122.

3 Suggestion and Auto-Suggestion, Engl. Trans., 1920, p. 204.
4 Die hypnotische Heilweise und ihre Technik, 1922, S. 33.

5 Idem, op. cit. p. 26.

Op. cit S. 117.

in psycho-analytical terminology, a hyper-cathexis of the idea of the operator is correlated with a hypo-cathexis of all ideas in conflict with his. Now is there a group of phenomena, to be called auto-suggestion, in which there is a corresponding hyper-cathexis of a given idea to account for the general hypo-cathexis that Lipps maintains to be present, and, if so, what is known of the nature of this idea? The only suggestion he makes in this connection is1 that the part of the ego communicating the idea is to be regarded as a foreign object to the part that receives it, but he throws no further light on this remarkable splitting of the personality. Baudouin2 repeatedly insists also on the essential importance of relaxation in the practice of auto-suggestion, and it is evident that this relation is identical with the inhibition of mental excitability described by Lipps. Baudouin's3 conception of the concentration of attention necessary in addition to the general relaxation-the two features which in his opinion comprise the essentials of the practice of auto-suggestion— corresponds further with what we have called hyper-cathexis of a particular idea, but he never mentions any idea to which this applies except the idea which is being 'suggested.'

It might be supposed that psycho-analysis, adhering as it does to the second of the two schools described above, the school that lays stress on the part played in the depths of the subject's mind, would at once lend countenance to auto-suggestion as a phenomenon which obviously supports the view in question. On the other hand, it would appear to contradict the psycho-analytical view concerning the significance of the idea of the operator, at all events unless it can be shown that in autosuggestion there is a hyper-cathexis of another idea which is equivalent to that of the operator.

It is time to turn from this general discussion of the problem and consider the actual data bearing on it. It must be said, however, that it is by no means easy to ascertain these. To begin with, McDougall's criticism that in so many of the examples cited of auto-suggestion one cannot exclude the operation of hetero-suggestion is evidently justified; it obviously applies to a great part of Coué's performances. Indeed, this factor has also to be taken into account when a person practises ‘autosuggestion' after reading a book of instructions, for the idea of the authority behind this book must often play a considerable part. McDougall further objects to the wide application of the term 'auto-suggestion' to such phenomena as the ready acceptance of propositions which are con