consciousness in order to explain the facts, a consideration which points, as we shall see later, in the direction of the Freudian 'unconscious.'

Some of the difficulties in the application of the conception of dissociation disappear if we abandon the atomistic and spatial terminology in which the conception was originally described. Naturally the phenomena with which we are concerned are in reality absolutely devoid of any actual spatial aspect, and the introduction of a spatial metaphor, although it lends itself insidiously to the construction of attractive diagrams, can only lead to erroneous deductions unless its purely descriptive and illustrative function is rigidly controlled. Dissociation does not separate the mind into pieces, it only produces more or less independently acting functional units, each such unit comprising material which may be peculiar to itself, but which may just as well form a part of any number of other functional units. The distinguishing character does not lie in the material of which it is composed, but in the set or pattern. Instead of regarding dissociation as the splitting of conscious material into separate masses, it must be regarded as an affair of gearing, the various elements of mental machinery being organized into different functional systems by the throwing in of the appropriate gear, or, if it is desired to render the situation in physiological terms, we have the notion of engrams and neuronic patterns developed in Dr R. G. Gordon's recent book1. With this conception the difficulty in understanding how the same material can belong to several personalities, or how there can be a non-reciprocal amnesia between the normal and hypnotic consciousness, is largely overcome.

The spatial and functional conceptions of dissociation are radically distinct from one another in their angle of approach to the phenomena which they seek to describe. The former regards the dissociated consciousness as built up by the accretion of elements, the simplest example being provided by the cases where only a few such elements are dissociated, hysterical anaesthesia for instance, while the more complex cases are produced by the addition of more and more elements to the dissociated mass, until finally that mass attains dimensions to which the term 'personality' may reasonably be ascribed. The functional conception, on the contrary, starts at the other end. It lays stress on the synthesizing activity which brings the elements together, and regards this as the essential feature rather than the mere agglomeration of elements. Instead of seeing in personality the final result of an unusually extensive agglomeration, it assumes that some synthesizing agent comparable to personality is present in every case.

1 R. G. Gordon, Personality, London, 1926.

With a functional interpretation of this type the conception of dissociation can be considerably extended, and applied to phenomena which could not be subsumed under the spatial interpretation. Several cases of multiple personality have been described in which the different personalities share the same memories, but are sharply distinguished from one another by the diversity of their characters and activities. There seems to be no ground for excluding these cases from the group of multiple personality, although there is no mutual amnesia. Similar considerations apply to those fugues where there is no subsequent amnesia, but where the behaviour during the fugue is totally foreign to that characterizing the normal self. Perhaps it is even justifiable to extend the conception to the changes observed in cyclothymia and the manic-depressive psychosis. It can be applied, again, to the interpretation of hallucinations, to those conditions where elaborate delusional systems exist without effect upon the behaviour of the patient, and even to the logic-tight compartment mechanisms observed in everyday life. The objection to this kind of extension is that a generalization which is stretched over a wide and graded series of phenomena tends to lose its precision and definition, and therefore its utility. It has some value, however, in that it connotes a common factor, i.e. a lack of integration, throughout the whole series, though varying in degree from minor examples up to the complete splits of double personality. It should be noted, by the way, that the extension of the conception of dissociation whereby Janet endeavours to explain the phenomena of psychasthenia, is not an extension of the same order, but a further development of the atomistic notion of dissociation.

It was pointed out at the commencement of this address that the history of psychology shows numerous attempts to establish lines of division in the continuity of the mind. Now the divisions which have so far been considered, the divisions included under dissociation, even in the extended sense of dissociation described above, are divisions of consciousness. The evidence for the existence of a dissociated stream of consciousness is of precisely the same order as that which establishes the existence of any kind of consciousness in people other than ourselves. The dissociated stream is made of the same stuff as the remainder of consciousness, and its peculiarity lies simply in the fact that there is a lack of complete integration between it and the remainder of consciousness. Janet's 'subconscious' comprises those instances of dissociation where the lack of integration is such that there is a lack of mutual awareness between the two streams, but in every other respect the pro

cesses concerned have all the attributes of consciousness. The division of dissociation is therefore a division within consciousness or, as will be explained later, it is a division on the phenomenal plane.

Now the division created by Freud in his conception of the 'unconscious' is a division along an entirely different plane. It is true that the actual history of its development was by way of clinical investigations of much the same kind as those which led to Janet's conception, but, as has already been mentioned, its logical ancestry is altogether different. This ancestry is to be found in the hypotheses which assumed the existence of mental processes lying altogether outside consciousness, but whose activity explained the facts of consciousness, hypotheses formulated by Kant, Schopenhauer, and Hartmann. The unconscious of Freud has been created by him in order to explain the processes occurring in consciousness. It is not in itself a fact of consciousness, and its existence cannot be demonstrated in the way in which the existence of Janet's dissociated streams can be demonstrated, any more than we can demonstrate the existence of the ether which has been created in order to explain the facts of light and heat.

In a paper published in 1909 I endeavoured to formulate and analyse the essential difference of plane existing between the conceptions of Janet and Freud, and the chief conclusions reached therein still seem to me to be fundamentally sound1. The 'subconscious' of Janet is a description of phenomenal facts, while the 'unconscious' of Freud is a conceptual construction, an imagined entity created in order to explain phenomenal facts. The distinction is that which marks, in all branches of science, the difference between a generalization from observed facts, e.g. the laws of refraction of various coloured lights, and a conceptual construction designed to explain the facts, e.g. the ether and its waves. The validity of a generalization from observed facts is established by a procedure entirely different from that needed to establish the validity of a conceptual construction. In the former case only observation and experiment are required; the generalization states that certain phenomena occur under certain conditions, and the demonstration that the phenomena do so occur establishes the generalization. The conceptual construction, on the other hand, cannot be demonstrated in this way, because the elements of which it is manufactured have no existence on the phenomenal plane. The procedure here is to deduce the consequences which would follow from the conceptual construction, as it is conceived

1 Bernard Hart, "The Conception of the Subconscious," Journ. of Abnorm. Psychol. 1909. Reprinted in Morton Prince's Subconscious Phenomena, Boston, 1915.

to be and to act, and then to compare these deduced consequences with the phenomena which are actually observed. If the deduced consequences are identical with the observed phenomena, then the validity of the conceptual construction is established.

These considerations lead to the conclusion that the difference of plane between Janet's 'subconscious' and Freud's 'unconscious' is that which lies between observable phenomena and conceptual constructions. This, indeed, was the conclusion reached in the paper to which I have referred, but there are certain difficulties in the way of its acceptance. It may be objected that the so-called phenomena of psychology, or at any rate of psychopathology, are not observed phenomena at all, but inferences. The conscious processes of others cannot be directly observed; they can only be inferred, and such inferences are of the order of conceptual constructions rather than phenomena. If this be so, no rigid line of demarcation can be established in psychology between phenomena and conceptual constructions, and the distinction becomes one of degree rather than of an essential difference in kind. Even if this be admitted, however, the difference of degree between Janet's and Freud's conception is so pronounced as to constitute a fundamental distinction of method, and it is from the side of methodology that the divergence in plane can be best represented. From this standpoint the conscious processes of others, although strictly inferences rather than actually observed phenomena, are treated as observed phenomena.

Karl Pearson1 has shown that the method of science consists in three steps, firstly the observing and recording of phenomena, secondly the classification of the phenomena observed, and thirdly the construction of formulae or laws which will resume or explain the phenomena. The three steps may be exemplified by the history of our knowledge of planetary motion: the first by the observation of the positions occupied at different times by the planets; the second by the classification of these observations by Kepler, resulting in his generalization that the planets travel in ellipses; and the third by Newton's law of gravity, which resumed in a single formula the whole movement and track of the planets. In this third step it is legitimate to use a conceptual imagery, and to construct hypothetical entities conceived to behave in a certain manner. This was done to some extent by Newton, but the more clearly apparent in such a conception as that of the ether. Now if we apply this methodological approach to the conceptions of the subconscious and the unconscious, it is obvious that Janet's dis

1 Karl Pearson, The Grammar of Science, 1892.

process

is

sociation belongs to the second step; it is a classification of observed phenomena, while Freud's unconscious belongs to the third step; it is a conceptual construction. This distinction of plane between subconscious and unconscious, manifesting itself in the radically divergent methods by which the conceptions are reached and by which they can be tested, is of fundamental importance. Unless it is recognized, confusion of thought is inevitable, and it is not difficult to find examples of such confusion. Hitschmann1, for instance, after quite correctly pointing out that 'unconscious, in the Freudian sense,... means something which one does not really know, while one is compelled in the analysis by conclusive inferences to recognize it," goes on to compare the division into conscious and unconscious to the splitting of consciousness which occurs in double personality, and states that, "if in such a splitting of personality the consciousness remains constantly joined to one of the two conditions, then this is called the conscious mental condition, the one separated from it, the unconscious." Clearly this comparison is not permissible, because the doubling of personality is a phenomenal event capable of being directly observed, while the unconscious of Freud is a conceptual abstraction. Moreover, while it is possible that one dissociated personality may exert some influence upon another, it is obviously not the same order of influence as that which Freud conceives to exist between the unconscious and the conscious.

In Freud's conceptions of the ego and the id2 a further division of the mind is formulated, whose relations to the conscious and the unconscious, and to Janet's dissociation, now call for consideration. Freud describes the ego as "the connecting organization of the mental processes in an individual3," and regards it as centred round the perceptual system of the psychical apparatus. The remainder of the psyche he calls the id. The ego is properly a differentiated portion of the id, developed by the influence of the outer world acting through the perceptual system. It endeavours to bring the influence of the outer world to bear upon the processes arising from the id, and to replace the pleasure-principle, which reigns without restraint in the id, by the reality-principle. Perception plays in it the part which falls to instinctual force in the ida.

It does not require demonstration that the division into ego and id is a division of a conceptual and not of a phenomenal kind, and one therefore

1 Hitschmann, Freud's Theories of the Neuroses, English translation, New York, 1917, p. 80.

2 Freud, Das Ich und das Es, Wien, 1923. 3 Ibid. p. 14.

4 Ibid. p. 27.