To avoid such a consequence, there is, we must admit, a conceivable loophole. It consists in maintaining on principle an absolute analogy, an exact similitude between internal reality and external objects. The forms which suit the one would then also suit the other.
But it must be observed that such a principle constitutes in the highest degree a metaphysical thesis which it would be on all hands illegal to assert previously as a postulate of method. Secondly, and above all, it must be observed that on this head experience is decisive, and manifests more plainly every day the failure of the theories which try to assimilate the world of consciousness to that of matter, to copy psychology from physics. We have here two different "orders." The apparatus of the first does not admit of being employed in the second. Hence the necessity of the attitude adopted by Mr Bergson. We have an effort to make, a work of reform to undertake, to lift the veil of symbols which envelops our usual representation of the ego, and thus conceals us from our own view, in order to find out what we are in reality, immediately, in our inmost selves. This effort and this work are necessary, because, "in order to contemplate the ego in its original purity, psychology must eliminate or correct certain forms which bear the visible mark of the outer world." ("Essay on the Immediate Data of Consciousness", Conclusion.) What are these forms? Let us confine ourselves to the most important. Things appear to us as numerable units, placed side by side in space. They compose numerical and spatial multiplicity, a dust of terms between which geometrical ties are established.
But space and number are the two forms of immobility, the two schemes of analysis, by which we must not let ourselves be obsessed. I do not say that there is no place to give them, even in the internal world. But the more deeply we enter into the heart of psychological life, the less they are in place.
The fact is, there are several planes of consciousness, situated at different depths, marking all the intervening degrees between pure thought and bodily action, and each mental phenomenon interests all these planes simultaneously, and is thus repeated in a thousand higher tones, like the harmonies of one and the same note.
Or, if you prefer it, the life of the spirit is not the uniform transparent surface of a mere; rather it is a gushing spring which, at first pent in, spreads upwards and outwards, like a sheaf of corn, passing through many different states, from the dark and concentrated welling of the source to the gleam of the scattered tumbling spray; and each of its moods presents in its turn a similar character, being itself only a thread within the whole. Such without doubt is the central and activating idea of the admirable book entitled "Matter and Memory". I cannot possibly condense its substance here, or convey its astonishing synthetic power, which succeeds in contracting a complete metaphysic, and in gripping it so firmly that the examination ends by passing to the discussion of a few humble facts relative to the philosophy of the brain! But its technical severity and its very conciseness, combined with the wealth it contains, render it irresumable; and I can only in a few words indicate its conclusions.
First of all, however little we pride ourselves on positive method, we must admit the existence of an internal world, of a spiritual activity distinct from matter and its mechanism. No chemistry of the brain, no dance of atoms, is equivalent to the least thought, or indeed to the least sensation.
Some, it is true, have brought forward a thesis of parallelism, according to which each mental phenomenon corresponds point by point to a phenomenon in the brain, without adding anything to it, without influencing its course, merely translating it into another tongue, so that a glance sufficiently penetrating to follow the molecular revolutions and the fluxes of nervous production in their least episodes would immediately read the inmost secrets of the associated consciousness.
But no one will deny that a thesis of this kind is only in reality a hypothesis, that it goes enormously beyond the certain data of current biology, and that it can only be formulated by anticipating future discoveries in a preconceived direction. Let us be candid: it is not really a thesis of positive science, but a metaphysical thesis in the unpleasant meaning of the term. Taking it at its best, its worth today could only be one of intelligibleness. And intelligible it is not.
How are we to understand a consciousness destitute of activity and consequently without connection with reality, a kind of phosphorescence which emphasises the lines of vibration in the brain, and renders in miraculous duplicate, by its mysterious and useless light, certain phenomena already complete without it?
One day Mr Bergson came down into the arena of dialectic, and, talking to his opponents in their own language, pulled their "psycho-physiological paralogism" to pieces before their eyes; it is only by confounding in one and the same argument two systems of incompatible notations, idealism and realism, that we succeed in enunciating the parallelist thesis. This reasoning went home, all the more as it was adapted to the usual form of discussions between philosophers. But a more positive and more categorical proof is to be found all through "Matter and Memory". From the precise example of recollection analysed to its lowest depths, Mr Bergson completely grasps and measures the divergence between soul and body, between mind and matter. Then, putting into practice what he said elsewhere about the creation of new concepts, he arrives at the conclusion—these are his own expressions—that between the psychological fact and its counterpart in the brain there must be a relation sui generis, which is neither the determination of the one by the other, nor their reciprocal independence, nor the production of the latter by the former, nor of the former by the latter, nor their simple parallel concomitance; in short, a relation which answers to none of the ready-made concepts which abstraction puts at our service, but which may be approximately formulated in these terms: ("Report of the French Philosophical Society", meeting, 2nd May 1901.)