Time and Free Will · Henri Bergson
FREE WILL
A thorough examination of these two conceptions of nature will show that they involve two very different hypotheses as to the relations between laws and the facts which they govern. As he looks higher and higher, the believer in dynamism thinks that he perceives facts which more and more elude the grasp of laws: he thus sets up the fact as the absolute reality, and the law as the more or less symbolical expression of this reality. Mechanism, on the contrary, discovers within the particular fact a certain number of laws of which the fact is thus made to be the meeting point, and nothing else: on this hypothesis it is the law which becomes the genuine reality. Now, if it is asked why the one party assigns a higher reality to the fact and the other to the law, it will be found that mechanism and dynamism take the word _simplicity_ in two very different senses. For the first, any principle is simple of which the effects can be foreseen and even calculated: thus, by the very definition, the notion of inertia becomes simpler than that of freedom, the homogeneous simpler than the heterogeneous, the abstract simpler than the concrete.
But dynamism is not anxious so much to arrange the notions in the most convenient order as to find out their real relationship: often, in fact, the so-called simple notion--that which the believer in mechanism regards as primitive--has been obtained by the blending together of several richer notions which seem to be derived from it, and which have more or less neutralized one another in this very process of blending, just as darkness may be produced by the interference of two lights. Regarded from this new point of view, the idea of spontaneity is indisputably simpler than that of inertia, since the second can be understood and defined only by means of the first, while the first is self-sufficient. For each of us has the immediate knowledge (be it thought true or fallacious) of his free spontaneity, without the notion of inertia having anything to do with this knowledge. But, if we wish to define the inertia of matter, we must say that it cannot move or stop of its own accord, that every body perseveres in the state of rest or motion so long as it is not acted upon by any force: and in both cases we are unavoidably carried back to the idea of activity. It is therefore natural that, _a priori,_ we should reach two opposite conceptions of human activity, according to the way in which we understand the relation between the concrete and the abstract, the simple and the complex, facts and laws.
Physical determinism, in its latest form, is closely bound up with mechanical or rather kinetic theories of matter. The universe is pictured as a heap of matter which the imagination resolves into molecules and atoms. These particles are supposed to carry out unceasingly movements of every kind, sometimes of vibration, sometimes of translation; and physical phenomena, chemical action, the qualities of matter which our senses perceive, heat, sound, electricity, perhaps even attraction, are thought to be reducible objectively to these elementary movements. The matter which goes to make up organized bodies being subject to the same laws, we find in the nervous system, for example, only molecules and atoms which are in motion and attract and repel one another. Now if all bodies, organized or unorganized, thus act and react on one another in their ultimate parts, it is obvious that the molecular state of the brain at a given moment will be modified by the shocks which the nervous system receives from the surrounding matter, so that the sensations, feelings and ideas which succeed one another in us can be defined as mechanical resultants, obtained by the compounding of shocks received from without with the previous movements of the atoms of the nervous substance.
But the opposite phenomenon may occur; and the molecular movements which go on in the nervous system, if compounded with one another or with others, will often give as resultant a reaction of our organism on its environment: hence the reflex movements, hence also the so-called free and voluntary actions. As, moreover, the principle of the conservation of energy has been assumed to admit of no exception, there is not an atom, either in the nervous system or in the whole of the universe, whose position is not determined by the sum of the mechanical actions which the other atoms exert upon it. And the mathematician who knew the position of the molecules or atoms of a human organism at a given moment, as well as the position and motion of all the atoms in the universe capable of influencing it, could calculate with unfailing certainty the past, present and future actions of the person to whom this organism belongs, just as one predicts an astronomical phenomenon.[1] [Sidenote: If principle of conservation of energy is universal, physiological and nervous phenomena are necessitated, but perhaps not conscious states.]
We shall not raise any difficulty about recognizing that this conception of physiological phenomena in general, and nervous phenomena in particular, is a very natural deduction from the law of the conservation of energy. Certainly, the atomic theory of matter is still at the hypothetical stage, and the purely kinetic explanations of physical facts lose more than they gain by being too closely bound up with it. We must observe, however, that, even if we leave aside the atomic theory as well as any other hypothesis as to the nature of the ultimate elements of matter, the necessitating of physiological facts by their antecedents follows from the theorem of the conservation of energy, as soon as we extend this theorem to all processes going on in all living bodies. For to admit the universality of this theorem is to assume, at bottom, that the material points of which the universe is composed are subject solely to forces of attraction and repulsion, arising from these points themselves and possessing intensities which depend only on their distances: hence the relative position of these material points at a given moment--whatever be their nature--would be strictly determined by relation to what it was at the preceding moment. Let us then assume for a moment that this last hypothesis is true: we propose to show, in the first place, that it does not involve the absolute determination of our conscious states by one another, and then that the very universality of the principle of the conservation of energy cannot be admitted except in virtue of some psychological hypothesis.
Sidenote: To prove conscious states determined, we should have to show a necessary connexion between them and cerebral states. No such proof. Even if we assumed that the position, the direction and the velocity of each atom of cerebral matter are determined at every moment of time, it would not at all follow that our psychic life is subject to the same necessity. For we should first have to prove that a strictly determined psychic state corresponds to a definite cerebral state, and the proof of this is still to be given. As a rule we do not think of demanding it, because we know that a definite vibration of the tympanum, a definite stimulation of the auditory nerve, gives a definite note on the scale, and because the parallelism of the physical and psychical series has been proved in a fairly large number of cases. But then, nobody has ever contended that we were free, under given conditions, to hear any note or perceive any colour we liked. Sensations of this kind, like many other psychic states, are obviously bound up with certain determining conditions, and it is just for this reason that it has been possible to imagine or discover beneath them a system of movements which obey our abstract mechanics.
In short, wherever we succeed in giving a mechanical explanation, we observe a fairly strict parallelism between the physiological and the psychological series, and we need not be surprised at it, since explanations of this kind will assuredly not be met with except where the two series exhibit parallel terms. But to extend this parallelism to the series themselves in their totality is to settle _a priori_ the problem of freedom. Certainly this may be done, and some of the greatest thinkers have set the example; but then, as we said at first, it was not for reasons of a physical order that they asserted the strict correspondence between states of consciousness and modes of extension. Leibniz ascribed it to a preestablished harmony, and would never have admitted that a motion could give rise to a perception as a cause produces an effect. Spinoza said that the modes of thought and the modes of extension correspond with but never influence one another: they only express in two different languages the same eternal truth. But the theories of physical determinism which are rife at the present day are far from displaying the same clearness, the same geometrical rigour.
They point to molecular movements taking place in the brain: consciousness is supposed to arise out of these at times in some mysterious way, or rather to follow their track like the phosphorescent line which results from the rubbing of a match. Or yet again we are to think of an invisible musician playing behind the scenes while the actor strikes a keyboard the notes of which yield no sound: consciousness must be supposed to come from an unknown region and to be superimposed on the molecular vibrations, just as the melody is on the rhythmical movements of the actor. But, whatever image we fall back upon, we do not prove and we never shall prove by any reasoning that the psychic fact is fatally determined by the molecular movement. For in a movement we may find the reason of another movement, but not the reason of a conscious state: only observation can prove that the latter accompanies the former. Now the unvarying conjunction of the two terms has not been verified by experience except in a very limited number of cases and with regard to facts which all confess to be almost independent of the will. But it is easy to understand why physical determinism extends this conjunction to all possible cases. [Sidenote: Physical determinism, when assumed to be universal, postulates psychological determinism.]
Consciousness indeed informs us that the majority of our actions can be explained by motives. But it does not appear that determination here means necessity, since common sense believes in free will. The determinist, however, led astray by a conception of duration and causality which we shall criticise a little later, holds that the determination of conscious states by one another is absolute. This is the origin of associationist determinism, an hypothesis in support of which the testimony of consciousness is appealed to, but which cannot, in the beginning, lay claim to scientific rigour. It seems natural that this, so to speak, approximate determinism, this determinism of quality, should seek support from the same mechanism that underlies the phenomena of nature: the latter would thus convey to the former its own geometrical character, and the transaction would be to the advantage both of psychological determinism, which would emerge from it in a stricter form, and of physical mechanism, which would then spread over everything. A fortunate circumstance favours this alliance. The simplest psychic states do in fact occur as accessories to well-defined physical phenomena, and the greater number of sensations seem to be bound up with definite molecular movements. This mere beginning of an experimental proof is quite enough for the man who, for psychological reasons, is already convinced that our conscious states are the necessary outcome of the circumstances under which they happen. Henceforth he no longer hesitates to hold that the drama enacted in the theatre of consciousness is a literal and even slavish translation of some scenes performed by the molecules and atoms of organized matter. The physical determinism which is reached in this way is nothing but psychological determinism, seeking to verify itself and fix its own outlines by an appeal to the sciences of nature.
We must not overrate the part played by the principle of the conservation of energy in the history of the natural sciences. In its present form it marks a certain phase in the evolution of certain sciences; but it has not been the governing factor in this evolution and we should be wrong in making it the indispensable postulate of all scientific research. Certainly, every mathematical operation which we carry out on a given quantity implies the permanence of this quantity throughout the course of the operation, in whatever way we may split it up. In other words, what is given is given, what is not given is not given, and in whatever order we add up the same terms we shall get the same result. Science will for ever remain subject to this law, which is nothing but the law of non-contradiction; but this law does not involve any special hypothesis as to the nature of what we ought to take as given, or what will remain constant. No doubt it informs us that something cannot come from nothing; but experience alone will tell us which aspects or functions of reality must count for something, and which for nothing, from the point of view of positive science.
In short, in order to foresee the state of a determinate system at a determinate moment, it is absolutely necessary that something should persist as a constant quantity throughout a series of combinations; but it belongs to experience to decide as to the nature of this something, and especially to let us know whether it is found in all possible systems, whether, in other words, all possible systems lend themselves to our calculations. It is not certain that all the physicists before Leibniz believed, like Descartes, in the conservation of a fixed quantity of motion in the universe: were their discoveries less valuable on this account or their researches less successful? Even when Leibniz had substituted for this principle that of the conservation of _vis viva,_ it was not possible to regard the law as quite general, since it admitted of an obvious exception in the case of the direct impact of two inelastic bodies. Thus science has done for a very long time without a universal conservative principle. In its present form, and since the development of the mechanical theory of heat, the principle of the conservation of energy certainly seems to apply to the whole range of physico-chemical phenomena.
But no one can tell whether the study of physiological phenomena in general, and of nervous phenomena in particular, will not reveal to us, besides the _vis viva_ or kinetic energy of which Leibniz spoke, and the potential energy which was a later and necessary adjunct, some new kind of energy which may differ from the other two by rebelling against calculation. Physical science would not thereby lose any of its exactitude or geometrical rigour, as has lately been asserted: only it would be realized that conservative systems are not the only systems possible, and even, perhaps, that in the whole of concrete reality each of these systems plays the same part as the chemist's atom in bodies and their combinations. Let us note that the most radical of mechanical theories is that which makes consciousness an _epiphenomenon_ which, in given circumstances, may supervene on certain molecular movements. But, if molecular movement can create sensation out of a zero of consciousness, why should not consciousness in its turn create movement either out of a zero of kinetic and potential energy, or by making use of this energy in its own way?
Let us also note that the law of the conservation of energy can only be intelligibly applied to a system of which the points, after moving, can return to their former positions. This return is at least conceived of as possible, and it is supposed that under these conditions nothing would be changed in the original state of the system as a whole or of its elements. In short, time cannot bite into it; and the instinctive, though vague, belief of mankind in the conservation of a fixed quantity of matter, a fixed quantity of energy, perhaps has its root in the very fact that inert matter does not seem to endure or to preserve any trace of past time. But this is not the case in the realm of life. Here duration certainly seems to act like a cause, and the idea of putting things back in their place at the end of a certain time involves a kind of absurdity, since such a turning backwards has never been accomplished in the case of a living being.
But let us admit that the absurdity is a mere appearance, and that the impossibility for living beings to come back to the past is simply owing to the fact that the physico-chemical phenomena which take place in living bodies, being infinitely complex, have no chance of ever occurring again all at the same time: at least it will be granted to us that the hypothesis of a turning backwards is almost meaningless in the sphere of conscious states. A sensation, by the mere fact of being prolonged, is altered to the point of becoming unbearable. The same does not here remain the same, but is reinforced and swollen by the whole of its past. In short, while the material point, as mechanics understands it, remains in an eternal present, the past is a reality perhaps for living bodies, and certainly for conscious beings. While past time is neither a gain nor a loss for a system assumed to be conservative, it may be a gain for the living being, and it is indisputably one for the conscious being. Such being the case, is there not much to be said for the hypothesis of a conscious force or free will, which, subject to the action of time and storing up duration, may thereby escape the law of the conservation of energy?
After this it would be absurd to consider time, even _our_ time, as a cause of gain or loss, as a concrete reality, or a force in its own way. Thus, while we ought only to say (if we kept aloof from all presuppositions concerning free will) that the law of the conservation of energy governs physical phenomena and _may,_ one day, be extended to all phenomena if psychological facts also prove favourable to it, we go far beyond this, and, under the influence of a metaphysical prepossession, we lay down the principle of the conservation of energy as a law which _should_ govern all phenomena whatever, or must be supposed to do so until psychological facts have actually spoken against it. Science, properly so called, has therefore nothing to do with all this. We are simply confronted with a confusion between concrete duration and abstract time, two very different things. In a word, the so-called physical determinism is reducible at bottom to a psychological determinism, and it is this latter doctrine, as we hinted at first, that we have to examine. [Sidenote: Psychological determinism depends on associationist conception of mind.]
Psychological determinism, in its latest and most precise shape, implies an associationist conception of mind. The existing state of consciousness is first thought of as necessitated by the preceding states, but it is soon realized that this cannot be a geometrical necessity, such as that which connects a resultant, for example, with its components. For between successive conscious states there exists a difference of quality which will always frustrate any attempt to deduce any one of them _a priori_ from its predecessors. So experience is appealed to, with the object of showing that the transition from one psychic state to another can always be explained by some simple reason, the second obeying as it were the call of the first. Experience really does show this: and, as for ourselves, we shall willingly admit that there always is some relation between the existing state of consciousness and any new state to which consciousness passes. But is this relation, which explains the transition, the cause of it? [Sidenote: The series of associations may be merely an ex post facto attempt to account for a new idea.]
May we here give an account of what we have personally observed? In resuming a conversation which had been interrupted for a few moments we have happened to notice that both we ourselves and our friend were thinking of some new object at the same time.--The reason is, it will be said, that each has followed up for his own part the natural development of the idea at which the conversation had stopped: the same series of associations has been formed on both sides.--No doubt this interpretation holds good in a fairly large number of cases; careful inquiry, however, has led us to an unexpected result. It is a fact that the two speakers do connect the new subject of conversation with the former one: they will even point out the intervening ideas; but, curiously enough, they will not always connect the new idea, which they have both reached, with the same point of the preceding conversation, and the two series of intervening associations may be quite different. What are we to conclude from this, if not that this common idea is due to an unknown cause--perhaps to some physical influence--and that, in order to justify its emergence, it has called forth a series of antecedents which explain it and which seem to be its cause, but are really its effect?
If we question ourselves carefully, we shall see that we sometimes weigh motives and deliberate over them, when our mind is already made up. An inner voice, hardly perceivable, whispers: "Why this deliberation? You know the result and you are quite certain of what you are going to do." But no matter! it seems that we make a point of safe-guarding the principle of mechanism and of conforming to the laws of the association of ideas. The abrupt intervention of the will is a kind of _coup d'état_ which our mind foresees and which it tries to legitimate beforehand by a formal deliberation. True, it could be asked whether the will, even when it wills for willing's sake, does not obey some decisive reason, and whether willing for willing's sake is free willing. We shall not insist on this point for the moment. It will be enough for us to have shown that, even when adopting the point of view of associationism, it is difficult to maintain that an act is absolutely determined by its motive and our conscious states by one another. Beneath these deceptive appearances a more attentive psychology sometimes reveals to us effects which precede their causes, and phenomena of psychic attraction which elude the known laws of the association of ideas. But the time has come to ask whether the very point of view which associationism adopts does not involve a defective conception of the self and of the multiplicity of conscious states.
Nevertheless language would have still expressed the movement and the position in the same way; and associationism would have distinguished the two cases by saying that with the idea of the same movement there was associated this time the idea of a new end: as if the mere newness of the end to be attained did not alter in some degree the idea of the movement to be performed, even though the movement itself remained the same! We should thus say, not that the image of a certain position can be connected in consciousness with images of different ends to be attained, but rather that positions geometrically identical outside look different to consciousness from the inside, according to the end contemplated. The mistake of associationism is that it first did away with the qualitative element in the act to be performed and retained only the geometrical and impersonal element: with the idea of this act, thus rendered colourless, it was then necessary to associate some specific difference to distinguish it from many other acts. But this association is the work of the associationist philosopher who is studying my mind, rather than of my mind itself. [Sidenote: Illustration from "associations" of smell.]
Ideas discussed in this chapter
Text from Project Gutenberg, public domain in the United States.