Memory. Term used for a variety of systems in the brain with different
characteristics. In all cases, however, it implies the ability to reinvoke or
repeat a specific mental image or a physical act. It is a system property that depends on changes in synaptic strengths. (Edelman 2004, ‘Glossary’, 166).
In this chapter on Bergson and memory I shall focus on two key questions which Bergson sought to establish as the foundation for a philosophical treatment of memory:
1) What is the relation between past and present? Is it merely a difference in degree, or it possible to locate the difference between them as one of kind? If we can do the latter, what will this reveal about memory?
2) What is the status of the past? Is it something merely psychological or might it be possible to ascribe an ontological status to it? In other words, what is the reality of the past?
Matter and Memory (first published in 1896) is widely recognised as Bergson’s major work. William James, a great admirer of Bergson’s work, described it as effecting a Copernican Revolution that was comparable to Kant's Critique of Pure Reason. Although the text fell into neglect in the second half of the twentieth century, it exercised a tremendous influence on several generations of French philosophers, including Emmanuel Levinas, Maurice Merleau-Ponty, Jean-Paul Sartre, Paul Ricouer, and Gilles Deleuze. In addition, there have been important engagements with the text, and with the phenomenon of Bergsonism, in the writings of critical theorists such as Walter Benjamin, Theodor Adorno and Max Horkheimer. If Bergson’s texts are being rediscovered today this is largely as a result of the influence of Deleuze‘s writings on current intellectual work. The current interest being shown in Bergson is not, however, confined to fashionable developments in continental philosophy. Bergson is gaining a renewed presence in psychology and the philosophy of mind. I shall devote most of this chapter to an explication of the main ideas we encounter in Bergson’s text. In the final section I shall say something on the reception of Bergson’s ideas in some key strands of twentieth century thought.
Bergson’s approach to memory was highly innovative. He was one of the first thinkers to show the importance of paying attention to different types of memory (episodic, semantic, procedural), and he sought to provide a sustained demonstration of why memory cannot be regarded as merely a diluted or weakened form of perception. Bergson is close to Freud insofar as both are committed to the view that a radical division must be made between memory and perception if we are to respect the radical alterity of the unconscious. Bergson calls memory 'a privileged problem' precisely because an adequate conception of it will enable us to speak seriously of unconscious psychical states. In this respect Bergson anticipates the arguments Freud put forward four years later in the Interpretation of Dreams (see Freud 1976, 770ff.).1 In his text of 1966, Bergsonism, Deleuze contends that Bergson introduces an ontological unconscious over and above the psychological one, and it is this which enables us to speak of the being of the past and to grant the past a genuine existence. The past is not simply reducible to the status of a former present, and neither can it be solely identified with the phenomenon of psychological recollection (Deleuze 1988, ch. three). However, as one commentator has rightly noted, Bergson’s conception of the unconscious does not concern itself with the problems of psychological explanation that so occupied the attention of Freud (Gardner 2003, 112).
Bergson always sought to think time in terms of duration (durée), the preservation or prolongation of the past and entailing the co-existence of past and present. He insists that a “special meaning” is to be given to the word memory (MM 222). In one of the finest essays ever written on Bergson’s text, Jean Hyppolite notes that the new sense memory comes to have in Bergson consists in conceiving its operation in terms of a synthesis of past and present and with a view to the future. This goes against the prevailing conception which conceives memory as a faculty of repetition or reproduction, in which the past is repeated or reproduced in the present and is opposed to invention and creation. For Bergson memory is linked to creative duration and to sense. As Bergson notes, if matter does not remember the past since it repeats it constantly and is subject to a law of necessity, a being which evolves creates something new at every moment (223).
But just how are we to draw this distinction between past and present? Following Bergson we can note:
Nothing is less than the present moment, if we understand by this the indivisible limit which separates or divides the past from the future.
This, however, is only an 'ideal' present; the real, concrete, 'live' present is different and necessarily occupies a tension of duration. If the essence of time is that it goes by, that time gone by is the past, then the present is the instant in which it goes by. However, we cannot capture this present by conceiving it in terms of a mathematical instant (as a point in time).
Bergson’s thinking is focused on the problem of how to draw a distinction between past and present whilst recognising the indivisible continuity of durational time. He claims that whilst the distinction we make between our present and our past is not arbitrary it is “relative to the expanse of the field that our attention to life can embrace” (MM 151-2). If memory is a form of duration, then it is one with the impetus of consciousness itself (understood in the broad sense that Bergson gives to it as that which is bound up with discernment),2 and what in fact needs explaining is forgetting. Bergson’s problem, then, is how to account for the distinction between past and present in the context of our recognition of the indivisibility of duration. Later philosophies of temporality, including the work of Martin Heidegger and Gaston Bachelard, criticized Bergson for conceiving duration as cohesion, and so failing to develop an account of the separations and ruptures of time, including the ecstasies of past, present, and future. However, as Jean Hyppolite points out, Bergson’s second major work, Matter and Memory, was precisely an attempt to raise this problem and to resolve it (Hyppolite 2003, 113-14). In his Huxley lecture of 1911 on life and consciousness Bergson makes it clear that consciousness is both memory (the conservation and accumulation of the past in the present) and anticipation of the future (ME 8).
Bergson's treatment of memory is not without difficulties or problems. But it is a valuable resource for mapping memory, and in this chapter I wish to explicate its novel and distinctive features. As we shall see, Bergson's presentation contains some highly unusual and unorthodox aspects, at least when one first encounters them and struggles to give them a sense.
Matter and Memory In Matter and Memory Bergson seeks to establish the ground for a new rapport between the observations of psychology and the rigours of metaphysics (by metaphysics Bergson means that thinking which endeavours to go beyond the acquired and sedimented habits of the human mind, which for him are essentially mechanistic and geometrical in character). His argument on memory is not advanced in abstraction from consideration of work done on mental diseases, brain lesions, studies of the failures of recognition, insanity and the whole pathology of memory. He poses a fundamental challenge to psychology in seeking to show that memories are not conserved in the brain. We have to hear him carefully on this point. In not wishing to privilege the brain as the progenitor of our representations of the world Bergson shows that he has an affinity with phenomenological approaches. He conceives perception and memory, for example, in the context of the lived body, conceives of cognition as fundamentally vital not speculative, and grants primacy to action or praxis in our relation to the world.
Bergson's argument rests on two hypotheses being put to work: pure perception and pure memory. Imagine a perception without the interlacing of memory (impossible but helpful). Imagine a memory that is not actualised in concrete and specific memory-images and thus not reducible to our present recollection: less impossible perhaps but equally helpful. The central claim of the book is that while the difference between matter and perception is one of degree the difference between perception and memory is one of kind. Regarding the first: unless we see it in this way the emergence of perception out of matter becomes inexplicable and mysterious. Regarding the second: unless we see it this way then memory is deprived of any unique and autonomous character and becomes simply a weakened form of perception (indeed Locke called it a 'secondary perception'). Bergson’s argument for the autonomy of memory is twofold:
a thesis on the active character of perception, the interest of which is vital and not speculative. In cases of failed recognition it is not that memories have been destroyed but rather that they can no longer be actualised because of a breakdown in the chain that links perception, action, and memory.
an argument from time conceived as duration: independent recollections cannot be preserved in the brain, which only stores motor contrivances, since memories are 'in' time, not in the brain which is seated in the present. Since memories concern the past (which always persists and exists in multiple modes), an adequate thinking of memory must take the being of memory seriously.
It is as if Bergson is saying: memory is not in the brain but rather 'in' time, but time is not a thing, it is duration, hence nothing can be 'in' anything. Hence his argument, curious at first, that when there takes place a lesion to the brain it is not that memories are lost, simply that they can no longer be actualised and translated into movement or action in time. Memory and psychological recollection are not the same. As Edward Casey has noted, the language of containment has taken a deep hold over our thinking on memory, whether it is the brain or the computer that provides the container which cribs and confines memory (Casey 2000, 310); but it is this language that Bergson attempted to expose as fundamentally flawed and to move beyond.
Bergson is concerned with the relation between the mental and cerebral and is keen to make such a distinction, simply because our psychical life, while bound to its motor accompaniment, is not governed by it. Rather, he argues that there are diverse tones, rhythms, and intensities of mental life. Our psychic life is lived at different tensions relative to the degree of our attention to life. Thus the relation of the mental to the cerebral is neither a simple nor a constant one. A psychical disturbance is to be explained on the basis of this conception of life: a disease of the personality can be understood in terms of an unloosening or breaking of the tie which binds psychic life to its motor accompaniment, which involves an impairing of attention to outward life. Bergson thus resists interpretations of disorders like aphasia in terms of a localisation of the memory-images of words. Bergson is not, of course, denying that there exists a close connection between a state of consciousness and the brain. His argument is directed against any reified treatment of the brain in separation from the world it is a part of and from 'life' treated as a sphere of praxis or activity. He thus argues against the idea that if we could penetrate into the inside of the brain and see at work the dance of the atoms which make up the cortex we would then know every detail of what is taking place in consciousness. The brain is in the world, not in the head, and it's only a small part of the life of the organism, the part which is limited to the present.
Bergson's starting point is to criticise the notion of some detached, isolated object, such as the brain, as the progenitor of our representation of the world. The brain is part of the material world. Thus, if we eliminate the image that is the material world we at the same time destroy the brain and its cerebral disturbances. The body is in the aggregate of the material world, an image which acts like all other images, receiving and giving back movement. The body is a centre of action and not a house of representation. It exists as privileged image in the universe of images in that it can select, within limits, the manner in which it shall restore what it receives (for insight into Bergson‘s theory of images see Sartre 1962, and more recently, Moore 1996, and Casey 2004). The nervous system, Bergson argues, is not an apparatus that serves to fabricate or even prepare representations of the world. Its function, rather, is to receive stimulation, to provide motor apparatus, and to present the largest possible number of such apparatuses to a given stimulus. The brain is thus to be regarded as an instrument of analysis with regard to a received movement and an instrument of analysis with regard to executed movement. Its office is to transmit and divide movement. Let us posit the material world as a system of closely-linked images and then imagine within it centres of action represented by living matter - that is, matter which is contractile and irritable – and around these centres there will be images that are subordinated to its position and variable with it. This is how we can understand the relation between matter and its perception and the emergence of conscious perception. Matter, therefore, can be approached in terms of the aggregate of images; the perception of matter is these same images but referred to the eventual (possible or virtual) action of one particular image, my body. It is not, therefore, a question of saying simply that our perceptions depend upon the molecular movements of the cerebral mass; rather, we have to say that they vary with them, and that these movements remain inseparably bound up with the rest of the material world. We cannot conceive of a nervous system living apart from the organism which nourishes, from the atmosphere in which the organism breathes, from the earth which that atmosphere envelopes, and so on.
Bergson insists: 'There is no perception which is not full of memories' (MM 33). With the immediate and present data of our senses we mingle a thousand details out of our past experience. Why does he use the hypothesis of an ideal perception? He comes up with the idea of an impersonal perception to show that it is this perception onto which are grafted individual accidents and which give an individual 'sense' to life; owing to our ignorance of it, and because we have not distinguished from it memory, we are led to conceive of perception mistakenly as a kind of interior, subjective vision which then differs from memory simply in terms of its greater intensity. At the end of chapter one Bergson turns his attention memory and insists that the difference between perception and memory needs to be made as a difference in kind. He fully acknowledges that the two acts, perception and recollection, always interpenetrate each other and are always exchanging something of their substance as by a process of endosmosis. So, why does he insist on drawing the difference as one of kind? He has a number of reasons: first and foremost, to make the difference between past and present intelligible and to ascribe a genuine ontological character to the past (the past is real in its past-ness); to develop an adequate understanding of the phenomenon of recognition (in what situations does my body recognise past images?); and finally, to explain the mechanism of the unconscious.
So, what is Bergson going to claim about memory? First, that in actuality memory is inseparable from perception; it imports the past into the present and contracts into a single intuition many moments of duration, ‘and thus by a twofold operation compels us, de facto, to perceive matter in ourselves, whereas we, de jure, perceive matter within matter' (MM 73). Second, whilst the cerebral mechanism conditions memories, it is not sufficient to ensure their survival or persistence.
The Types of Memory
In the opening argument of chapter two Bergson addresses what he regards as the two main types of memory. Only the second, what he calls independent recollection, can be called memory proper.
The essential dimension of the body is activity, specifically adaptation in the present (solving a problem, overcoming an obstacle in the environment). It is only in the form of motor contrivances that the action of past can be stored up. Past images are preserved in a different manner. The past survives, then, under two distinct forms: in motor mechanisms and in independent recollections. Both serve the requirements of the present. The usual or normal function of memory is to utilise a past experience for present action (recognition), either through the automatic setting into motion of mechanism adapted to circumstances, or through an effort of the mind which seeks in the past conceptions best able to enter into the present situation. Here the role of the brain is crucial: it will allow only those past images to come into being or become actualised that are deemed relevant to the needs of the present. A lived body is one embedded in a flux of time, but one in which it is the requirements of the present that inform its constant movement within the dimension of the past and horizon of the future. If the link with reality is severed, in this case the field of action in which a lived body is immersed, then it is not so much the past images that are destroyed but the possibility of their actualisation, since they can no longer act on the real: 'It is in this sense, and in this sense only, that an injury to the brain can abolish any part of memory' (MM 79).
Let's consider in a little more detail how Bergson conceives the contraction of the past taking place as a way of addressing the present. Here I draw on the helpful account provided by Patrick McNamara. When a level of the past gets contracted the contraction is experienced by present consciousness as an expansion, simply because its repertoire of images and moments of duration are increased and intensified (McNamara 1999, 37). Memory enables us to contract in a single intuition multiple moments of time. In this way it frees us from the movement of the flow of things and from the rhythm of mechanical necessity. The activation of memory involves a series of phases. Firstly, there is a relaxation of the inhibitory powers of the brain; this is followed by a proliferation of memory-images that can flood the cognitive system; and then, finally, there takes place a selection phase in which the inhibitory processes are once again called upon. The proliferation of images opens up a plurality of possible states of affairs and possible worlds; the process of actualization, however, requires contraction take place in order to contextualize a cue and provide an adequate response to the problem in the environment that has been encountered. What is selected may not, however, be the 'best match or the most optimal solution to a current perception' (38). Bergson does not subscribe to a straightforwardly Darwinian model of the selection process at work in memory.
Bergson's theory of memory rests on understanding these contractions and expansions in relation to the syntheses of past and present. However, our grasp of this theory remains inadequate so long as we do not appreciate its addition of a third term, that of pure memory. Bergson provides in fact a tripartite theory with a 'pure memory' advanced alongside those of habit- and representational-memory. How do we arrive at this third term of memory?
When we learn something a kind of natural division takes place between the contractions of habit and the independent recollection of events that involve dating. If I wish to learn a poem by heart I have to repeat again and again through an effort of learning, in which I decompose and recompose a whole. In the case of specific bodily actions and movements habitual learning is stored in a mechanism that is set in motion by some initial impulse and that involves releasing automatic movements within a closed system of succession and duration. The operations of independent recollection are altogether different. In the formation of memory-images the events of our daily life are recorded as they take place in a unique time and providing each gesture with a place and a date. This past is retained regardless of its utility and practical application. The past is preserved in itself and, at the same time, contracted in various states by the needs of action that are always seated in an actual present. This repetition of memory-images through action merits the ascription of the word memory not because it is involved in the conservation of past images but rather because it prolongs their utility into a present moment. The task of this kind of memory is to ensure that the accumulation of memory-images is rendered subservient to praxis, making sure that only those past images come into operation that can be coordinated with a present perception, and so enabling a useful combination to emerge between past and present images: 'Thus is ensured the appropriate reaction, the correspondence to environment - adaptation, in a word - which is the general aim of life' (MM 84). Without this coordination of memory-images by the adaptive consciousness the practical character of life would be distorted and the plane of dreams would mingle with the plane of action (in fact, as Bergson fully concedes, the planes do communicate and cannot be treated as isolable dimensions of consciousness and unconsciousness; the issue is rather to be approached in terms of different tensions and situations of lived time).
The pure past - by which is simply meant the preservation of the past independent of its actualisation in a present - is inhibited from freely expressing itself by the practical bent of our bodily comportment, 'by the sensory-motor equilibrium of a nervous system connecting perception with action' (MM 95). Not only is there more than one kind of memory, but memory-images enjoy more than the one kind of existence, being actualized in multiple ways: 'Memory thus creates anew the present perception, or rather it doubles this perception by reflecting upon it either its own image or some other memory-image of the same kind' (101). Our life moves - contracts, expands, and relaxes - in terms of circuits and it is the whole of memory that passes over into each of these circuits but always in a specific form or state of contraction and in terms of certain variable dominant recollections: 'The whole of our past psychical life conditions our present state, without being its necessary determinant' (148). We shift between virtual and actual states all of the time, never completely virtual or completely actual.
Bergson holds that perception and memory interlace and all memories must become actualized in order to become effectively real (127). Personal recollections make up the largest enclosure of our memory. He writes: 'Essentially fugitive, they become only materialized by chance, either when an accidentally precise determination of our bodily attitude attracts them or when the very indetermination of that attitude leaves a clear field to the caprices of their manifestation' (106). The pathology of memory has its basis in an appreciation of the vitality of memory. Memory, Bergson argues, has 'distinct degrees of tension or of vitality'. Pathology confirms this insight: 'In the "systematized amnesias" of hysterical patients', he writes, 'the recollections which appear to be abolished are really present, but they are probably all bound up with a certain determined tone of intellectual vitality in which the subject can no longer place himself' (170). He further notes that there are always dominant memories for us, which exist as 'shining points round which the others form a vague nebulosity' (171). These shining points get multiplied to the extent to which our memory is capable of expansion. The process of localising a recollection in the past does not consist in simply plunging into the mass of our memories as into a bag in order to draw out memories closer and closer to each other and between which the memory to be localised may find its place. Again, he finds helpful the pathology of memory:
In retrogressive amnesia, the recollections which disappear from consciousness are probably preserved in remote planes of memory, and the patient can find them by an exceptional effort like that which is effected in the hypnotic state. But, on the lower planes, these memories await, so to speak, the dominant image to which they may be fastened. A sharp shock, a violent emotion, forms the decisive event to which they cling; if this event, by reason of its sudden character, is cut off from the rest of our history, they follow it into oblivion (171).
In short, Bergson has posited an assemblage made up of three components: pure memory, memory-images, and perception. The latter is never simply a contact of the mind with a present object but is impregnated with memory-images; in turn these images partake of a pure memory that they materialize or actualise and are bound up with the perceptions which provide it with an actual embodiment.
Perception and Memory
It is necessary to dispel a number of illusions that shape and govern our thinking about memory, a key one being that memory only comes into existence once an actual perception has taken place. This illusion is generated by the requirements of perception itself, which is always focused on the needs of a present. While the mind or consciousness is attending to things it has no need of pure memory which it holds to be useless. Moreover, although each new perception requires the powers afforded by memory, a reanimated memory appears to us as the effect of perception. This leads us to suppose that the difference between perception and memory is simply one of intensity or degree, in which the remembrance of a perception is held to be nothing other than the same perception in a weakened state, resulting in the illegitimate inference that the remembrance of a perception cannot be created while the perception itself is being created or be developed at the same time (Bergson ME 160-1).
It is by recognising the virtual character of pure memory that we can perhaps better appreciate that the difference between perception and memory is one of kind and not merely degree. Memory is made up of memory-images but the recollection of an image is not itself an image (it is closer to a concentrated act of intellectual effort). Bergson insists that 'To picture is not to remember' (Imaginer n'est pas se souvenir) (MM 135). As a recollection becomes actual it comes to live in an image, 'but the converse is not true, and the image, pure and simple, will not be referred to the past unless, indeed, it was in the past that I sought it…' (ibid.).
Bergson's claim is that at every moment of our lives we are presented with two aspects, even though the virtual aspect may be imperceptible owing to the very nature of the operations of perception:
Our actual existence, then, whilst it is unrolled in time, duplicates itself all along with a virtual existence, a mirror-image. Every moment of our life presents two aspects, it is actual and virtual, perception on the one side and memory on the other. Each moment is split up as and when it is posited. Or rather, it consists in this very splitting, for the present moment, always going forward, fleeting limit between the immediate past which is now no more and the immediate future which is not yet, would be a mere abstraction were it not the moving mirror which continually reflects perception as a memory (ME, 165).
It is because the past does not simply follow the present but coexists with it that we can develop an explanation of paramnesia or the illusion of déja-vu, in which there is a recollection of the present contemporaneous with the present itself. The illusion is generated from thinking that we are actually undergoing an experience we have already lived through when in fact what is taking place is the perception of the duplication we do not normally perceive, namely, of time into the two aspects of actual and virtual. There is a memory of the present in the actual moment itself. I cannot actually predict what is going to happen but I feel as if I can: what I foresee is that I am going to have known it - I experience a 'recognition to come', I gain insight into the formation of a memory of the present (if we could stall the movement of time into the future, this experience would be much more common for us; we can note that current empirical research on the phenomenon of déja-vu focuses on the regions of the brain involved in producing it and explains it in terms of gaps in our attentive system).
This difference between past and present can be explained in the following terms: our present is the 'very materiality of our existence' in the specific sense that it is 'a system of sensations and movements and nothing else' (MM 139). This system is unique for each moment of duration 'just because sensations and movements occupy space, and because there cannot be in the same place several things at the same time' (ibid.). One's present at any moment of time is sensory-motor, again in the specific sense that the present comes from the consciousness of my body: actual sensations occupy definite portions of the surface of my body. The concern of my body, manifest in the consciousness I have of it, is with an immediate future and impending actions. By contrast one's past is 'essentially powerless' in the specific sense that it interests no part of my body conceived as a centre of action or praxis. No doubt, Bergson notes, it begets sensations as it materializes, but when it does so it ceases to be a memory and becomes something actually lived by passing into the condition of a present thing. In order for such a memory to become materialized as an actual present I have to carry myself back into the process by which I called it up, 'at it was virtual, from the depths of my past'. Bergson insists that this pure memory is neither merely a weakened perception nor simply an assembly of nascent sensations. When conceived in terms of the latter, memory becomes little more than the form of an image contained in already embodied nascent sensations. Let us once again clarify the difference between the present and the past: it is because they are two opposed degrees that it is possible to distinguish them in nature or kind.
Bergson's innovation, then, is to suggest that a recollection is created alongside an actual perception and is contemporaneous with it: 'Either the present leaves no trace in memory, or it is twofold at every moment, its very up-rush being in two jets exactly symmetrical, one of which falls back towards the past whilst the other springs forward towards the future' (ME 160). The illusion that memory comes after perception arises from the nature of practical consciousness, namely, the fact that it is only the forward-springing jet that interests it. Memory becomes superfluous and without actual interest: 'In a general way, or by right, the past only reappears to consciousness in the measure in which it can aid us to understand the present and to foresee the future. It is the forerunner of action' (175). Because consciousness is bound up with an attentiveness to life, to action, it 'only admits, legally' those recollections which provide assistance to the present action (177). This explains Bergson's interest in the anomalies (illegalities) of the life of ésprit, such as deliriums, dreams, hallucinations, etc., which, Bergson insists, are 'positive facts' that consist in the presence, and not in the mere absence, of something: 'They seem to introduce into the mind certain new ways of feeling and thinking' (151).
The past can never be recomposed with a series of presents since this would be to negate its specific mode of being. To elaborate an adequate thinking of time, including the time of the present, requires that we make the move to an ontological appreciation of the past. Psychological consciousness is born and emerges into being only when it has found its proper ontological conditions. On this movement Bergson writes:
Whenever we are trying to recover a recollection, to call up some period of our history, we become conscious of an act sui generis by which we detach ourselves from the present in order to replace ourselves, first, in the past in general, then, in a certain region of the past - a work of adjustment like the focusing of a camera. But our recollection still remains virtual…' (MM 133-4).
In short, we cannot reconstitute the past from the present but must make the move into the past itself as a specific region of being. The past will never be comprehended as something past unless we follow and adopt the movement by which it expands into a present image, and this movement by definition is something virtual: 'In vain do we seek its trace in anything actual and already realized; we might as well look for darkness beneath the light' (135). Bergson contends that this is, in fact, one of the chief errors of the school of associationism, which dominated the study of memory in the second half of the nineteenth century: 'placed in the actual, it exhausts itself in vain attempts to discover in a realized and present state the mark of its past origin, to distinguish memory from perception, and to erect into a difference in kind that which it condemned in advance to be but a difference of magnitude' (ibid.) What is in need of explanation is not so much the cohesion of internal mental states but rather 'the double movement of contraction and expansion by which consciousness narrows or enlarges the development of its content' (MM 166). Associationism conceives the mechanism of linkage in terms a perception remaining identical with itself, it is a 'psychical atom which gathers to itself others just as these happen to be passing by' (165). In Bergson's model of recollection, however, the linkages and connections forged by the mind are not simply the result of a discrete series of mechanical operations. This is because within any actual perception it is the totality of recollections that are present in an undivided, intensive state. If in turn this perception evokes different memories:
…it is not by a mechanical adjunction of more and more numerous elements which, while remaining unmoved, it attracts round it, but rather by an expansion of the entire consciousness which, spreading out over a larger area, discovers the fuller details of its wealth. So a nebulous mass, seen through more and more powerful telescopes, resolves itself into an ever greater number of stars (165-6).
The first hypothesis, which rests on a physical atomism, has the virtue of simplicity. However, the simplicity is only apparent and it soon locks us into an untenable account of perception and memory in terms of fixed and independent states. It cannot allow for movement within perception and memory except in artificially mechanical terms, with memory-traces jostling each other at random and exerting mysterious forces to produce the desired contiguity and resemblance (for a critique of Hume‘s early associationist account of memory see Mounce 1999, 30, and compare Deleuze 1991, 93ff., and Biro 1993, 50). Bergson's theory of memory in terms of pure memory, memory-images, and actual perception, is designed to provide a more coherent account of how associations actually take place and form in the mind.
We find ourselves, largely out of force of habit, compelled to determine or ascertain the place or space of memory: where is it? How can the past, which has ceased to be, preserve itself if not in the brain? Bergson is not denying that parts of the brain are playing a crucial role in our capacity for memory and in the actualization of memory. Memories cannot be in the brain (except habit-memory), because the brain occupies only a small slice or section of becoming, namely, the present: 'The brain, insofar as it is an image extended in space, never occupies more than the present moment: it constitutes, with all the rest of the material universe, an ever-renewed section of universal becoming'. Moreover, the difficulty we have in conceiving the survival of the past - which has ceased to be useful but not ceased to be - comes from the fact that
…we extend to the series of memories, in time, that obligation of containing and being contained which applies only to the collection of bodies instantaneously perceived in space. The fundamental illusion consists in transferring to duration itself, in its continuous flow, the form of the instantaneous sections which we make in it. (MM 149).
Our reluctance to admit the integral survival of the past has its origin in the very bent of our psychical life - 'an unfolding of states wherein our interest prompts us to look at that which is unrolling, and not at that which is entirely unrolled' (150). As Deleuze points out in Bergsonism, the question ‘where are recollections preserved?’, involves a false problem by supposing a badly analysed composite (Deleuze 1988, 54). Why suppose that memories have to be preserved somewhere? Furthermore, a fundamental feature of Bergson’s novel empiricism is to insist on their being different ‘lines of fact’; as Deleuze insists, whereas the brain is situated on the line of ‘objectivity’, recollection is part of the line of ‘subjectivity’. It is thus ‘absurd to mix the two lines by conceiving of the brain as the reservoir or the substratum of recollections’ (ibid.). For Bergson memory is primarily affective, and as soon as we attempt to isolate the affects of memory, setting out time in space and confusing the different lines of fact, they become lifeless.
Reception and Influence Bergson has a relation to phenomenology and was read by all the key figures that contributed to this school in the twentieth century. Although there are no references to Bergson in the work of Husserl he was aware of Bergson’s contributions and, in spite of their differences in method and ultimate theoretical commitments, there are parallels between the two thinkers in how they conceptualise time and memory (see Husserl 1964, 1973; and see the remark Deleuze makes on Bergson’s relation to Husserl‘s phenomenology in the afterword to the English translation of his Bergsonism). Important engagements with Bergson’s thinking on time and memory can be found in the work of Levinas (1987), Merleau-Ponty (1989, 2001), and Sartre (1962, 1989), each one of whom made a seminal contribution to phenomenology. The main criticism made of Bergson by the likes of Merleau-Ponty and Sartre is that he is unable to adequately account for the intentional structure of consciousness and, as a result slides back into a pre-phenomenological realism (Merleau-Ponty 2001, 89-90 and Sartre 1962, 39-40). In his study of 1953 a young Jean-Francois Lyotard argued that phenomenology separates itself from Bergsonism on the question of time by replacing a flowing time in consciousness with a consciousness that positively constitutes time for itself. This critique of Bergson has been challenged in recent theoretical work where he is seen as having closer affinities with post-phenomenological notions of agency and subjectivity to be found, for example, in the work of so-called post-structuralist figures such as Derrida and Deleuze (see, for example, Game 1991). Bergson’s work, especially Matter and Memory, is seen as containing valuable resources for calling into question the primacy of the ‘For-Itself’ and its idealistic stress on the unitary and transparent character of self-consciousness (this move is prefigured in the work of Levinas, 1983; Sartre‘s reading of Bergson was effectively challenged by Hyppolite in his essay of 1949, see Hyppolite 2003, especially 121-5). On this point Levinas wishes to go as far as underlining the importance of Bergsonism 'for the entire problematic of contemporary philosophy’ on account of the fact that it is no longer a thought of a 'rationality revealing a reality which keeps to the very measure of a thought'. In effecting a reversal of traditional philosophy by contending the priority of duration over permanence, Bergson has provided thought with 'access to novelty, an access independent of the ontology of the same' (Levinas 1987, 132).
Walter Benjamin is one thinker to have appreciated the rich character of Bergson’s treatment of memory and its significance for our understanding of certain critical aspects of modernity. In his essay ‘On Some Motifs in Baudelaire’, first published in 1939 in the Zeitschrift für Sozialforschung, in which he develops a wide-ranging treatment of Proust, Freud, Baudelaire, the disintegration of the aura and the shock experience, he situates Bergson’s text in the context of attempts within philosophy to lay hold of the ‘“true” experience’ in opposition to the manufactured kind that manifests itself in the ‘standardized, denatured life of the civilized masses’. For Benjamin, Bergson’s ‘early monumental work’, as he describes it, towers above the body of work associated with the philosophy of life of the late nineteenth century - he mentions the work of Wilhelm Dilthey - on account of its links with empirical research and the richness of its account of the structure of memorial experience (Benjamin 1973, 159). Bergson’s text needs to be taken to task, however, on account of its failure to understand its own historical conditions of possibility and reflect on its historical determinations. On this issue Benjamin goes on to note some important differences in the figuration of the experience of memory we find in Bergson’s text and in Proust’s great modern novel, In Search of Lost Time. Benjamin contends that Bergson’s conception of durée is estranged from history (186), and this point informs Horkheimer’s critical engagement with Bergson. Horkheimer acknowledges that he owes ‘decisive elements’ to Bergson’s philosophy for his own thinking, but argues that Bergson offers a metaphysics of time that privileges an interior spiritual world, rests on a disavowal of human history, and suffers from a ‘biological realism’ (Horkheimer 2005; on Bergson and the idea of ‘historical duration’ see also Hyppolite 2003, 114).
It is interesting to note that the critical reception of Bergson we find in the work of critical theorists such as Horkheimer, is similar to that we find in phenomenology, namely, that his thinking on memory is seen to grant too much importance to its contemplative aspects over its critical and intentional ones. For phenomenologists this manifests itself in an alleged failure to account for the synthesising powers of an intentional subject (Bergson grants intention to memory itself over and above the subject; the subject is implicated in memory; ‘subjectivity is never ours, it is time…the virtual’, as Deleuze puts it, 1989, 82-3). For critical theorists, by contrast, it reveals itself in the failure to provide a constructivist, and activist, account of history and historical agency (Bergson is oblivious, Horkheimer says, to the meaning of theory for historical struggle). To what extent these criticisms are fair, and to what extent they have been called into question by more recent intellectual developments, are questions that cannot be treated here. I would simply point out that Bergson set himself a specific task in Matter and Memory - taking the psychology of his day to task on account of what he regarded as its inadequate and impoverished approach to the life of memory - which, to a large extent, he fulfilled and admirably so, and it is necessary to respect the integrity of his project (which is not to say that all kinds of critical questions cannot, and should not, be asked of it). It is quite clear that Bergson’s heart lies not with contemplation but with creative action. His complaint is there is too much contemplation in philosophy. In his prescient final text The Two Sources of Morality and Religion, published in 1932, Bergson pays homage to those great spiritual and ethical leaders, from Christian saints to social revolutionaries, who have brought something dynamically new into existence and helped to push humanity forward.
In terms of recent work in psychology and the philosophy of mind, Bergson’s work has been positively received in some quarters and is seen to provide a set of rich resources for thinking memory beyond simple-minded mechanical models of mind and memory. The neurologist Oliver Sacks often cites Bergson's ideas in support of his call for a neurology of identity, which would move away from a rigid physicalist paradigm, centred on notions of algorithm and template, and that supposes notions of rigid cerebral localisation and a rigidly programmed cerebral machine, towards a neurology able to match the 'richness and density of experience', what he calls its sense of scene and music, its 'ever-changing flow of experience, of history, of becoming' (Sacks 1991, 178). More substantially, Patrick McNamara puts Bergson’s ideas on mind and memory to instructive and productive use in his important study, Mind and Variability: Mental Darwinism, Memory, and Self(1999), whilst the attempt by Israel Rosenfield in his The Invention of Memory (1988) to expose the view that we can remember because we have fixed memory images permanently stored in our brains for what it is - a myth (that of localisation) - continues the work Bergson began over a century ago. This is echoed in NcNamara’s more recent study, as when he writes for example: ‘The representational-instructionist view of memory is still what I would call the modern standard view of the nature of memory. It and its related “trace theory” of how the brain “stores” memory constitute the background assumptions of much of modern research into memory’ (1999, 23). In his book, Memory, History, Forgetting, one of the most important studies of memory in recent years, Paul Ricoeur acknowledges the original and innovative character of Bergson’s thinking on memory. For Ricoeur, Bergson is ‘the philosopher’ to have best understood the close connection between the ‘survival of images’ and the phenomenon of recognition (Ricoeur 2004, 430-1). Furthermore, with this insight into the survival of images, which require that we acknowledge that memory has the character of endurance, Ricoeur believes that Bergson’s thinking holds the resources required for understanding the working of forgetting, even if Bergson himself was only able to think this in terms of effacement. It is the self-survival of images that can be considered as a figure of fundamental forgetting. Ricoeur poses the question, ‘On what basis, then, would the survival of memories be equivalent to forgetting?’ (ibid., 440). His answer is to propose that forgetting be conceived not simply in terms of the effacement of traces, but rather in terms of a reserve or a resource: ‘Forgetting then designates the unperceived character of the perseverance of memories, their removal from the vigilance of consciousness’ (ibid.). On this conception forgetting can be understood not simply as an inexorable destruction, but as an immemorial resource.
Bergson’s great text is significant for a number of reasons, including its attempts to demonstrate the ontological status of the past, to provide a genuinely dynamical model of memory’s operations, to show the virtual character of (pure) memory, and, finally, its advancement of the argument that memory is not simply the mechanical reproduction of the past but sense. Without memory life is, quite literally, devoid of meaning. Matter and Memoryis a text we are still catching up with.
Abbrevations Used for Bergson ME Mind-Energy, trans. H. Wildon Carr (New York, Henry Holt, 1920). This
volume of essays now exists as a critically annotated edition published by
Palgrave Macmillan (2007).
MM Matter and Memory, trans. N. M. Paul & W. S. Palmer(New York, Zone
Bibliography Benjamin, Walter, ‘On Some Motifs in Baudelaire’, Illuminations, trans. H. Zohn
(London, Collins, 1973), pp. 157-202.
Biro, John (1993), ‘Hume’s new science of the mind’, in David Fate Norton (ed.), The
Cambridge Companion to Hume (Cambridge & New York, Cambridge
University Press), pp. 33-64.
Casey, Edward S. (2000), Remembering (Bloomington, Indiana University Press).
--- (2004), ‘Image and Memory in Bachelard and Bergson’, in Spirit and Soul: Essays
in Philosophical Psychology, second and expanded edition (Putnam,
--- (1989), Cinema 2: The Time-Image, trans. Hugh Tomlinson & Robert Galeta
(London and New York, Continuum Press).
--- (1991), Empiricism and Subjectivity: An Essay on Hume’s Theory of Human
Nature, trans. Constantin V. Boundas (New York, Columbia University
Edelman, Gerald (2004), Wider than the Sky: The Phenomenal Gift of Consciousness
(London, Allen Lane).
Freud, Sigmund (1976), The Interpretation of Dreams, ed. James Strachey
Game, Ann (1991), Undoing the Social: Towards a Deconstructive Sociology (Milton
Keynes, Open University Press).
Gardner, Sebastian (2003), ‘The unconscious mind’, in Thomas Baldwin (ed.), The
Cambridge History of Philosophy 1870-1945 (Cambridge, Cambridge
Gross, David (1985), ‘Bergson, Proust, and the Revaluation of Memory’,
International Philosophical Quarterly, 25 (4), pp. 369-80.
Horkheimer, Max (2005), ‘On Bergson’s Metaphysics of Time’, trans. Peter Thomas,
Radical Philosophy 131, pp. 9-20 (originally published as ’Zu Bergsons
Metaphysik der Zeit’, Zeitschrift für Sozialforschung, Heft 3, 1934, pp. 321-
Husserl, Edmund (1964, The Phenomenology of Internal Time Consciousness, ed. M.
Heidegger, trans. J. S. Churchill (The Hague, Martinus Nijhoff; based on
lecture courses 1893-1917).
--- (1973), Experience and Judgment: Investigations in a Genealogy of Logic, trans. J.
S. Churchill and K. Ameriks (London, Routledge, Kegan & Paul).
Hyppolite, Jean (2003), ‘Various Aspects of Memory in Bergson’ (1949), trans.
Athena V. Colman, in Lawlor, The Challenge of Bergsonism, pp. 112-28.
Lawlor, Leonard (2003), The Challenge of Bergsonism (London and New York,
Levinas, Emannuel (1983), ‘Beyond Intentionality’, in Alan Montefiore (ed.),
Philosophy in France Today (Cambridge & New York, Cambridge University
--- (1987), Time and the Other, trans. R. A. Cohen (Pittsburgh, Duquesne University
Locke, John (1976), An Essay Concerning Human Understanding (London, Dent).
McNamara, Patrick (1999), Mind and Variability: Mental Darwinism, Memory, and
Self (London, Praeger).
Merleau-Ponty, Maurice (2001) The Incarnate Subject: Malebranche, Biran, and
Bergson on the Union of Body and Soul, trans. Paul B. Milan (New York,
Humanity Books) (lecture course given at the Ecole Normale Supérieure,
--- (1989), Phenomenology of Perception, trans. C. Smith (London & New Yorik,
Routledge; originally published 1945).
Moore, F. C. T. (1996), Bergson. Thinking Backwards (Cambridge & New York,
Cambridge University Press).
Mounce, H. O. (1999), Hume’s Naturalism (London and New York, Routledge).
Mullarkey, John (1999), Bergson and Philosophy (Edinburgh, Edinburgh University
Ricoeur, Paul (2004), Memory, History, Forgetting, trans. Kathleen Blamey and
David Pellauer (Chicago and London, University of Chicago Press).
Rosenfield, Israel (1988), The Invention of Memory: A New View of the Brain (New
York, Basic Books).
Sacks, Oliver, A Leg to Stand On (London, Picador, 1991).
Sartre, Jean-Paul (1962), Imagination, trans. Forrest Williams (Ann Arbor, University
of Michigan Press; 1936).
--- Being and Nothingness: An Essay on Phenomenological Ontology (1989), trans.
Hazel E. Barnes (London & New York, Routledge).
1 In an essay of 1922 Bergson writes: ‘…my idea of integral conservation of the past more and more found its empirical verification in the vast collection of experiments instituted by the disciples of Freud’, The Creative Mind, trans. M. L. Andison (Totowa, Littlefield, Adams, & CO., 1965), p. 75.
2 This is very much in line with how neuroscientists frame consciousness today: ‘Consciousness reflects the ability to make distinctions or discriminations among huge sets of alternatives’ (Edelman 2004, 141).