Matter is an aggregate of images.  An image is ‘a certain existence which is more than that which the idealist calls a representation, but less than that which the realist calls a thing—an existence placed halfway between the “thing” and the “representation”.  This conception of matter is simply that of common sense’ (vii—viii).  The notion of the image works both ways, to deny both idealism and simple materialism. [In practice, the debate will be between idealist/psychological/phenomenological and material/physiological accounts of memory andperception. Much of the discussion tunrs on how we can interpret the detailed 'facts' provided by accounts of various pathologies such as 'word blindness'  -- I'm going to pass over these]. We already have the notion of a relation of movement replacing static binary divisions [and again this is going to be developed in terms of attempts to locate specific mental functions in separate centres of the brain versus accounts of the processes by which, for example, memories come to fuse with perceptions].

Ch 1.  Of the selection of images for conscious presentation.  What our body means and does

We live in the middle of the set of images, but one is given a particular privilege—my body and its affections[which gives knowledge from within] , which operate between receiving stimuli from outside and executing action.  Everything turns toward action.  Habitual or automatic action requires no conscious affection.  The body can be understood as transmitting stimuli along nerves to the central processing system, and then redistributing them back in the form of action.  However, nerves and brains are still images, so there is no mechanism.  At the same time, the brain is an image like all the others, part of the external world [only a part though, so there is an external world beyond brain activity].  However, bodies are privileged in that they can act on the other images, and this places our body as something that sees the world instrumentally, with the other images organized in zones according to their distance, their ability to affect my body and vice versa, the 'possible action upon them' (7).  Perception is a matter of referring images of the external world to  'the eventual action of one particular image, my body' (8).  Thus perception is a function of movement. 

Again this is more than a matter of information processing or representing the world.  The brain does not initiate representation, which is itself instrumental and connected to action.  Representations permit 'virtual acts' of the body (10), producing 'nascent' reactions.  In general, material objects themselves are not isolated, but each 'borrows its physical properties from the relations which it maintains with all others' (11-12), and this variation is related to perceptions [as a whole to a part].  The problem is to see how the images of the external world, which vary in themselves, connect to the way in which the body reduces the variation in order to act.  Whether or not objects have some real existence is a typical false problem, whereas seeing everything in terms of images is at least common ground between materialists and idealists: the difference in emphasis is whether the external images determine the subjective ones or vice versa.  However, materialists have to allow for [subjective] perception, while idealists have to ignore the independent variation of the external images [later rendered as 'the system of images which has no center', 15], especially when considering continuities between past or future.  In neither case does 'every change [give] the exact measure of its cause'.  In both cases, some extra process has to be invoked, like the ones whereby consciousness becomes epiphenomena, perception a mere accident, or you have to assume some pre established 'harmony between things and mind', so that scientific understanding becomes a mere accident [the form of this argument appears throughout, that reductionism requires some additional extra process].

Both systems also assume that the function of perception is to acquire pure knowledge, in a 'wholly speculative interest' (17).  However, perception is always tied to reaction, even if it is delayed, and even if human beings can distinguish between automatic and voluntary activity.  This distinction, and the spiritualized notion of consciousness that often follows is 'merely a difference of complication and not a difference in kind'(18).  Again the issue is connected to representation, which is sometimes seen as a special result of some magical process of transforming sense data.  There is choice arising from how to react to the same disturbance, but the brain is 'no more than a kind of central telephonic exchange'.

Action can vary in intensity, and this does affect the amount of conscious perception.  If immediate reaction is required, perception is limited, more or less to registering a contact.  With increased distance, and a wider 'zone of indetermination' (23) it is easier [and necessary] to consider interests and calculate advantage.  Distance produces the extent of 'indetermination of the act which is to follow'.  As a result 'perception is master of space in the exact measure in which action is master of time'[rather obscurely put --'function of' would be better than 'master of'?].  This indetermination is the consequence of conscious perception, rather more than is representation [as others argue]  [representation is explained later as a necessary stage in perception].

In practice, perception cannot be separated from memory.  Some theories of consciousness seem to want to operate with pure perception, however, focused only on the present and the task of understanding the external objects.  It is right to preserve the focus on the external object, and to deny that perception is some kind of 'interior and subjective vision' (25).  But the effects of memory provide the experience of subjectivity, reducing external moments to one internal moment.  Pure perception is useful only as an hypothesis.

Everything must be understood as a matter of images, although these are not always perceived or represented: but this is a matter of distance or interval.  We should understand representation of an image as reducing some of its qualities, for example in its relations with past and future versions of itself, and what actually 'fills it', as opposed to its 'external crust, its superficial skin'(28).  Present images are always more complex, and therefore more objective [because they connect fully with other aspects of reality].  Representation is therefore 'always virtual...  neutralized' and therefore unable to be actual.  [Confusing here, I think this means because representations have to be connected to other representations, rather than pursuing the objective connections of the object itself. Or that representation is only a matter of appearances again, too vague to be useful in action?].  The actual becomes the virtual representation by losing aspects of itself to become a picture, and this is also  governed by practical interest. 

This is how our perceptions arise [so they are representations here?].  The qualities emanating from the object are like light, reflected back onto the object thus making objects visible.  Our body at the centre reflects back those rays of light which interest it, and aspects that we do not influence are allowed to pass.  Objects are indifferent to each other and so can relate much more widely and fully with each other.  When their action is diminished in some way, say by some spontaneous reaction, we can form a representation of them [so representations are not directly related to practical interests, but something that we can use to acquire knowledge in a more indifferent way, 'reflected by our freedom' from the immediate impulse to action, in Bergson's terms, 29].  If there is full reflection of light, we get a luminous point, producing 'a virtual image which symbolizes...  The fact that the luminous rays cannot pursue their way' (30) [virtual here defined in optical terms, as a kind of 'mirage'?].  But if we have a center of activity, the real connections are reduced to a matter of 'virtual action', here meaning  'the eventual influence of the living being upon them' (30) [so virtual here means potential].

We have been describing differences of degree for images—there is no different in kind between objective being and being perceived, and all are part of the reality of matter.  Our representations depend on our 'possible action upon bodies', something which can be postponed, or chosen [and both imply reduction of the qualities of the object and its relations].  In more positive terms, this selection might be called 'discernment' (31).  We do not perceive in a kind of photographic way, partly again because objects are always in relation to each other.  Our own zones of indetermination act as a kind of screen to filter out elements of real action leaving only virtual [potential, conceivable?] action.

It is impossible to operate without some notion of the material world that is perceived virtually, and which can be broken down into isolated objects like bodies.  But it would be a mistake to fully isolate human bodies separated from the rest of reality, as we saw above.  Separating out particular aspects of reality only leaves the problem of having to join them back up again, like separated representations produced by consciousness have to be joined back to the material world.  [While we are here, trying to restrict movement to specified kinds makes it difficult to explain qualitative change].  Instead, the whole mechanism of perception involves movement from external images through to voluntary action, with perception as something limited, a reduction of the image of the whole.  This limitation is because perception always refers to some variable center with a degree of indetermination around the body.  This is what makes it look as if perception arises from within human beings in some way.  At its most general, this requires that 'we restore to movement the unity, indivisibility, and qualitative heterogeneity denied to it by abstract mechanics' (36).  To take an example, we might perceive the light coming from a particular point, but we must immediately assume it is an image, and that we are choosing it as part of our perception [a bit obscure and technical, 36-37) -- where it leads to is the fundamental importance of the 'sensori-motor process' in perception (37) with no direct transmission of vibrations from objects to us, but with a number of zones of indetermination both in perception and in subsequent action.  These aspects of indetermination might be sensibly ignored in natural science, but not in studies of human perception, [although we can borrow various black boxes in scientific expression].  We must also be aware of how these zones of indetermination appear to split up objects into say images and conscious images— in reality they  'form a single whole'.

It is certainly true that it looks as if the brain can produce its own images, as when we discuss dreams, but these really depend on the influence of memory.  [Then physiological arguments about various pathologies are used in support.]  There is continued insistence on the relation between objective images and our perceptions, the connectedness between perception and motor activity, and a growing sense in which the objective world can be seen as 'queries or demands addressed to my activity' (40).  The body appears to us as something invariable as it moves, 'a center to which I refer all the other images' (43), and this also confirms belief in the external world: it cannot simply be an extension of our own consciousness.  Consciousness mostly appears as choice and discernment, rather than attempts to reconstruct whole objects: a special kind of education of the senses is needed here to restore continuity.  The activities of such consciousness provide us with 'virtual action'[that is potential].

[More argument about the results of various physiological and psychological experiments]

[Affection is also discussed].  One view says that perceptions tends to become affections, such as pain, and as pain decreases, so does the perception of its cause: the result can be an external projection, a representation. Instead, affections too 'arise out of the image' (55).  We see this by connecting pain to the effort 'to set things right', although acting at a local level.  The whole organism has to repel the stimulation, which is the reverse of perception, a difference in kind, Bergson assures us (57).  That's because living things have to not just absorb forces from the outside but also struggle with them.  Perception shows the 'reflecting power of the body, affection measures its power to absorb' (57).  [There's also the idea of affection as filling up the interval between perception and action, prompting real action from all the virtual possibilities, 58].  Affection can therefore be seen as a kind of specialized perception, operating at the particular level of the surface between external images and the body.  It follows that affective states also emanate from external objects, representing actual effort on external images.  Real actions like this also compound perceptions, so in practice 'there is no perception without affection' (60). 

This mixture is often misunderstood by positivist psychology that wants to separate out affection and perception as a matter of differences of degree between virtual and real action.  Sometimes affection is seen as the basis of perception, its primary matter, [because it is seen as localized while perception is more extended] but Bergson sees it as 'rather the impurity with which perception is alloyed'(60).  But affections are different in kind from perception – it is a double action of the image, and provides a doubled characteristic of the image of the living body.  In the case of affection, we are offered a different kind of knowledge [maybe?].  [The context here is the continuing one about how to reconcile the internal dimensions of consciousness with the external 'extensity' of space].  It would be wrong to start with affection.  Instead we need to start with action, the ability of affecting changes in things.  As a living being, we are aware of centers of indetermination, our own and those of other living beings.  We need to deal with movements or influences  emanating from these other images and use them.  The first division of labour concerned the specializing of different sorts of organs, of nutrition and of action: just as perception develops into more specialized forms so it can deal with variation, so does affective sensation. 

It's hard to go any further without considering the effects of past experience or memory, 'the survival of past images' and their mixture with perceptions of the present.  Such survival must be useful, and help to throw a 'a better light on our decision' (71).  There is no doubt that the effect of memory can be understood subjectively, but again there is  'an impersonal basis' for memories in externality itself, just as there is with perception.  Although perceptions are interlaced with memories, and memories can only be activated by linking with perception, they can be separated abstractly [wrongly, though, for positivism, as a difference in degree].  It is a mistake, for example, to see recollection merely as a weaker form of perception [as some other thinkers do].  Activity actualizes perceptions and prolong them in the present as something that is 'ideo-motor'(74), whereas the past is only idea.  It is not that the past is only 'that which acts no longer'[but that it can no longer penetrate the real].  Nor is recollection simply a reconstruction pursued by the mind, even though memory provides us with a sense of ourselves as subjective [since it is the 'particular rhythm of duration of our consciousness' which helps us know things about matter, not our subjective qualities alone]. 

Perception alone offers 'a series of pictorial, but discontinuous views of the universe'(76) but this makes it impossible to predict qualitative change.  'The qualitative heterogeneity of our successive perceptions of the universe results from the fact that each, in itself, extends over a certain depth of duration'.  What memory does is to condense into each perception 'an enormous multiplicity of vibrations which appear to us all at once, although they are successive' (77).  Again this is a selection from the multiplicity of matter and movement and if we could somehow unpack the multiplicity, we would 'pass thereby from perception to matter, from the subject to the object'.  We would be able to grasp the whole 'system of homogeneous vibrations' of which matter consists, and we would not need even to assume space without movement, or 'consciousness without extended sensations'.  The division between subject and object would disappear in this 'extended perception'.  We would see the subjective as simply resulting from 'the contraction effected by memory', and see the whole reality of matter without these divisions of vibrations that you find in perception.  Hence [the famous quotation]: 'Questions relating to subject and object, to their distinction and their union, should be put in terms of time rather than space' (77) [in other words, our memory compresses and condenses things which limits our perceptions and makes them subjective].

The notion of pure perception leads the conclusion that 'there is in matter something more than but not something different from that which is actually given', a difference in degree between the qualities detected in perception and matter itself, a relation of part and whole.  There is no other mysterious power of matter outside what can be perceived, despite what some materialists argue.  Material elements do not interplay among themselves to produce perceptions, with a necessary relativism attending perceptions.  Nor is the opposite view ['spiritualism'] adequate, seeing qualities only as subjective matters, while matter itself remains mysterious.  In practice 'matter is precisely that which it appears to be' (80) with no hidden powers, but with a certain independence granted to 'spirit'. 

This is common sense, but it needs one philosophical tweak.  Memory contracts moments of duration into a single intuition and this produces a practical perception of 'matter in ourselves'.  At the same time, we are able [theoretically] to perceive matter within matter.  This is why memory is crucial, why it makes perception looks so subjective, and why we must try to eliminate its effects in order to philosophize.  Have the same time, it is memory that appears as spirit, the heart of subjectivity independent of matter.  Again, memory extends far beyond the usual physiological attempt to explain it in terms of the activities of the brain.  It is devoted to action not knowledge, and it follows that the conception of the spirit is linked to action similarly.  Nevertheless, work on the operation of the brain is an important area [hence the discussion in chapter two—Bergson seems to think that 'the empirical study of memory' (83) will settle different philosophical issues—if memory is a recollection, of an absent object, then it appears as an entirely mental phenomenom, but if the operation of memory is connected to activity then we make links again with the external world.  However, experimental data cannot decide one important issue, whether perception means we are actually placed outside or not, 'since the practical results are absolutely the same whether the reality of an object is intuitively perceived or whether it is rationally constructed'.  However, if memory is only the same sort of mental phenomenon as perception, that leans towards construction, but if it is different in kind, it means that perception must have an additional element, a contact with reality].  This is why the study of memory is crucial.

Chapter two.  Of the recognition of images.  Memory and the brain.

The body only conducts movements from the outside into certain actions.  This implies an independent memory gathering images, with our body as only another image, 'the last, that which we obtain at any moment by making an instantaneous sections in the general stream of becoming' (86).  The actions of the past must be stored somewhere.  There are two distinct forms, 'motor mechanisms' and 'independent recollections' (87).  These are used in action either automatically or following an 'effort of the mind' finding representations.  The first involves movements proceeding from objects directly, the second representations from the subject.  Already we have a notion of the body as 'an ever advancing boundary between the future and the past' (88), over time 'situated at the very point where my past expires in a deed'.  So representations are prolonged into the present and linked with the real through action.  There is a process involving recollection, which offer 'nascent or possible action in space'.

Experience can inform our understanding here.  For example the two forms of memory are displayed when we might study a lesson and attempt to learn it by heart: we read it repeatedly to make progress in linking elements to make a continuous whole.  Each reading can be recollected separately, though, as a definite event.  Overall, once learned by heart, we know it through habit, just like an  habitual bodily exercise, but the recollection of each stage is not habitual, and is linked with other recollections just like events are.  Each reading 'is entirely sufficient to itself' (91) and thus can be seen as 'an original moment of my history'.  It appears as a representation, embraced by 'an intuition of the mind' that can be extended or shortened, in any length of duration.  However, the overall lesson is better understood as an action not a representation, something that 'requires a definite time'.  It becomes part of the present, something lived and acted.  It's not innate because I can still remember the processes that produced it, but it is autonomous from them.

This shows that we have two different memories, 'theoretically independent' (92).  One records all the events of a daily life in the form of memory - images, and we use these in recognition of perceptions.  However, they can all be seen as nascent actions as well, and there's a second type of memory turning them into action, which requires a set of 'dispositions towards action'.  In this way, our experience consists of memory images but also a series of 'mechanisms wound up and ready', or 'answers already prepared to an ever growing number of possible solicitations'.  We become conscious of these as well, but they are always linked to action: we've selected them from the past, and the past acts in us.  Thus one memory imagines and the other repeats, although it is not always easy to tell the difference in practice.  However, recalling the first kind of memory as an image requires a temporary withdrawal from action and from instrumentalism, and this is uniquely human.  However, doing this is always being challenged by the other kind of 'more natural memory' oriented 'to action and to life'(94).

Memories which we acquire through repetition 'are rare and exceptional' compared to those that are recorded 'at every moment of duration', but we tend to prioritize them and see them as 'model memory' (95).  However spontaneous recollection is complete and cannot be amended [so it looks natural or just part of our personality], while learned recollections become more and more impersonal and foreign.  Repetition acts to alter spontaneous recollections through organization and the formation of habit [which can also be a matter of selective amnesia?]  so if anything, spontaneous recollection seems to be the better model for memory.  There is also a focus on objects routinely encountered, which develop attitudes which 'automatically follow our perception of things' (96), and this is a mechanism of adaptation, 'the general aim of life'.  What use is the other kind of memory, that focuses on images?  Our consciousness is able to set them aside from more useful forms, but sometimes the threads that bind memories to action lose their tension or relax, and the other kind of images assert themselves.  This is what happens in dreams or in other disturbances of 'sensori-motor equilibrium'(98).  Normally, however, those memories linked to action dominate and prioritize, and this is the same sort of mechanism as learning things by heart, which also regulates spontaneous images.  However, we can recall the spontaneous ones, and sometimes they are latent and able to intervene in habit-memory.

We can find some clues from examining pathological cases, including those that have shown automatism or dementia.  [I'm not going to review this set of arguments, 99 F.  Apparently they show that memory of images {'image memory'} can be detected and pursued, although they are fugitive and disappear as soon as an effort is made to pin them down.  Interestingly, mnemonics are ways of attempting to tap spontaneous memories.  Mental photography is also discussed].  Summarizing, there are two forms of storing the past, one which uses 'motor mechanisms to make use of it', the other consists of 'personal memory images which picture all past events with the outlines of their color and their place in time' (102).  We can tame the first one to exercise our will, but the second one is only useful in doing things like associating ideas, rather vaguely. 

We are describing here memory in its 'pure state'(103).  We need to do this because most psychology mixes the two [as a bad composite in deleuzian terms].  This has also led them to identify origins in particular conceptions of the brain, and postulate some simple process linking image to action.  We can see what really happens by examining recognition.

One process involves locating a present perception in remembered surroundings.  This implies a process of evoking those surroundings or circumstances, through 'some association of resemblance' (106).  What we're doing is associating a perception with a memory, but this is not always a result of a conscious suggestion of resemblance [common versions of which are again associated with notions of 'cerebral tracks', 107].  However, simple associations between perceptions and memories do not explain recognition [with more evidence from pathological cases say of word blindness, 108, or cases where people can retain visual memories, but still not recognize objects].  Instead, there is sometimes  'an instantaneous recognition, of which the body is capable by itself' requiring no explicit memory image (109), demonstrated by action not representations [the example seems to be walking around a town automatically, a bit like driving a car without being able to specify the exact route in advance: before that develops fully, there's a mixed state, but recognition is still being linked to actions, in 'a motor order', 111]. 

Recognition usually means knowing how to use something, taking an attitude towards it, releasing habitual sets of movements and perceptions.  It has already been argued that perception is always tied to action, and the links can be consolidated by repetition.  This might be a form of adaptation again.  What is involved here is a whole system of movement with a preformed order [sketched out by rival mechanists etc?]  so that parts are immediately associated with wholes [the example here is learning to play an instrument by ear so that one note immediately identifies and leads to playing a whole melody.  We're also getting close here to 'muscle memory'].  However, as humans, we have a 'past psychical life' as well (113), and that persists as events localized in time, waiting for some temporary problem with the sensori-motor system.  We require a definite effort to locate these personal memory images, a withdrawal from action, and a choice of representation from among the available ones.  Movements and actions might have already marked out a field for these, so that the sensori-motor system both edges out and encourages recollection of personal memory images, when they seem to be able to fit present circumstances.

[More ensues on pathologies like various kinds of psychic blindness, 115 F].  It leads to a more deliberate kind of recognition—'attentive recognition' (118).  This is dominated by memory images in an effort to perceive the present.  Such images have to lose a lot of detail, however,in order to be made to relate  One issue is whether perception dominates here, or whether there is some spontaneous connection between memory and perception [with more attempts to solve it by studies of the brain and pathologies].

Attention is a process that renders perception 'more intense' (120).  Details are spread out.  Attention appears to be something that comes from within, through an attitude, through some faculty of concentration, maybe related to higher levels of cerebral energy, but these views involve either metaphors or 'physiological symbolism'(121).  Instead, attention is better seen as an adaptation of the body, which are associated with workings of the mind.  It takes a positive form as it exposes detail, and it involves memory, because those memory images are used to extend perception.  More and more of them are called upon, and the extended perceptions may themselves  call up the memories.  This involves being able to reproduce a memory image, which works by synthesizing a series of 'hypotheses' [although it is often described as a process of analysis] (124).  Analogous images are chosen in a 'movement of imitation'.  This shows us that active or attentive perception involves 'reflexion', a process of projecting 'an actively created image, identical with, or similar to, the object on which it comes to mould itself'—we can see this in the phenomenon of after-images.  A first image is then connected with others stored in memory, with increasingly diminished resemblances.  This shows that active perception is thoroughly infused with memory [as in the through and through interconnectedness of subjective time].  We see this with rapid reading, where a few limited perceptions of actual words, for example, are rounded out with memory images.  Attentive perception is not so much a linear series of processes, but a circuit, in which 'all the elements, including the perceived object itself, hold each other in a state of mutual tension' (127).

There is no radical separation between the mind and the object in this case, but rather 'a solidarity'.  If we continue to concentrate, we develop more and more circuits between memory and the object, with the smallest circle represented by immediate perception containing only the object, and a series of larger and larger circles offering 'intellectual expansion' [diagram 128].  The object and these circles of memory form a whole system, and provide 'deeper strata of reality'.  The mind can operate at these different levels of depth and detail, operating with different degrees of tension [here meaning tight bonds to the initial object].  This is a way to show how personal memories become materialized, by chance or by attention.  These memories blend with perceptions so that we cannot separate them, and here memory can actually inform ['follow'] the movements of the body.

In this process, images are shaped and focused ['thinned and sharpened' 130] so as to penetrate actual experience, and to project more memories into automatic action [again rival theories are discussed, 131].  Everything turns on the body being engaged with action.  [Then more stuff on various disorders of the memory, and where recollections actually exist in the brain, 132 f].  The example that best fits this analysis is when we hear sounds, recognize them, attend to them, and finally locate them in a system of sense drawn from a memory.  We see how memory modifies present perceptions with this example, where sounds perceived pick up auditory images in memory [and again studies of brain lesions are not fully adequate to explain psychological processes].  What happens is that we can construct  'the motor diagram, as it were, of the speech we hear' (136), a motor operation which coordinates potential voices to sounds.  [The use of the term 'diagram' resonates with Deleuze again, but I think it also refers to some specific models of the brain developed by physiologists, brain circuits as it were]

The same goes with physical exercises which represent a compound of various muscular movements.  To the extent to which it seems a complex compound in our perception, we have already begun a 'virtual decomposition' or analysis of the overall exercise.  We begin by isolating autonomous movements and practising them without losing their connection with the others, by appealing to 'the intelligence of the body', drawing the attention of the body to more and more details until 'the body has been made to understand it'(138). So diagrams [maybe literal ones again in this field]  used initially to analyze into elements are only outlines, only initial attempts to distinguish what is different from other similar movements.  We still need the body to understand movement, and this requires us to follow 'the logic of the body' (139), where there are no 'tacit implications', and everything is made explicit in a complete analysis and an actual synthesis rather than the diagram which is a mere sketch.  [Followed by more stuff on the pathologies of speech. It is possible to detect in speaking subjects, 'inner movements of repeating and recognizing' which are 'a prelude to voluntary attention' (145)].

Attentive recognition is a circuit in which memories adopt higher degrees of tension [that is get focused more and more on the object].  What happens [when we interpret speech as an example] is that actors put themselves in the middle of corresponding ideas and develop them into detailed memories to flesh out the initial diagram.  'Crude' perceptions are used to identify an initial level in the circuitry to permit more and more memories to add detail.  The example of speech shows another characteristic: different voices do not produce different specific auditory images which are stored and then recognized [which is the way machines 'recognize' words, and one of those rival theories that are to be attacked].  Recall of these images require both 'a motor ally', and 'a kind of mental attitude' which itself has to be embodied [to produce some sort of active interrogation?] (152).  [the bodily connection explains why nouns are the first to be lost from memory, and verbs the last!].  Again, what we do in speech is to form a motor diagram first of all as a guide, into which can be inserted various memory images. 

This is 'a continuous movement' to condense ideas into more distinct images, and it is a mistake to try to separate this process into objective phases with finished things [discussed further 154-55].  Usually, the limits are revealed as subsequent details are added, although even here, the most complex diagrams still failed to grasp the full complexity of reality [in terms of the specific discussion of aphasia 157 F].  In more general terms, perceptions are only solidified once they have been identified with memories, and similarly, pure memories are virtual, and can only be actualized by being connected with perceptions: there are two opposite currents and it is a mistake to trace just one of them [more discussion about brains and how they work, 164f, and another model, this time of 'centres of stimulation' in the brain identified in some mechanist theories being better understood as acting as a keyboard with external objects striking individual notes that are harmonized—physiology can explain the ways in which external objects strike a note, but not how the keyboard works .  Keyboards or psychological centres  have two sides: from the front they receive impressions sent in by sense organs and real objects, from the back they are subject to the influence of virtual objects [similar to Deleuzian terminology about the body without organs turned both towards the strata and towards the virtual multiplicity.]

Pure memory is the source of intentions and it produces intermediate memories realized as memory images by the action of these various keyboards or centres.  Thus ideas become gradually embodied in images or perceptions.  Overall, then, recognition 'is not centripetal but centrifugal' (168).  [ie impressions do not come from material objects to become mental constructs but the other way around? I also think this contradicts the earlier bit about a circuit -- but then I realized this applied to recognition not perception].  Pure memories become actual and in the process awakens sensations in the body.  These are still potential [virtual in this sense] until action arises.  There's also a process 'by which the virtual image realizes itself'.  Its final stage [not a first stage]  is the stimulation of sensory centres, a prelude to a motor reaction, an action in space.  Thus real movement realizes sensations and therefore images, a process whereby 'the past tends to reconquer, by actualizing itself, the influence it had lost'.

[I am sure there is another pithy saying in here somewhere about how the truth continually regresses in positivism, I think as more and more ad hoc hypotheses have to be added to make the simplified model work].

 Chapter three.  On the survival of images.  Memory and mind.

So there are three processes—pure memory, memory-image and perception, and they always occur together: perception is 'impregnated with memory images' (170), which complete and interpret it; memory image is linked with purer memory which it materializes, and with perception with which it embodies itself.  Thought goes from pure memory to perception.  When we try to call up some period of history we detach ourselves from the present and place ourselves in a region of the past.  The metaphor here is focusing a camera.  The recollection is virtual then actualized, becoming more distinct and imitating perception, but it is still linked to the virtual.  [Associationism is then criticized as reducing mixtures to atomistic psychical states, 172 F].  Memories are not initially linked with the actual at all, but become actual: it is not that a weak sensation becomes stronger: memories and sensations are different in kind not in degree.  This theory is linked to a deeper one assuming that we have purely speculative spirit which lead us to perceptions, whereas the real difference is between present perceptions linked to action, and a powerless past.

This connects to discussion about links between present and past.  There may well be an ideal present separating past from future, but the 'real, live present...  necessarily occupies a duration' (176).  This duration spans past and present.  Action is linked to the future, and by the time we talk about the present, it is already past.  The future can be seen as a movement, and the past as sensation [because it takes time to be realized].  This makes the present, 'in its essence, sensori-motor'  (177), a combination of sensations and movements.  This is a form of consciousness associated with the body.  Because sensations and movements are linked in a system, the present seems to be particularly 'absolutely determined', compared to the past [by immediate needs etc?].  The body is a center of action representing  'the actual state of my becoming, that part of my duration which is in process of growth'(178), and we see it as a 'quasi instantaneous section' of the flowing mass of the material world. In this way, matter extended in space, is 'a present which is always beginning again', and our present is a guarantee of the materialty of our existence.  The system is 'determined, unique for each moment of duration' [determined in detail then], because only one connection between sensation of movement can be in the space at the time.

Philosophers have misunderstood this by thinking of sensation and memory as differences in degree.  The difference in kind arises because of the connection with the body [memory is not connected with the body directly, at least until it is actualized].  Positivist psychology sees memory as only a weakened perception, already an image and so already embodied.  As soon as the past becomes an image it is starting to be actualized.

Part of the problem is the unwillingness to acknowledge unconscious states.  Consciousness is attached to the active and the present, and does not occupy the central role in existence for humans.  There is an unconscious or ineffective state as well, and we can see this if we examine 'the immediate antecedents of the [conscious] decision and on those past recollections which can usefully combine with it' (182).  The mistake again is to see consciousness as aiming at speculation, which prevents us seeing the role played by immediate interests.  It is as absurd to deny the unconscious as it is to deny the existence of the material world when we cease to perceive it.  Indeed, the material world beyond our perception is an unconscious representation.  Consciousness is but a point on a line sketching all the simultaneous objects in space.  [Diagram 184]. That point itself extends in a line upwards to include recollections gained over time.  The two series are temporal and spatial [and there are unperceived elements on both].

The problem has been discussed before in chapter one on objectivity, but we can generalize to consider matter as well.  In a way, unperceived objects retain their reality for us because they are seen as something for future action, showing us the 'vitality of the present perception' (185), but it would be wrong to erect this into a 'metaphysical distinction'. We can arrange these objects spatially, in terms of distance from us, with horizons opening on to further horizons to infinity.  An actual perception seems to be but a limited content of a vaster and unlimited experience.  Memories, on the other hand, are 'so much dead weight that we carry with us' (186) that we think we can ignore: time is shut off behind us, and the present seems to be the only reality.  If a memory intrudes, there must be a special reason for it.  The parallel between non perceived objects and memories sketched above is hard to represent, because the conventional relationships have become habitual.

Certainly, action that takes us into the future can be much more predictable, while memories appear capricious: the first as a necessary order of representation, but not the second.  In practice, there may well be a necessary chain between memories as well, but this is obscured because of the action of the consciousness in selecting only the act which is useful, which can help present perceptions.  Consciousness usually [always?] has to jump distances in time, missing out the intermediate stages.  What is at stake here is an underlying conception of existence itself, which depends first of all on something being presented in consciousness, and secondly developing some logical or causal connection with elements in the past or future, seemingly determined by the characteristics of spatial and temporal series.  However, there are degrees of determination in both: only external objects can be seen as following necessary laws, but presentations in consciousness are never fully fulfilled, and aspects of it always hidden.  This leads to the normal sense of the division between external and internal states, although they are both actually engaged.

This common sense division leads to misunderstandings of matter as well as mind [the first illusion is discussed in chapter one], the second one means we tend to ignore the unconscious and its effects.  Our normal thinking is dominated by spatial images which leads to questions like where is memory stored [leading to theories of brain mechanism, 'depositing memories in matter' discussed 192 F].  More generally, this is an attempt to manage the flow of duration by thinking of it as instantaneous sections.  The real argument is that the past has not actually ceased to exist, but merely ceased to be useful.  Indeed, most attempts to grasp the present necessarily involve the immediate past, so that perceptions appearing to be instantaneous actually already contain 'an incalculable multitude of remembered elements' (194).  When engaged in practice, we only perceive the past, and the present itself is really only 'the invisible progress of the past gnawing into the future'.  [Lots of metaphors here about consciousness illuminating elements of the continuous flow of history].  Immediate interests structure our understanding of the mind.

So, there are two kinds of memory: 'intelligently constructed mechanisms'(195) which help us reply to particular demands in the present, to adapt ourselves through actions, sometimes accomplished, sometimes nascent, embedded in habits.  It does not involve calling up images.  The second one does and is 'true memory'.  It structures all our bodily states in temporal sequence, and involves 'truly moving in the past'.  It is now necessary to see how the two are connected, drawing on the notion of the present as really only the immediate past.  The body is itself an image and cannot store images in brains.  But the body 'constitutes at every moment...  a section of the universal becoming' (196) [so this notion of duration is clearly linked to Deleuze and the importance of becoming].  It represents a link between received impulses and movements, or it can be seen as a point of a cone, with its base in the past, unmoving, while the point is the present moving forward unceasingly.  The point also touches the 'moving plane... of my actual representation of the universe'[diagram 197 – this bit about the plane is what I missed first time].  Since the body is an image on this plane, it can also relate to all the other images [as well as to its own past], and it does this through the sensori-motor system and habitual [practical] memory.  Because the cone touches the plane, true memory is in contact with practical memory, supplying mechanisms with recollections to help them, while the body helps to materialize true memory, and gives it 'the warmth which gives it life'.  We can detect within the cone the passages of memory.  When everything works well, we have a well balanced mind and good adaptation to life, but some people overdo the action and live only in the present, giving way to impulses, while others live in the past as dreamers [and there is a bit on children and the way in which they balance intellect with memory, 199, together with a nasty racist bit about how 'African savages' mimic without understanding].

We can access this hidden past by escaping from the necessities of present action, as when we sleep, for example, which relaxes 'the tension of the nervous system'[meaning close connections to action], and there is some evidence that people near to death relive their lives.  [Further pathologies include those who are obsessed with particulars and details, leaving only in their memory, compared to those who follow habitual action which simply recognizes resemblances between images, without being able to think of universals].

We can analyze general ideas that work on this perception of similarity, and see how pure memory constantly attempts to assert itself into 'motor habit'(202).  It is difficult to explain the ability to generalize, which requires abstraction [and nominalism and conceptualism are criticized, 203 F—both run the risk of circularity].  The mistake is to assume that we start by perceiving individual objects which are then classified or analyzed.  Such discrimination is really a ' luxury of perception'(205), however, and a clear general idea is 'a refinement of the intellect'.  To notice the individuality of objects involves a memory of images, and this usually provides us with 'an intermediate knowledge, from a confused sense of the striking quality or of resemblance'.  This intermediate knowledge therefore has to be further refined by reflective thought to get to general ideas, or by using  'discriminative memory' to get down to the level of individuality.  Again, this is really driven by utilitarian action which leads to perceptions of resemblance or quality—as in noticing food items rather than separate objects.  Connections between these general items and the individual acts 'like a force ...  operating a purely physical law... which requires that the same general effects should follow the same profound causes'(207).  This is grasped by simple perception first, as the individual responds to the surroundings that 'take hold' of them.  It is the perception of something useful that provides the common link, even in advanced organisms where different sensations of the sense organs provided varied information.  The similarity begins in the objective world, and is turned into some habitual similarity in the mind.  These similarities are developed by understanding and memory, and maybe subdivided.  Habitual action leads to reflexivity and to general ideas.  Speech arises in the same way, as a motor apparatus assembled in such a way as to respond to individual objects and to classify them.

However, representations that arise from these stable images are unstable.  General ideas involve a constant transformation between action and memory, via levels in the cone.  There is no simple translation of general ideas into actions, but rather a 'double current which goes from one to the other, always ready either to crystallize into uttered words or to evaporate into memories' (211).  It is possible to operate at many levels inside the cone, according to our ability to detach ourselves from the sensori motor scheme.  The 'normal self' is constantly in movement between these levels, again normally determined by utility.  [Then associationist approaches are denounced again 212 F].

Chapter four.  The delimiting and fixing of images.  Perception and matter.  Soul and body

The body is deeply engaged in action and this limits the 'life of the spirit'.  We see this in pure perception.  Memory, on the other hand helps us to choose recollections that assist in action, as useful memory, and recollections are more individual, and many can be linked to actual situations.  This leads to a new dimension to human action.  Memory is also being constantly pushed on bodily action, explaining the role of the imagination. 

This has led to a serious problem about the link between soul and body, matter and spirit.  The two are not linked through negation, since both are positive and real.  The difficulties are compounded because the issue is connected to old philosophical arguments about the extended and the unextended, and about quality and quantity.  The mind offers pure unity in the face of 'an essentially divisible multiplicity' (235), and it seems to add 'heterogeneous qualities' to an otherwise homogenous and calculable extended universe.  Materialism and idealism simply try to explain one as a consequence of the other and both have been rejected, since there is more in perception than the material, and more in material reality as well.  We seem to end with a dualist position although there might be some reconciliation between the opposing terms.

Perception develops images of things to replace the things themselves, but things still 'participate in the nature of our perception' (237).  The old ideas of extensity, based on geometry, have also been revised [more below]. This already implies that there is some connection between the extended and the unextended.  The same might be said of quality and quantity once we consider pure memory.  It is argued that there are differences that remain between qualities that succeed each other in perceptions, and apparently homogenous changes going on in matter in space, according to science.  However, science has had to develop the notion of pure quantity in order to achieve its calculations, diluting their heterogeneity.  This still shows us that [mental activity] is necessarily involved, that we are really synthesizing qualities in a contracted form.  Memory does this, just as it offers a more relaxed form with qualities: differences in tension explain the apparent differences between the extended and the unextended [I'm still having trouble with this notion of contraction and relaxation—it makes sense to me to put it in the way above, but Bergson actually puts it the other way around, that heterogeneity arises from contraction and homogeneity from relaxation!  I suppose what he's getting at is that the tighter our attention to reality the more heterogeneity we see, while relaxing a bit from the tight focus on reality helps us think out and subsequently impose some homogeneity. Anyway, at the very least, it is memory that provides these characteristics of quantity and quality -- it's a bit clearer below]

We also have to discuss method.  We take as a 'fact' that which helps us adapt the real to practice.  There's also pure intuition, which grasps reality as an undivided continuity.  This continuity is subsequently subdivided and broken up into separate elements such as words and objects.  We then feel we should restore the bond between them as something external, to replace the original 'living unity' (239).  The bond takes the form of 'an empty diagram as lifeless as the parts which it holds together', and it can take an empiricist or a dogmatist form [239 f], the one emphasizing the parts and the other the relations. [To take one side for more discussion] the emphasis on the parts in particular colours the notion of experience which empiricism apparently enshrines, and, as with other philosophical positions, the argument is that empiricism enshrines far too much of the sensori-motor system.  Both usually end by recognizing the strengths and weaknesses of each other and this leads to relativism.  Instead, we must look to human experience, but at that stage above the 'decisive turn...  in the direction of our utility' (241), in the passage between 'the immediate' and the 'useful' (242).  This is still a connection with experience, not just speculation, but we hope to restore intuition's 'original purity and so recover contact with the real'.

This will involve a considerable effort to give up habitual thinking and perception and the difficulties are recurrent.  However, it is then a matter of reconstructing from this segmented passage 'the real curve, the curve itself' (242), just as mathematicians develop functions which will explain the entire curve, the differential.  The method has been already established in Time and Free Will, where the utilitarian work of the mind refracts [so there is some distortion] 'pure duration into space', so that our very psychical states can become more impersonal and distinct [which may be necessary] 'to make them enter the current of social life'.  This is what empiricism has done, resulting in seeing the world as a succession of facts.  What we need to do is put ourselves back in pure duration, where the flow is continuous, and where one state succeeds and other 'insensibly' in a proper lived continuity.  This will result in conceptions such as the evolution [sic] of action from the processes of receiving and forming sensations as we saw before.  Since this adds something new [from memory], this can be seen as the basis for a free act, where feelings and ideas produce 'a reasonable evolution'(243): spontaneity has got nothing to do with it.  Understanding this is the basis of true knowledge, putting ourselves in a duration 'wherein we see ourselves acting', not attempting to break up the elements and explain them externally.  The case of free action would be an exemplary case.

What are the implications for matter?  Can we grasp the 'vague tendency towards extension' before it is fully deployed to produce homogenous space?  We have already argued that there's a difference between pure knowledge and more practical knowledge.  The latter leads to the notion of concrete extension, together with its implied 'amorphous and inert space'.  In such a space, separate figures appear as arbitrary divisions, and movement is particularly misunderstood—as 'a multiplicity of instantaneous positions, since nothing there can ensure the coherence of past with present' (245).  We actually perceive extensity differently, while homogeneous space is only 'a kind of mental diagram', which is taken for reality itself.  Reverting to experience, or 'immediate knowledge' can be justified if it helps us solve some of the contradictions and difficulties with the notion of homogeneous space.

As suggestions towards an adequate account of matter:

(1) Movement is indivisible.  Experience tells us that this is a fact, derived both from seeing a movement and from being aware of muscles being engaged when I make it.  The movement of a hand from A to B involves a passage from rest to rest involving an undivided act.  If we actually trace the path of the movement as a line, we can then go ahead and divide it, of course, but we have had to construct an image outside of experience of an act.  Experience tells us there is a radical difference between passing and halting, movement and immobility.  We are aware that our hand might stop at any point, in imagination.  However, 'what the moving body has to do is, on the contrary, to move' (247).  It is tempting to see the moving body as immobile at each point, but this again is an 'artifice of the mind', a reconstruction.  The senses tell us what is real movement.  It is also the space which is traversed that is divisible, but movement is not the same as the path within this space [which we have already argued is an assumption]. We must not 'substitute the path for the journey' (248), and there is no substitution possible between something that progresses and something that is immobile [which would be one of these bad composites discussed by Deleuze].  The original flaw actually lies in an attempt to divide up duration into 'an indivisible instant'.  Consciousness misunderstands duration in this way: it is whole and undivided but it describes a trajectory in space which is promptly considered conventionally as running between two indivisible points.  However if the starting and finishing points are indivisible, then duration must start and stop with them—but duration is indivisible.  The line is actually a poor representation of duration and movement, and it would be wrong to attribute characteristics of lines to the characteristics of movement and duration.  This mistake underpins the paradoxes of Zeno [discussed very nicely 250 F].  There's a tendency in common sense and language to 'always regard becoming as a thing to be made use of', with no need to investigate its interior any further.  For the purposes of common sense, it is only important that every movement describes a space, and that every movement might stop.  Here lies the error of Zeno in taking these common sense understandings, in each case confusing movement and path.

(2) There are real movements.  Mathematicians operate with a relative notion of movement, turning on the distance of a point from points of reference [such as the abscissa].  Movement becomes a change of distance, and it doesn't matter which point actually does the changing.  In physics, it does, and movement becomes 'an indisputable reality' (254).  This has led to some contradictions between mathematical and physics understandings (255 F).  Absolute notion involves more than just a change of place, but do we have to make each position absolute and absolutely different?  [Different options including Newton discussed 256 F].  Perhaps a real motion has a real cause, resulting from a force, but force itself needs to be understood in science as a matter of mass and velocity, and acceleration again can only be relative.  Perhaps there are absolutely real forces, 'profound causes'(257), akin to the efforts experienced in consciousness, but this is unsuccessful too.  All these problems disappear if we see that movement is mobile in its essence, that the sensation of movements are real sensations and 'something is effectually going on' (258), whether it is me that is moving or the object.  Thinking of movements induced by human beings, they naturally appears a change of state or of quality following an act of will to produce them, some absolute, not just a relation.  This is how common sense normally decides between real and apparent movement, according to whether objects are independent of each other or not.  Without this independence, the question shifts to one of how changes in parts should be seen as a 'change of aspect' in the whole.

(3) 'All division of matter into independent bodies with absolutely determined outlines is an artificial division.'(259).  Apparently independent material objects seem to be a system of qualities detectable by sight and touch, with several other dependent qualities.  However, these senses produce a misleading impression because they are the ones that 'most obviously have extension in space'.  Apparent intervals between things, like silences between sounds or gaps between odors are more detectable by hearing and the sense of smell and taste, but sight and touch produce whole fields.  It seems as if the continuity of extensity given in these primary perceptions can be subdivided into individual bodies.  This is partly because different aspects are presented to us, but we still do not see this as a change of the whole.  We divide objects into permanent and changing ones, and then represent permanence in terms of homogenous movements in space.  This is not necessary either for intuition or science: indeed modern science notes 'the reciprocal action of all material points upon each other' (260).  Again we see the impact of life upon knowledge, the necessity of living and acting.  Action requires distinct material zones corresponding to bodies.  My own body needs to distinguish itself from others.  Seeking nutrition involves discrimination like this, efforts converging on an object as the centre.  Life itself produces discontinuities between needs and things that satisfy them.  Needs act as 'so many searchlights' singling out distinct bodies (262).  Such practical action simply persists in dividing and subdividing, but this gets  'unsuitably transported into the domain of pure knowledge'.  Chemistry is a good example, contenting itself with studying bodies rather than matter itself.  However, when physics investigates the atom, its materiality dissolves more and  more [so physics has escaped this lingering utilitarianism?]: Solids interest us in practice, but they don't really exist for physics [citing Clerk-Maxwell].  Images produced by practical life do not help us understand 'the inner nature of things' (264).  There are also constant interactions of forces, and atoms themselves can be seen as immaterial focuses of forces.  Locating things at precise points matter in practice, but not for a theory of matter which must reject these practices.  Force is material, objects 'more and more idealized'[citing Kelvin and Faraday].  In this rejection of homogenous and extensive space, they are converging with intuition in identifying something beneath, some independent reality.

(4) real movement is a transference of the state not of a thing.  Qualities and movements are closer than they seemed at first.  Both can appear in the form of sensations of movement.  Things like quality and quantity may be distinguishable in their pure states, but both can be seen as 'vibrations' at work in real movement.  This is quite unlike the notion of movement in mechanics, but in reality, even these movements are 'indivisibles which occupy duration, involves a before and an after, and link together the successive moments of time by a thread of variable quality' (268), and this gets close to how we see things in consciousness.  As an example, perceived color actually shows a narrow duration, with billions of vibrations of light contracted in it.  The slow vibrations of deep notes in music can be grasped by consciousness, and here we can actually see that quality is a matter of vibration.  Normally, however we tend to think of solids, even atoms, instead of vibrations, because moving bodies are sustained by our consciousness, and these are taken as something stable at least which can inform perception.  However, this is only really 'an outward projection of human needs' (269) and it runs into trouble as soon as we start to explore it.  Perceptions always displayed both the state of our consciousness and an independent reality [back to the definition of the image] and this contradiction helps us believe in the external world [and this is misunderstood by realism, 270, which sees sensations as just an echo of movement].  Consciousness also features acts which go beyond sensation, especially when we are attributing qualities.  [Confusing here, but it leads to the argument that] we find in movement 'the immense multiplicity' of internal vibrations.  Perception itself indicates this excess, as soon as we start to think what might happen.  Again it is not that we deduce homogenous qualities which permits calculation, because that leads to a difficulty explaining how the homogenous ones are linked to the heterogeneous ones [we cheerfully take them as essential or typical, surely?].  We are forced to conclude [I can't quite see how] that there are movements underpinning qualities, 'internal vibrations'(271), which are not homogenous, although they also produce qualities that are less heterogeneous: the reason for this is that both originate in 'an endless multiplicity of contracting' (272) and these contractions into a narrow duration make it impossible to separate the moments.

The duration of consciousness has its own rhythm, which is not that of time found in physics, where a large number of phenomena can be compressed and stored, like the vibrations of light massively compressed into a second [400 billion vibrations per second in the case of red light, it seems]. To separate them into something that we could perceive would occupy a huge interval of time [25,000 is the calculation here, at the rate of 500 per second].  In human duration, only a highly limited number of phenomena can be contained if we are to be aware of them.  The notion of infinitely divisible time involves another kind of duration.  It is different with space, where we can think of extending notions of human duration: it is something outside us, and, in principle something that can be divided.  It is 'nothing but the mental diagram of infinite divisibility'(273).  With duration, the division must be terminated when we exceed the number of manageable phenomena.  We know, however, that there must be millions of phenomena compressed into short intervals of time.  However, it would be a mistake to identify this form of duration with 'homogenous and impersonal duration' which apparently flows onward and affects everything that exists: this is only a fiction of language, and it arises because it is very difficult to imagine different sorts of duration in reality, with different source of tension. 

Instead, it is easier to think of independent Time.  This can even be supported by our consciousness such as experiencing different forms of time in dreams, or when we imagine [some alien] examining human history in a very short time, 'at a higher degree of tension' (275) [my confusion about contraction and tension isn't helped by this at all. It might make sense if the alien was contracting things in order to better act towards us and this would involve close attention to the reality?], and we do this when we condense evolutionary time into a few epochs of intense life.  In these examples, 'to perceive means to immobilize'.  In this way, representation exceeds simple perception [not exactly what Bergson says—he says perception outruns perception itself], although the content of reality is not different.  Perception contracts into moments of duration.  The rhythm of duration for humans is driven by the need to act.  Matter itself is not contracted in this way, but consists of 'numberless vibrations, all linked together in uninterrupted continuity...  and traveling in every direction' (276).

We can grasp this by trying to connect together 'the discontinuous objects of everyday experience', then think of their qualities as vibrations as above, then focus on movements, abstracting them from standard divisible space.  The result is a conception of mobility akin to the one that consciousness has of our own movements.  This will provide us with a pure conception, freed from utilitarian considerations.  We can then replace this gradually with practical life, and see what has happened to the 'inner history of things', how they have been reduced to snapshots, with everything condensed into vivid colors, replacing the 'multiplicity of inner repetitions and evolutions,...  discontinuous' (277) with a smoothed out continuous version provided by experience and connotations of separate objects in space.  We see how we can localize change and constitute stable bodies as a manageable version of 'the universal transformation'. 

We see separate objects and beings, and they are really there, but it's not so easy to separate them from their environment.  That they interact together shows that they are not tightly confined.  Our perception sketches them, gives us their 'nucleus', but only to the extent that we act on them and that they interest us.  We then have to persuade ourselves that reality is divisible like this, and to do this, we have to construct 'a network', whose mesh is variable.  This diagram is 'homogenous space'.  Perceptions of separate objects and their qualities are solidified in memory.  A perception of the flow of things between past and present is also guided by our notion of action and reaction, in the same duration and in the present.  This provides us with a sense of necessity.  We only gain freedom, some indetermination, by being able to grasp becoming in general and as it relates to us, and see how this produces specific actions that seem necessary.  We will see the affects of the greater or less intensity of life, which provides duration with greater or less tension [greater tension seemingly to do again with having to pay attention in order to adapt].  We can free ourselves from this particular localized rhythm, and disentangle the elements that have been solidified, but only by grasping succession in general in terms of homogenous time.

So homogenous space and homogenous time are not properties of things, nor essential categories for human knowledge as in Kant.  They arise from a useful solidification and division of the real so that we can act on it, 'the diagrammatic design of our eventual action upon matter' (280).  To misunderstand this leads to metaphysical dogmatism [more criticisms 280f], and assumes a speculative rather than a practical or vital interest driving human action.  Instead, we need to consider real duration and real extensity, without homogenizing it.  However, this would involve taking on deeply rooted habits, in common sense and in idealism or realism.  It would be to argue that real extensity persists between apparently different orders, and there's a connection with quality.  [Rival philosophies are explored 282 F --this time, they  tend to isolate particular sections of a process as privileged bases for the rest].  Overall, concrete extensity is projected into space, and real motion 'deposits space beneath itself' (289), although it is easy to invert these.  Because human action operates on the world as apparently fixed motionless images, imagination tends to see rest as the point of reference for motion, motion as only a variation of distance, through a preprovided space which is infinitely divisible.  This has enabled us to map reality, but we should not mistake the diagram for reality itself [in other words, a good summary 290f].  We overcome these limits by placing ourselves 'face to face with immediate reality' (291), and this enables us to break down these absolute divisions 'between perception and the thing perceived, between quality and movement'.  Everything depends on a requirement for action projecting notions of divisibility of space and matter.  It is the 'prejudices of action' that separate perception and matter.

We can now address the old issue of the link between body and soul.  Again this is a philosophical mistake which sees them as separated in the first place, with matter as divisible and the soul as inextensive, with matter further subdivided according to some homogenous space and so on.  By contrast, there are degrees and transitions between the two, the passage from the idea to the image and then to the sensation, for example.  This means that spirit has found a way to unite with matter in perception, especially if we consider it as memory, 'a synthesis of past and present with a view to the future' (294), which contracts and condenses in order to become useful for action.  Again we can assert that distinctions between body and mind are matters of time not space.  Ordinary dualism still operates with the conventional notions of space and extension, and can only link mind and matter with some parallelism or some assumed harmony.  However if we start from pure perception, subject and object already coincide, and if we place them in duration we see that matter tends to be more and more 'a succession of infinitely rapid moments', while spirit as memory, prolongs the past into the present in the form of 'progress, a true evolution' (295). 

A temporal distinction rather than a spatial one is what is needed, and we can also understand the transition between spirits and matter if we see that spirit serves to 'bind together the successive moments of the duration of things' in order to come into contact with matter.  It also offers a number of levels between itself and matter, allowing for relatively less determined action, and for the operation of reason and reflection.  At each level [moving in a downward direction?], the intensity of life grows, and duration becomes more tense , and the sensori-motor system develops.  But this takes place only by an independence of mind, which is misunderstood as a development in the complexity of the nervous system.  In particular, the past is contacted more and more, after moving out of the immediacy of the practical.  There are 'all possible intensities of memory' between matter and mind, and thus many degrees of freedom [from necessity].  But mind and matter are joined in a process, not separated radically from each other.

Pure perception shows this happening, sense devoid of memory, it is part of matter.  Memory intervenes in a way which is compatible with matter as a series of repetitions of the past, a series of moments which are equivalent.  For matter too, 'its past is truly given in its present' (297) [because we can deduce the present from the past].  Living things evolve, however creating 'something new every moment' and we can only recover their past by examining memory.  This means that the past cannot be 'acted by matter' but only imagined 'by mind'(298).

Summary and conclusion

The body is an instrument of action and does not create representations.  This is shown in a similarity of function between the brain and the spinal cord.  The brain does not store our recollections or images and therefore does not contribute directly to representation.  This seems to involve us in a radical split between body and soul, but this will eventually be resolved.

The mistakes of materialism or idealism arise from seeing the physical and mental as duplicates of each other, but the duplication is difficult to explain.  How does it happen that the consciousness reflects material reality?  How is it that the conscious can easily be disconnected from a material reality that persists nevertheless and that is studded by science.  Materialism and idealism serve as two poles, and if these are seen as equal, relativism arises.  The direct correspondence between the material and the ideal obviously also sacrifices freedom.  The common era is to see the operation of the mind as an operation of 'pure knowledge' (302), designed to duplicate external reality or to pursue its own interests completely separated from material reality.  This problem disappears if we see that memory and perception are really connected to bodily action.  Memory increasingly also allows us to escape from 'the rhythm of necessity' (303).  When contracted, memory increases the characteristics of matter.

Turning to perception, it involves seeing the body as a certain image, therefore located in a totality of other images, implying that perception is and must be 'only some parts of those objects themselves' (304).  Perception selects from the totality possible actions of the body, it chooses but does not create.  We have already established some connection between realism and idealism, concrete are used by calling things 'images', with relations both to other objects and to consciousness.  Perception is limited, however, and science has managed to establish more of the characteristics of matter or material images.  Perception offers a relation between part and whole in order to prepare for action, and this explains the link between consciousness and matter that idealism fails to grasp.  Again, seeing consciousness as speculative is at the heart of the problem [I think the argument is that it requires a special kind of speculation in the form of science to really grasp the characteristics of objects and their relations with other objects.  One consequence is to find, with Kant, that there must be some unknowable reality].  Bergson's approach firmly links consciousness and matter, through a relation of part and whole [this is also inexplicable with strict materialism]. 

An additional flaw is the conception of space as homogeneous, logically anterior to material things.  Instead, there is extensity, which is 'prior to space' (308), since action conceives of it as a network underneath material continuity, enabling intervention.  This overcomes problems raised by conceptions of the divisibility of space, and notions of the extended world which exceed those of conscious perception.  The argument here is that extensity is not really divided, and immediate perception is extended.  Reality consists of an 'extended continuum'(309) structured around our body as a source of activity.  This involves dividing the continuum into distinct bodies, but virtual action of things upon our body and vice versa also takes place, in the form of nascent reactions or possible actions, and here 'the brain exactly corresponds to the perception', continuing this virtual action.

Pure perception has to be modified by considering affections and memory. Both will help us identify the extensity of the body and the duration of perception.  Affections represent the intrusion of our own body into perception, especially the effects of bodies which are close to us.  Our own bodies coincide with the centre of this perception [so we can experience real effects of other bodies].  Pain is an example, resulting from an attempt to repair local damage, something real but within the body.  The difference between interior sensation and exterior images lies at the heart of our distinction between subjectivity and objectivity, but again we must remember that these distinctions are connected to action, not speculation.  It would be a mistake to see perceptions and sensations as only offering differences of degree [as rival psychologists do], with sensation as something on extended acting as the building block for images.  Instead we need to start with whole experiences as 'the necessary field of our activity' (312).  Perceptions as images come first, and sensations should be understood as an impurity, intruding into pure perception a part of our own body.

We have to also consider spirit.  Being able to choose perceptions from among images in general hints at it.  There might also be a consciousness in the universe itself, [a kind of action that guarantees stability].  However, the best way to understand it is to start with the ways in which individual consciousness escapes from necessity, from repetition.  This involves understanding memory, where 'we definitely abandon matter for spirit' (313).  In pure perception, the image of a body is actually  given [that is no virtual dimension], so that consciousness duplicates the image.  However, we also have representations of absent objects, involving even more the power of our consciousness to construct images.  Repetition of an image can be understood as weakening perception, and making it a similar in kind to recollection.  However recollection can be seen as something different in kind, adding to perception and converting it into action.  This is because it focuses on absent objects [which can no longer contribute anything].  Psychological studies of pathology can help to show us this [315 F].  There can be 'passive recognition' leading to automatic movement as a habitual response to a perception, located in the motor apparatus.  But there is also active recognition, produced by memory images: sometimes this will not activate motor machinery, however, and will not become actual, as in some pathologies.  Consciousness itself has to be engaged in order to materialize memories, and this is pure memory.

Idealists tend to assume that perception is only an intense version of memory [318], while some psychologists have argued the reverse.  However, memory is not a progression from present to past, but the reverse.  When we remember something we place ourselves in the past, which is a virtual state, and gradually actualize the image through a series of different planes of consciousness until it is finally materialized in a perception, an active state in the present.  Again this shows the importance of the practical orientation of psychical states.  This process ads vitality to what would otherwise be automatic action.  Pure memory like this can also not be seen as a a result of brain states, but should be understood instead as a genuine 'spiritual manifestation' (320).