Why is there an unconscious?

This may seem like a strange question.  From the perspective of our personal experience, it's not clear why it shouldn't all just be conscious.  With a little training in introspection, I can succeed in making significantly more of that unconscious experience accessible to my consciousness, and this can be really useful.  So what's the point of having this hidden stuff to begin with?  You might answer that it conserves mental processing power, or something of the sort, and since this processing is evolutionarily expensive you want to use as little a possible.  However, it's not immediately clear why being conscious of running some mental algorithm should require more resources than just running the algorithm.  And anyway, this line of thought leads directly to a different version of the same problem.

You could equally well pose the question the opposite way around -- why is there consciousness at all?  This direction would probably make more sense from an evolutionary perspective.  What exactly does consciousness add to our ability to survive?  You might answer that it makes possible learning, or abstraction, or reason, or language.  But it's not clear that any of those things require consciousness.  Or perhaps it's not clear that consciousness is anything different from having those things?  Even if conscious computation doesn't turn out to require more resources, what's the point of having this curlicue added to the brain.  We sure make a big deal about it, but it's not obvious that it has any purpose.

If you put these two versions of the question together, you realize that an even better version is probably -- why is there a difference between the conscious and the unconscious?  A world with only one or the other wouldn't raise any issues.  But there's something mysterious about having both.  And isn't perhaps the oddest thing about this setup that they so often seem to be in opposition to one another?  Why are we fighting our selves?

I'm not sure I understand completely where this idea takes us, but I'd like to think our new understanding of repetition as process has something to contribute here.  We concluded that the process of repetition constructs identity.  In other words, repetition as a relation differs in kind from the identities of the things related and necessarily hides itself behind them.  Is this distinction exactly why there must be both a conscious and an unconscious?   The virtual needs the real and the real needs the virtual?  

Deleuze never seems to use the word "process" in D&R, despite the fact that one of his big influences, Whitehead, explicitly called his thinking a "process philosophy".  I'm not completely sure why Deleuze avoids the term.  Part of the explanation may be that it's all to easy to claim (as I have so many times already) that, "X is a process, not a thing" as if this just immediately elucidated something profound.  Have we really said anything important when we do that though?  Aren't we still using a concept of identity -- a process -- that just treats it like a more abstract thing?  Don't we push the concept around in our head like a little box, just like we would with anything else?  Perhaps Deleuze avoids using the term directly because the book itself is written as a definition of what a process is.  We have to discover what's important in the distinction between a process and a thing.  The movement, the displacement, the hiding, the lack of identity, the intersubjectivity.  A process necessarily happens in time.  But what is time?

We'll see in the next section that thinking about the virtual seems to culminate in in contemplating the difference in nature between problems or questions and their solutions.  The unconscious will be seen as a problem posing machine, a questioning machine.  These problems have a real 'objective' existence for Deleuze.  They aren't just a reflection of some fact that we don't happen to know right now.  They are, to use the jargon, ontological, and not epistemological problems.  While they may be objectively real in the colloquial sense, their reality is not of the same type as the solutions.  Solutions are real, but problems are virtual.  And questions are virtual objects.  Their being is ?-Being, (non)-Being, a non-Being that is not negation.

Being is also non-being, but non-being is not the being of the negative; rather, it is the being of the problematic, the being of problem and question.

Just in the place where I might expect a discussion of the unusual "being of a process", what makes a process distinct from a thing, we get instead a section on problems and their relationship to the unconscious.  It makes me think that the short but deep answer to our original question is: "because Time".  

Repetition is Virtual

In our last episode I added yet another word (intersubjectvity) to describe what's quickly becoming the most over-loaded concept of all time -- the virtual object.  

I tried to illustrate how me might consider our self as a virtual object.  This object would be something like the process that establishes the intersubjective relation between all our various selves at various time points.  It aims to account for how we can experience an ongoing self across time without taking any particular instantaneous self, or even the general collection of all our selves, as the model for our true self.  There is nothing more than ongoing flow, the process of transformation of one possible temporal self into another.  The virtual object 'self' is the structure of that process, an invariant, in the mathematical sense, of those transformations.

Like I say, this just added yet another way to describe the virtual object.  It's a reflection of a reflection.  It's a sexual drive.  It's a partial object, a part of a real object and a part of our self.  It's a missing object, a part of its own self, without identity and in constant motion.  It's a fragment of the pure past, a memory that never actually happened.  It's a symbol.  It's intersubjective.  It's not a thing but a process.  This proliferation of terms is meant to help us 'identify' something inherently unidentifiable. 

Ultimately, what we're trying to 'identify' is really the concept of identity itself.  And in a world of difference, identity can only mean repetition.  Not repetition of something, because this would presume a repeated identity.  But repetition for itself, repetition repeating itself.  Identity is produced when repetition for itself succeeds in relating differences to one another.  Different possible outcomes are related to one and the same process.  The repetition of this process of relation becomes prior to the things repeated.

But this is a very difficult idea.  It might help to back up a couple of pages and talk a little about what Deleuze is not saying.  In fact, he starts this section with a critique of the failure of every psychoanalytic concept of repetition.  

There are a variety of places where repetition appears in psychoanalysis.  The most concrete is the suggestion that our neurotic fixations stem from hidden aspects of our childhood sexual experience that, so long as the analyst is unable to help us see them, we are doomed to repeat.  In other words, it claims that our counterproductive present behaviors repeat our childhood traumas because we have repressed those memories.  Slightly more abstractly, psychoanalysis makes use of repetition as a therapeutic device.  We are meant to project our problematic childhood relationships onto the analyst himself in the form of a transference neurosis.  With his help, we then are able to not merely see, but to dramatize those lost events.  We play them out again in therapy in the hopes of achieving some catharsis.  Repetition also appears in psychoanalytic theory more abstractly.  Freud thought of the death drive as the desire of a living thing to repeat or return its 'natural' or original state -- inorganic material.  And Jung thought that our individual unconscious was constantly repeating or reincarnating archetypes contained in some larger collective unconscious.  It's not immediately obvious why the unconscious should be so dominated by repetition, by psychoanalysis keeps coming back to it again and again.

I lay all these out because Deleuze goes after each one and shows us how all of them are based on an idea of repetition that necessarily involves a first time, an original model, repeated a second time, as copy.  

The question is whether repetition may be understood as operating from one present to another in the real series, from a present to a former present. In this case, the former present would play the role of a complex point, like an ultimate or original term which would remain in place and exercise a power of attraction: it would be the one which provides the thing that is to be repeated, the one which conditions the whole process of repetition, and in this sense would remain independent of it.  The concepts of fixation and regression, along with trauma and the primal scene, express this first element. As a consequence, repetition would in principle conform to the model of a material, bare and brute repetition, understood as the repetition of the same: the idea of an 'automatism' in this context expresses the modality of a fixated drive, or rather, of repetition conditioned by fixation or regression.

This bare or material repetition could also be called numerical repetition.  It involves counting the number of distinct instances of the same thing.  

Of course, if recognizing a repetition were as simple as counting sheep, we wouldn't need Herr Doktor Freud monkeying about in our dreams.  So the psychoanalytic theory of repetition has to have a second element to it to explain why repetition is not immediately obvious -- each time the same thing returns, it has a new disguise.  This disguise is created by our repression, some force which tries to prevent the original term from repeating itself.  This force that is opposed to the power of repetition never really succeeds completely.  It only manages to change the form in which the repeated original manifests itself, either by introducing some code for it (my what a big cigar you have Herr Doktor!) or, at the limit, by sublimating the original sexual drive and turning it into some 'higher' principle (how aesthetically magnificent is the world's tallest tower!).  But the most important point for Deleuze is that this force is separate from and opposed to the force of the repetition.  It comes from outside the repetition, which always strives to produce exactly the same thing, not some modified or disguised copy.

And if this material model is in fact perturbed and covered over with all kinds of disguises, with a thousand and one forms of disguise or displacement, then these are only secondary even if they are necessary: the distortion in the majority of cases does not belong to the fixation, or even to the repetition, but is added or superimposed on to these; it necessarily clothes them, but from without, and may be explained by the repression which translates the conflict (within the repetition) between the repeater and what is repeated. The three very different concepts of fixation, automatic repetition and repression testify to this distribution between a supposed last or first term in relation to repetition, a repetition which is supposed to be bare underneath the disguises which cover it, and the disguises which are necessarily added by the force of a conflict.

I think this critique is pretty clear when it comes to Freud's psychoanalytic method.  But Deleuze also critiques psychoanalytic theory by extending the same idea to cover a more abstract range of possible "ultimate or original terms".  For example, the thing repeated by the death drive is the state of being inorganic matter.  But this too is just a state, a fixed type of identity, albeit a more abstract one defined by the property 'not-alive'.  

Even - and above all - the Freudian conception of the death instinct, understood as a return to inanimate matter, remains inseparable from the positing of an ultimate term, the model of a material and bare repetition and the conflictual dualism between life and death.

Likewise, making the original or ultimate term imaginary or spiritual instead of material doesn't help either.  So long as we continue to think of it as a state defined by some pre-established identity we will still just be counting the number of copies of some model. 

It matters little whether or not the former present acts in its objective reality, or rather, in the form in which it was lived or imagined. For imagination intervenes here only in order to gather up the resonances and ensure the disguises between the two presents in the series of the real as lived reality. Imagination gathers the traces of the former present and models the new present upon the old.

Even if my childhood was actually blissful and I merely imagined it as traumatic, I would still be copying the model of that same imagination by reliving it as a neurotic adult.  In fact, the problem persists even if we assume the original imagination wasn't my own, but an image selected from some universal or cultural unconscious, since in both cases, this mysterious ur-memory would simply be repeated twice in my own imaginative experience.  

Nor do we believe that the Freudian discovery of a phylogenesis or the Jungian discovery of archetypes can correct the weaknesses of such a conception. Even if the rights of the imaginary as a whole are opposed to the facts of reality, it remains a question of a 'psychic' reality considered to be ultimate or original; even if we oppose spirit and matter, it remains a question of a bare, uncovered spirit resting upon its own identity and supported by its derived analogies; even if we oppose a collective or cosmic unconscious to the individual unconscious, the former can act only through its power to inspire representations in a solipsistic subject, whether this be the subject of a culture or a world.

In summary, every aspect of the psychoanalytic theory of repetition is based on the model of recognition and representation.  It counts identities.  This theory will never help us build a self, to follow the construction of a subject capable of having experiences.  Instead, it is constantly presuming the identity of that subject experiencing a real or imagined object at the outset, as an original model which keeps coming back in the form of a disguised copy that partially resembles it.  It takes the conclusion as a premise.

The traditional theory of the compulsion to repeat in psychoanalysis remains essentially realist, materialist and subjective or individualist. It is realist because everything 'happens' between presents. It is materialist because the model of a brute, automatic repetition is presupposed. It is individualist, subjective, solipsistic or monadic because both the former present - in other words, the repeated or disguised element - and the new present - in other words, the present terms of the disguised repetition - are considered to be only the conscious or unconscious, latent or manifest, repressed or repressing representations of the subject. The whole theory of repetition is thereby subordinated to the requirements of simple representation, from the standpoint of its realism, materialism and subjectivism. Repetition is subjected to a principle of identity in the former present and a rule of resemblance in the present one.

Okay, but then so if we can't understand the repetition that Deleuze wants us to think about by counting things that resemble an original identity, how can we understand it?  Basically, by paradoxically taking non-identity as the original term that is repeated.  Instead of seeing one thing repeating another we focus on the way the relationship between them is repeated.  We think of this relationship coming before the identity of the things that it relates and giving rise to this identity.  The virtual object is exactly an attempt to think of this between as a 'thing' in its own right.  It's a sort of figure-ground reversal of our usual way of looking at things.  Instead of things that then get related, we take the relation as a 'thing' in and of itself, and the distinct things to be various modifications of that relation.

However, while it may seem that the two presents are successive, at a variable distance apart in the series of reals, in fact they form, rather, two real series which coexist in relation to a virtual object of another kind, one which constantly circulates and is displaced in them.

Repetition is constituted not from one present to another, but between the two coexistent series that these presents form in function of the virtual object (object = x). It is because this object constantly circulates, always displaced in relation to itself, that it determines transformations of terms and modifications of imaginary relations within the two real series in which it appears, and therefore between the two presents.

This kind of repetition in some sense creates the identities of the things we can later count, and it creates them precisely as related, as instances of 'the same thing'.  It is exactly the idea we need to get past the problem of taking some identity for granted as we try to explain how we recognize and remember who we are through time.  Otherwise the whole action of remembering or identifying is left as fundamentally mysterious and undefined.

Consider the two presents, the two scenes or the two events (infantile and adult) in their reality, separated by time: how can the former present act at a distance upon the present one? How can it provide a model for it, when all its effectiveness is retrospectively received from the later present?

It's not that my current self happens to remember my childhood self.  Instead, my current self is produced as remembering that childhood self.  The memory, the middle term or relation, draws together the terms as a repetition.  It gives a commonality to those terms which didn't pre-exist this relation.  The only thing repeating here is the process of drawing together, which has as its product two related moments. 

Maybe we can make this more concrete by considering how biological memory actually works.  We tend to think of our memories by analogy to a computer (or formerly a book) -- we say they are stored somewhere, and that we have go locate and retrieve them.  In short, we treat them like objects.  But if a past memory is an object, how does it ever do anything in the present?  What is its mechanism of action-at-a-temporal-distance?  We say that something in the present object reminds us of this past object, and then we go call it up.  But then the present can't really be copied from that memory.  Why would the copy set off in pursuit of a model it doesn't have any reason to believe exists, and that can only be useful once it's dragged back to the present?  The model doesn't seem to have any causal effect here.  It's just a placeholder for some internal mechanism we don't see but by which present stimulus produces present perception or action.  On the other hand, if the present is the model and the memory is the copy, then what's the point of producing a memory at all, since it seems like this wouldn't invoke a real past, but just an imaginary one that anyhow doesn't seem to add anything.  The passage I quoted above continues:

Furthermore, if we invoke the indispensable imaginary operations required to fill the temporal space, how could these operations fail ultimately to absorb the entire reality of the two presents, leaving the repetition to subsist only as the illusion of a solipsistic subject? 

Either way you look at it, it's difficult to know how the memory, considered as a snapshot of a no longer real past experience, can have any effect in the present.

Of course, we know that memories are not like objects stored like photos in file cabinets somewhere in the brain.  We know that on some level a memory has to be something marked, in the present, as a re-creation of something marked as not-present.  As far as I know, no one really understands the neural mechanism for this, at least for conscious human memories.  But we do know something about the way memories are 'stored' in simpler neural systems.  They're stored in the structure of the connections between neurons.  "Neurons that fire together wire together".  What makes something a memory is that it reactivates a structure that had been modified by earlier activation patterns.  This neural mechanism of memory gives us a clear image of how two different triggers get related as memories by repeating the same process of activity.  In this case, there is really no original instance of which later memories would be copies.  Memories are continuously encoded in the ongoing modifications of the structure of the brain.  The water flows down this gully because of the pattern of erosion created by an earlier flow.  

Hopefully that digression helps us to better understand how the essence of memory, of identity across time, is a process of transformation that leaves behind a residue in a structure of the world.  Deleuze of course puts this more poetically by saying that disguise is an essential, and not an accidental, part of repetition.  There is no first instance which appears modified by disguise a second time.  What's repeated is the modification, the variation, the constantly transforming process of activation leading to reactivation.

The displacement of the virtual object is not, therefore, one disguise among others, but the principle from which, in reality, repetition follows in the form of disguised repetition. Repetition is constituted only with and through the disguises which affect the terms and relations of the real series, but it is so because it depends upon the virtual object as an immanent instance which operates above all by displacement.

I think there are a variety of reasons that Deleuze talks about this as form of "disguise".  It ties it to the symbolic disguise that psychoanalysis identifies in our dreams.  It relates it to Nietzsche's idea of masks.  But translated into plainer English, disguise and displacement are simply meant to describe the way we never as see a process itself directly, but always have to infer it from its products.  A process is an odd object because it is always hidden behind those products.  It can't be known in any other way.  In which case every time we attempt to grasp it we come at it through some relationship between multiple products.  This is the easiest way to understand the displacement here.   Nor can it ever be exhausted by any finite collection of products.  There is always a part of it that remains in reserve, as a potential for further transformation and new products.   It never coincides with itself, so to speak, as a plainly self-identical object, but is always moving.  It is motion.

Behind the masks, therefore, are further masks, and even the most hidden is still a hiding place, and so on to infinity. The only illusion is that of unmasking something or someone.

In the first place, the mask means the disguise which has an imaginary effect on the terms and relations of the two real series which properly coexist. More profoundly, however, it signifies the displacement which essentially affects the virtual symbolic object, both in its series and in the real series in which it endlessly circulates. (Thus, the displacement which makes the eyes of the bearer correspond with the mouth of the mask, or shows the face of the bearer only as a headless body, allowing that a head may none the less, in turn, appear upon that body.) 

The virtual is always a very slippery concept that can easily be misunderstood in a number of ways.  It is as inherently hard to grasp as the notion of time itself.

I am just a symbol of myself

In this next section (pg. 103-106) Deleuze comes back more explicitly to the theme of the chapter and connects virtual objects to the idea of repetition for itself.  We've already seen that these objects are partial, inherently mobile, always missing from themselves, and symbolic.  But what do any of those characteristics have to do with the concept of repetition?  

As always, to sort this out, we have to step back to the problem.  Chapter 1 tries to conceive an atomized world of pure flux, pure instantaneous difference.  Chapter 2 then begins to ask how, in such a setting, we could ever come up with a concept of identity.  Repetition in time, or across time, is the closest we can come to the concept of a stable identity given our starting point.  So the overall problem of repetition is really a problem of how identity can be constructed.  Within that general question, the whole psychoanalytic section we are working on now addresses the more specific problem of how a human identity that senses, remembers, and even thinks, can be constructed from infant milquetoast.  Finally, the specific part of that section we're working on today has been addressing memory.  We might state the problem as: how can I develop a sense of my self as an ongoing substantial entity that's the same from moment to moment?  How can I remember who I am?  If I don't take this identity for granted at the outset, if I think of each instant as a completely different and new moment, it would seem I have to develop some notion of how my current self now repeats my former self.  But then the question immediately becomes, "wait, repeats what exactly?"  Where would the first self have come from?  This is going to be the jumping off point for understanding Deleuze's difficult idea of the way a virtual object is a kind of symbolic repetition that is defined precisely by never having a first time. 

I think the easiest way into this idea is to frame it as a question of intersubjectivity.  In a world of difference, 'I' am not one subject but many -- an infinity of former and present selves.  The question is what holds these subjects together.  We usually ask this as: what stable underlying object is being repeated each time one of these subjects appear?  We imagine the production of a copy made according to a model.  There is an initial model, from which we can derive an infinite number of copies, each with some acceptable amount of variation (because if the variation is too big, it's not 'the same' as the model, it becomes something different).  In the case of a human self though, where would this initial model have come from?  For psychoanalysis, the answer is obvious -- it came from our childhood interaction with our parents.  This moment sets up the model that we repeat in disguised form for our whole life.  

For Deleuze, of course, this whole schema of the model and the copy is problematic, because it presumes that we can identify the original model.  He is therefore critical of the stress most psychoanalysis places on childhood memory.  In place of thinking about a model -- a thing or particular empirical memory -- being repeated, he wants us to put the process of repetition first.  This sounds a bit counter-intuitive.  How can we recognize a repetition without having a first thing to repeat in a second instance?  But this problem vanishes (or at least changes form) if what's repeated is a process which produces all N instances of the thing as a result.  Instead of considering each copy of ourselves as a chip off the old block, we have to see that the copies directly relate to one another through a process of causal transformation.  They need not be mediated by some central original from which they each differ.  As a result, what we usually call a self is actually not a thing, but a relationship between things, the process of the transformation of one into another.  In other words, an inter-subjectivity.  The virtual is intersubjective, or as we said earlier, it's a relationship between possible selves (or perhaps sets of possible selves).  

Clearly, the 'thing' repeated here, the process, is a different kind of thing than the repeated products.  That's why Deleuze describes it as virtual, and why it has so many paradoxical properties.  

Repetition is no more secondary in relation to a supposed ultimate or originary fixed term than disguise is secondary in relation to repetition. For if the two presents, the former and the present one, form two series which coexist in the function of the virtual object which is displaced in them and in relation to itself, neither of these two series can any longer be designated as the original or the derived. They put a variety of terms and subjects into play in a complex intersubjectivity in which each subject owes its role and function in the series to the timeless position that it occupies in relation to the virtual object.  As for this object itself, it can no longer be treated as an ultimate or original term: this would be to assign it a fixed place and an identity repugnant to its whole nature. If it can be 'identified' with the phallus, this is only to the extent that the latter, in Lacan's terms, is always missing from its place, from its own identity and from its representation. In short, there is no ultimate term - our loves do not refer back to the mother; it is simply that the mother occupies a certain place in relation to the virtual object in the series which constitutes our present, a place which is necessarily filled by another character in the series which constitutes the present of another subjectivity, always taking into account the displacements of that object = x. In somewhat the same manner, by loving his mother the hero of In Search of Lost Time repeats Swann's love for Odette. The parental characters are not the ultimate terms of individual subjecthood but the middle terms of an intersubjectivity, forms of communication and disguise from one series to another for different subjects, to the extent that these forms are determined by the displacement of the virtual object.

While this idea of an intersubjective virtual is basically what I wanted to convey in this post, I now find myself wanting to add all sorts of addenda to head off various possible misunderstandings.  Because it's very easy to take the words "the process not the product is what is repeated" at face value, to substitute one object for another and miss the most crucial point.  All of the paradoxical apparatus of virtual objects is meant to help us conceive in detail exactly what is the difference between a process and a thing. 

A process is not just a more abstract version of a thing. It's also not merely a part or aspect of a thing.  When we speak of the intersubjective nature of our virtual self, we don't just mean what attributes all of our different present selves have in common. The virtual self is not just the totality or overlap or least common denominator of all the real selves. This would not take us outside the concept of identity.  We would still be taking the individual identity of each of the selves for granted as a sort of substance with various attributes or properties.  Their coincidence would then be some essential core of our self, and each individual self would be related to this core analogously.  All the different incarnations would reference the same underlying 'me'.

A true process, on the other hand can't be conceived as a normal sort of object.  It inherently involves time, movement, the qualitative change of transformation.  It is never static but always circulating, perpetually missing from its place.  We never see it directly but only hidden in its products.  It's like the wind in this respect.  The great symbol of the spirit for a good reason.  

The Virtual and the Possible

In writing the previous post I felt the need to spend a few moments explaining how those characteristics of the virtual make it different in kind from the possible.  I have often used the two as synonyms in an effort to avoid relying on technical terms, but for Deleuze there is a crucial distinction.  In fact, this distinction in kind between the possible -- conceived as just a negated copy of the actual -- and the virtual -- conceived as positive potential different from the actual -- is the core motor of Deleuze's philosophy.  You might almost say that this distinction is difference in-itself.  We've touched on this idea before when we discussed Being and Non-Being (?-Being) in the Sophist, especially at the end when we pushed back against Quine's ontological 'urban renewal' (aka slum clearance).  It's so important that its literally the last words Deleuze ever wrote.

There is a big difference between the virtuals that define the immanence of the transcendental field and the possible forms that actualize them and transform them into something transcendent.

So I'm sure we'll come back to this distinction many times.  I pause to address it again now because the image of the two mirrors facing one another I used as inspiration for our discussion of virtual objects cast this difference in a slightly new light for me.

Possibility space is intuitively familiar to us -- its all of the things that might happen.  But how can we define an open-ended entity like that?  Well, in physics, they define it as phase space.  It's just a space constructed from all the possible values that the various variables that describe the system (the degrees of freedom, like position and momentum) can take on.  Any of the points in this space would describe an actual system at a given time, and the system is always currently at one or another of these points.  

A virtual space is subtly different because it does not take from granted that we already know which are the right variables to use in describing the system.  In other words, it doesn't take for granted that we know what the right question to ask will be.  For purpose P, we might want to describe the system using variables (x,y,z) but for purpose Q we better use (a,b,c).  Once it's established as a discipline, physics can skip over this question and just assume that things like position and momentum describe everything.  But this choice of variables was anything but obvious as the discipline got established.  Trying to define a virtual space takes us back towards figuring out what question we were trying to ask to begin with, and for what purpose.  Which makes it more like constructing a space of possible possibility spaces.  We want to investigate the different ways of describing a system and how these ways might be related.  

This seems to be a fairly straightforward generalization of constructing a phase space until you realize that this time you are unable to take for granted and bracket your purpose in choosing one set of variables instead of another.  All possible descriptions for all possible purposes doesn't seem like a well formed concept.  So its doesn't seem like you can lay out this whole space once and for all the beginning as we did with classical phase space.  All you can really do is bootstrap your way towards this space of total possibility, or possibility raised to another power (it's not clear quite how to describe this entity -- in some sense it's like describing the world as chaos).  You begin with a particular purpose in mind and characterize the variables that describe the possible behavior of the object relevant to this purpose.  Consider that if the behavior of the object were pushed into some interesting corner of that phase space, it might start to do something surprising, might begin to function in a new way that you had not expected.  It might now lend itself to a different purpose, or at least make you wonder about other aspects of its behavior that you had neglected in your first description.  

[To add a concrete example here, consider the way a tropical chimp would describe water.  It rains.  It forms ponds and rivers, etc ... It behaves in a certain well defined way, does certain well characterized things relevant to tropical chimp life.  Then the chimp goes on a ski vacation.  Or climate change brings the chalet to the chimp.  Now, water can also be ice, which has a whole different set of relevant properties and behaviors.  In fact, tropical chimp probably thinks of it as a completely different thing for a while.  Of course, physics reassembles this all as a single phase diagram of water.  But that only happens with the benefit of hindsight]

In short, you might discover a new experience of the object that leads you from examining one space of possible behaviors into examining a completely different one.  The virtual is meant to describe the way you move through this space of possible spaces that cannot be completely defined in advance.  The possible, by contrast, is always defined in advance and limited by an initial question that is supposed to define everything relevant about the object.  The possible contains no sense of progressive specification or qualitative change.  Everything is laid out at the start, and all the points in the phase space are equivalent aside from the empirical fact that the system happens to be in just one of them right now.  The possible also removes the observer from the problem of describing the system.  It assumes in advance the availability of a complete description, valid for every observer looking at this system in any way.  

I'm trying to very abstractly describe how the virtual entails some sort of shuttling back and forth between the object in question and the subject interacting with it.  And even more than just going back and forth between two fixed points, it enables the construction or progressive unfolding of the possibilities of those two sides.  The virtual is always a transcendental synthesis that goes beyond the subject and object and from which they emerge.  This idea of emergence from a transcendental field is at the heart of everything Deleuze wrote.  The requirements this idea imposes allows you to understand all the other concepts -- the need for multiplicity, for there always being more, for affirmation, etc ...  Right now, I'm finding that our discussion of the partiality and 'missing-ness' of virtual objects, they way they are in-process, in motion between subject and object, trapped between two mirrors, illuminates a new aspect of what we mean by a transcendental field.

We're Virtually There!

It's time to try and sum up virtual objects as best we can.  I'm still uncertain I really understand the psychoanalytic context of this concept, but at least we've explored it a little.  Now we'll try to use those ideas from Freud, Klein, and Lacan to better situate virtual objects in their proper place in Deleuze's scheme.  We should bear in mind two things in what follows.  First, the problem is: how do you make a self?  Second, we've already met the paradoxes of the virtual once before; the pure past is virtual totality, a past that never was present, a mythical, immemorial past.  The pure past is not a collection of memories of former presents, but like the space of all possible presents.  We can use some of that understanding to further elucidate the virtual objects.

Let's start with the problem.  We've got an infant.  They are just a bundle of passive local egos -- drives -- that are the product of pre-existing habits of evolution.  Gradually, that goo becomes a person.  On some level, we might consider this development from a bundle of joy to an integrated unity as another of evolution's habits, a duplicate version of the first passive synthesis that takes as its building blocks the initial habits.  For reasons I'm not entirely clear on right now, we're going to look at this development as a whole new type of passive synthesis -- the second passive synthesis of memory.  This makes the second passive synthesis into something more like learning.  The infant gradually learns to recognize objects in the real world, and learns to recognize its self and corresponding body as objects.  We are no longer in the domain of self-fulfilling habits, but are now asking how habits get established.  It's as if we've discovered the elements that become contracted in a habit need to have some sort of prior capacity to relate in order for them to get linked together at all, but that this only became apparent when we wondered how those habits themselves get linked into a larger structure.  

Habit is the foundation of time, the moving soil occupied by the passing present. The claim of the present is precisely that it passes. However, it is what causes the present to pass, that to which the present and habit belong, which must be considered the ground of time. It is memory that grounds time. We have seen how memory, as a derived active synthesis, depended upon habit: in effect, everything depends upon a foundation. But this does not tell us what constitutes memory. At the moment when it grounds itself upon habit, memory must be grounded by another passive synthesis distinct from that of habit.   

The second synthesis splits into two distinct but interdependent directions.  Both posit some object (you might even say some substance or "substantive") that holds together various parts of a whole.  On one side, reality is synthesized as something the infant needs to actively go get to satisfy its most important drives and keep its nascent self in business.  Hence the infant learns to distinguish real objects which hold together enough to hold itself together.  On the other side, the various drives can be put into relation to one another by locating them in the virtual space of memory.  Virtual objects are like attractors in memory space capable of drawing together the drives.  This 'drawing together' is the second passive synthesis of memory.

In fact the child is constructed within a double series: on the basis of the passive synthesis of connection and on the basis of the bound excitations. Both series are objectal: one series comprises real objects which serve as correlates of active synthesis; the other virtual objects which serve as correlates of an extension of passive synthesis. The extended passive ego fulfils itself with a narcissistic image in contemplating the virtual centres.

So far I've outlined what are basically the general job descriptions of the real and the virtual.  They are the stuff or material that is necessary for the infant to develop a world and a self.  Agglomerations or attractors in 'reality space' define real objects external to the infant or parts of its world.  These become the goals of an active ego.  Likewise "centers" in 'memory space' define virtual objects which are internal to the infant or parts of its self.  These become objects that an extended passive ego contemplates (without acting) to form itself.  These two sets of objects, and their corresponding types of ego, interact, and the full Ego is built from this interaction.

... the infantile world is in no way circular or egocentric but elliptical; that it has two centres and that these differ in kind, both nevertheless being objective or objectal.  In virtue of their dissimilarity, perhaps a crossing, a twist, a helix or a figure 8 is even formed between the two centres. What, then, would be the ego, where would it be, given its topological distinction from the Id, if not at the crossing of the 8, at the point of connection between these two intersecting asymmetrical circles, the circle of real objects and that of the virtual objects or centres?

Deleuze spends most of the time talking about the strange virtual objects needed for the second passive synthesis, since we mostly (think we) know what a real object is and how it governs our activity.  Ultimately, what makes virtual objects so strange and paradoxical is that the stuff they are made of -- the space of memory, the pure past -- is not real.   

We cannot say that it was. It no longer exists, it does not exist, but it insists, it consists, it is.

The virtual is, it has an effect, but it does not act, and so in that sense is not real.  Because the virtual objects are not real they have many other strange and wonderful properties.  Some of the psychoanalytic concepts we've talked about can help illustrate these.  

To begin with, a memory is obviously not like an external object, but something closer to the internal objects we saw Melanie Klein describe.  We have to be slightly careful with the way we construe this analogy because we don't want to presume the existence of the Ego we're trying to construct; the dividing line between inside and outside has to emerge.  Remember that the type of pure past we're talking about doesn't belong to anyone.  It's not a particular memory of a particular moment some individual once actually experienced as a present.  It's more like the space of possible experiences, possible presents that could incarnate that memory in different ways.  Which is also in a sense the space of possible selves who can have that experience (should we call these experimenters?).  

[Now suddenly I understand the sentence just before the quote about habit and memory I used at the beginning of this post.  The second synthesis has to provide a measure for the relationship between the possessor and its possessions.  It has to simultaneously produce a subject and an object, which is why Deleuze calls the second passive synthesis transcendental.

The first synthesis, that of habit, is truly the foundation of time; but we must distinguish the foundation from the ground. The foundation concerns the soil: it shows how something is established upon this soil, how it occupies and possesses it; whereas the ground comes rather from the sky, it goes from the summit to the foundations, and measures the possessor and the soil against one another according to a title of ownership.

]

Though they exist "objectally" (they are not simply created by the subject) memories are like internal objects in the sense that they relate to what it feels like, on the inside, to interact with some real external object.  But because they are possibility spaces -- immemorial memories that never were any subject's experience, and possible selves that might have that experience -- these internal objects differ in kind from the external ones.  Like two mirrors facing one another, the possible experiences reflect the possible experimenters that reflect the possible experiences ... and so on.  Like every good paradox, they directly capture a sense of movement.  In addition, as potentials, the virtual objects aren't subject to a binary either/or logic, but obey a logic of indefinitely iterated addition.  Deleuze cites Klein's idea of the "good and bad object", to which a principle of non-contradiction does not apply.

It lacks its own identity. The good and the bad mother -- or, in terms of the paternal duality, the serious and the playful father -- are not two partial objects but the same object in so far as it has lost its identity in the double.

The same recursive structure we saw where each part of a fractal past reflects the whole and implies an "ever-increasing coexistence of levels of the past" also means that the virtual object is never complete.  This is like the flip side of the possibility of addition.  There's always some more of a virtual object that's missing, as befits a pure potential.  Just like the virtual object is added to reality or "planted in it like trees from another world", as another possible experience of the real object, it is also subtracted from reality.  Neither procedure produces a whole though.  They only give us a partial object -- part of the real object, part of our self, and part of itself because of its perpetually unfinished circulation between these two parts.

We see both that the virtuals are deducted from the series of reals and that they are incorporated in the series of reals. This derivation implies, first, an isolation or suspension which freezes the real in order to extract a pose, an aspect or a part. This isolation, however, is qualitative: it does not consist simply in subtracting a part of the real object, since the subtracted part acquires a new nature in functioning as a virtual object. The virtual object is a partial object - not simply because it lacks a part which remains in the real, but in itself and for itself because it is cleaved or doubled into two virtual parts, one of which is always missing from the other. In short, the virtual is never subject to the global character. which affects real objects. It is -- not only by its origin but by its own nature -- a fragment, a shred or a remainder. It lacks its own identity.

Conversely, these virtual objects are incorporated in the real objects. In this sense they can correspond to parts of the subject's body, to another person, or even to very special objects such as toys or fetishes. This incorporation is in no way an identification, or even an introjection, since it exceeds the limits of the subject. Far from opposing itself to the process of isolation, it complements it. Whatever the reality in which the virtual object is incorporated, it does not become integrated: it remains planted or stuck there, and does not find in the real object the half which completes it, but rather testifies to the other virtual half which the real continues to lack.

Finally, our discussion of Lacan and his object a gives us the most succinct image of a structure that inherently lacks completeness in itself yet sets everything in perpetual motion.  This was supposed to be a "substantial lack", a perpetually missing piece, an object that we only find as lost.  All of these are descriptions of the partial, unfinished, and hence simultaneously creative nature of the virtual.  For Lacan, what causes desire, the object-cause of desire, was also a memory, but precisely a hazy sort of memory from a time before we were a fully formed subject.  So again, this memory comes to us as something that doesn't really belong to us, in the proper sense of the term.  It's a fragment of our formative infant experience from before there was an us to experience it.  But this possible experience sets in motion the search for a possible self to whom it would correspond.  This would be exactly whatever self it is that would enjoy this lost ur-experience.  The possibility on both sides of this equation is always just a part of the object or the self, and not those finished products in themselves.  In some ways it's more than just the object or self, since it includes variations of those (the lack could be construed as a surplus).  But in another sense it's always less than they are, since they never actually correspond to it directly but to aspects or parts of it.  Either way you look at it, I think the key concept is the sense of perpetually unfolding recursion of two mirrors facing another.  In fact, I think this is almost always the key concept in Deleuze.

Although it is deducted from the present real object, the virtual object differs from it in kind: not only does it lack something in relation to the real object from which it is subtracted, it lacks something in itself, since it is always half of itself, the other half being different as well as absent. This absence, as we shall see, is the opposite of a negative. Eternal half of itself, it is where it is only on condition that it is not where it should be. It is where we find it only on condition that we search for it where it is not. It is at once not possessed by those who have it and had by those who do not possess it. It is always a 'was.' 

I don't know whether I've really done justice to virtual objects and their relationship to psychoanalysis, but at this point we have to move on.  It's not as if we won't see all this same structures again and perhaps improve our description.  To set up the next section, though, we have to deal with one last point, which is the link between the virtual and the symbolic.

We've already hinted at this in discussing Klein's partial objects.  These seemed to relate to what various objects in the infant world meant to the infant.  I almost want to say what they 'represent' to the infant, but this is a very specific word in Difference & Repetition, and I think we are still investigating how a representative view could be formed.  The partial objects are (partially) an "isolation or suspension that freezes the real in order to extract a pose, an aspect, or a part".   

Deleuze continues this line of thought when he quotes Lacan talking about Edgar Allen Poe's The Purloined Letter.  It isn't actually a very interesting story in my opinion, but it does work well as an image for something that is hidden by being in plain sight.  The letter in question is eagerly sought after by a detective who searches every possible secret hiding place of the suspect's residence without success.  The armchair detective hero of the story deduces that this valuable letter is hidden not because its location is concealed, but because it is laid in the open and disguised as a worthless scrap of discarded paper.  This illustrates Lacan's ideas that there is a lack, a piece inevitably missing from the center of the story.   

In this sense, Lacan's pages assimilating the virtual object to Edgar Allan Poe's purloined letter seem to us exemplary. Lacan shows that real objects are subjected to the law of being or not being somewhere, by virtue of the reality principle; whereas virtual objects, by contrast, have the property of being and not being where they are, wherever they go: 

 

what is hidden is never but what is missing from its place, as the call slip puts it when speaking of a volume lost in the library. And even if the book be on an adjacent shelf or in the next slot, it would be hidden there, however visibly it may appear. For it can literally be said that something is missing from its place only of what can change it: the symbolic. For the real, whatever upheaval we subject it to, is always in its place; it carries it glued to its heel, ignorant of what might exile it from it.

We'll get a lot more explanation of the symbolic in the next section.  But we can already sketch some of its important characteristics.  A symbol is like a memory in that it never was anything concrete.  There's a double arbitrariness to the symbol, so to speak, in that it can use any old signifier to refer to a whole class of signifieds.  For example, I can use a whole bunch of distinct sounds or combinations of little marks to symbolize 'table', and each of these different instantiations of the symbol can refer to an infinity of different physical objects that would fall into the category 'table'.  In other words, the symbol "table" is not a one-to-one correspondence like we usually think of it, but actually a many-to-many correspondence.  It links possible tables to possible experiences of pointing to tables.  The missing aspect of the symbolic, its mobility or ability to change place, is like the flip side of this observation.  The 'same' idea 'table' always appears in some different disguise.  It's impossible to pin down or hold in place.  Being able to move around like this -- ultimately not being anything more substantial than this motion itself -- is literally the essence of something being symbolic.  

Lacan's Lack

Everything I know about Lacan I learned from this long (but quite readable) essay explaining his concept of "object a".  This term, which Deleuze mentions a couple of times, is meant to be read "object lowercase a".  Understanding it at least roughly can help us uncover more of what he has in mind by his virtual objects.  Which means I need to start by summarizing the important points I took from the essay; from there we can go on to see how it fits with Deleuze.

First, object a is a paradoxical object because it's the form we only retrospectively project on our experience.

The objet a is a paradoxical "object" directly because of the relation between its emergence and loss. Žižek clarifies this for us: "This coincidence of emergence and loss, of course, designates the fundamental paradox of the Lacanian objet petit a which emerges as being-lost" (The Plague of Fantasies, p. 15). The idea is objet a is not an actual object we once possessed but, then, lost. The very moment it emerges it does so as a lost object. This is its trick. We never really had a perfect drive satisfaction (jouissance), but we retroactively produce this illusion as soon as restrictions are placed on the jouissance we had at our mother's body (das Ding).

Jouissance could use a whole 'nother essay, but it seems to be roughly pleasure, or maybe more accurately the pleasure principle we saw associated with drives and which implies the whole cycle of alternating need and fulfillment.  Probably the easiest way to understand this paradoxical quality of object a is to think about our everyday reminiscences of pleasurable moments.  Memory is a funny thing, much more creative than we give it credit for.  We don't remember something as we actually experienced it, as if we were watching it on film.  What we live is a sort of recreation of some aspect of that experience.  In our airbrushed memory, the experience is way more fulfilling than it was at the time.  I get older, but it stays the same age, so to speak, and that immortality or immemoriality is what we love about it.  Object a is an image of this lost paradise, a reminder of something that never really existed on its own.

Second, Lacan's idea seems to be that the subject, the whole ego, develops around this longing to recapture something ineffable that never was.  This leads him to describe the subject as defined by its structurally lacking something.  The always missing part of ourselves is object a; this image of a pleasure we lack comes to define us.

Remember, the being of an infant is that of unmediated jouissance or libidinal plenitude. For a baby, to lose this sort of jouissance is to lose its very being. The moment language takes hold and places restrictions on jouissance is the moment when a structural lack is produced within the human being — the lack of immediate jouissance. Now one's being is a sort of non-being. "I am my inability to be." Now there's some-thing that I'm missing, that I lack, that I must have in order to be whole again. There's some "part" of myself that I have been separated from. This some-thing is objet petit a. We could say that objet petit a is the ghost of one's primordial jouissance that emerges through the socialization process. The objet petit a is that little remainder of the excessive jouissance we were once submerged in. As Lacan put it, "The objet a is something from which the subject, in order to constitute itself, has separated itself off as organ. . . . It must, therefore, be an object that is, firstly, separable and, secondly, that has some relation to the lack"

You can probably already see where this is going.  He fixes the cable.   The "organ" in question is the phallus, and we are separated from it by symbolic castration. So the phallus becomes the ur-symbol of how we need some symbolic memory of what never happened in order to experience what happens. The phallus is the head virtual object.  

For jouissance to become virtual is for it to cease to be immediately present. In other words, it is something the subject lacks. In fact, the subject is this very lack. The desiring subject, all the days of its life, will be unknowingly chasing this lost "object" in the form of the virtual jouissance we call objet petit a.

The objet a is both the void, the gap, and whatever object momentarily comes to fill that gap in our symbolic reality. What is important to keep in mind here is that the objet a is not the object itself but the function of masking the lack. 

From the sounds of this essay, Lacan and especially Žižek have written themselves in circles about the "substantial void" at the heart of the subject (its "constitutive lack" or "reified emptiness") and the tragic consequences of the fact that we only want what we can't get.  These psychoanalysts always seem frightfully obsessed with having their Johnson cut off.  Bunch of fucking amateurs.

I don't think that exploring lack and emptiness in any depth is going to help much with Deleuze.  The idea of a memory that functions as a creative force in its own right is clearly relevant and interesting.  But I don't think Deleuze intends to take us into the Hegelian and ultimately Christian territory Lacan's and Žižek's theory seems to lead towards.  Negation negating itself and our experience of ourselves as primordially fallen from joy is not really Deleuze's style.  In fact, though it's been 20 years since I read it, I now suspect that the sustained critique of "lack" that I recall occupying a big chunk of Anti-Oedipus Volume 1 is directed as much at Lacan as at Freud.  I think Deleuze prefers to conceive of the lack or partiality of object a less as a "substantial void" than as a line of flight -- a gap that sets things in motion and opens up the edge of an expanding space, ultimately a positive potential.  In other words, the lack isn't an identity, but a difference -- not a thing but a process.  We always get something different than what we want because we are always in a sense inventing exactly what it is we want.

Which brings us to the third and final takeaway from this essay on Lacan.  

The Lacanian object or objet petit a is not the object of desire. Instead, it is the object-cause of desire, that is, it is the object that causes you to desire the object you actually desire. Imagine being in a theater watching a graceful ballerina perform a spotlit solo. You find yourself completely captivated and memorized by this dancer. However, what in this analogy is the condition of this enchantment? It is the very spotlight in which the ballerina stands out from the darkness. In a sense, we are not even conscious of this light — it is "unconscious". Analogously, it is this "object" that causes the ballerina to attract our attention.

The object-cause of desire is like the proverbial donkey's carrot that functions as an unattainable lure or enticement.

[The essay also gives several other interesting images to illustrate the object-cause of desire.  Check them out if the prima donna doesn't do it for you]

I think this idea of object a as the object-cause of desire helps us understand better why Deleuze calls it both a virtual object and a partial object.  It's virtual because it differs in kind from a real object.  It's not the object itself we want, but the aura of satisfaction that we associate it with and that sets in motion a process which constructs both the "we" and the "what we want".  Like we saw with Klein, object a is not an external object, but an internal one, though now this internal object is understood not as something we have but as something we lack -- namely the direct pleasures the drives have become cut off from as the ego begins to form.  

There are two main senses in which objet a is the cause of desire (it's worth noting that Lacan devoted a whole session to this concept of the cause in Seminar X). First, objet a is literally the cause of all desire, that is, it's emergence is the very reason why human beings start to desire at all. Before the "falling away" of objet a, we are not desiring subjects, but, instead, are little bundles of wild drives and unregulated jouissance. The "breaking off" of objet petit a is precisely what causes desire as such. This is easy enough to understand, since we already know that the Law (name-of-the-father, prohibition) separates us from das Ding (maternal body of jouissance) and, thereby, produces a fundamental lack (objet a) "in" us that causes us to desire.  This constitutive, structural lack is one that all of us as desiring subjects have in common.

Virtual objects are also inherently partial because they are only a part of the real object and only a part of us.  We've already talked about how they are like idealized memories of some experience.  They abstract some trait or aspect of our particular experience.

However, the objet a also comes to cause specific desires. We all have our own particular histories of desire and objet a in its idiosyncratic dimension is the hidden cause at work behind the scenes.

Each of us in our own ways (via fantasy) come to unconsciously associate certain empirical features with that missing "part" of ourselves. If we can just find the right object of desire, then we will finally fill the void. If we can just get ahold of IT (no, not that killer clown), then we will be complete. Of course, this is impossible, but it's the impossibility that makes desiring subjectivity continue to be possible.

Put differently, objet petit a gets linked to certain idealized and libidinally-invested traits. One of the easiest ways to see this mechanism at work is to consider that many men end up marrying women that strikingly resemble their mothers.

Though we've also touched on it, but it's much more difficult to understand how virtual objects are partial "on the inside", so to speak.  They are our own autonomous drives reflected back to us as part of a self.  They no longer appear as self-sufficient forms seeking nothing more than to repeat themselves.  They have gained a "higher" purpose they didn't have to begin with.  Lacan seems to look at this partiality from the perspective of the global ego.  There's a primordial lack of fulfillment for some local egos implied by the ordering and regulation of all those "wild drives".  Deleuze might be trying to looks at the situation from the perspective of the component drives.  Some may indeed be cut off from their goal, from their self, while others are gratified, but all of them shift from being ends in themselves to being a parts of some global ego.  Perhaps this is why he seems to focus on the lack, the negative, or the void, and instead talks more about how the virtual object is missing from its place, no matter where it's at.  But we'll come back more specifically to what use Deleuze makes of Lacan next time.

Klein's Bottle

I know about as much about Melanie Klein and her partial objects as you do.  I think Deleuze introduces them here just to give some more tangible way to think about the paradox inherent in virtual objects.  Remember our working hypothesis is that the virtual object is a mirror image of something that doesn't exist before the mirroring; it's an image created from two mirrors facing one another.  To see how Klein's ideas are going to help though, we have to do a bit of background work. 

Fortunately, we already know Freud backwards and forwards, amirite?  Klein's position appears to be a modification of his idea that sexual drives have an external object, and that ego-drives and simply ones that have taken the self as their erotic object.  For Klein, the drives are aimed at objects that are only a part of what we would normally term a real object (apparently this idea also had some precedent in Freud).  Then later, when the ego begins to form and take itself as a narcissistic object, this does not represent the ego relating to itself as an external object (which in Klein's theory has become partial and hence not really a good model for the ego's wholeness and unity) but as the contemplation of a collection of internal objects.  At least, that seems to be how this paper describes her ideas. 

Klein further developed the notion of internal objects, and this was central in the expanded role of objects in her own and Fairbairn's work. In her early papers she had described more and more complex phantasies in young children concerning their mothers' "insides." The latter were believed to contain all varieties of substances, organs, babies, etc. During the late 1920's, Klein began to write of parallel phantasies which the child develops concerning his own insides, a place similar to his mother's interior, also populated by body parts, substances, people, etc. In contrast to Freud's super-ego concept, Klein suggests that these phantasies of internal presences begin in the first months of life. As development proceeds, Klein suggests, representations of all experiences and relations with significant others also become internalized, in an effort to preserve and protect them. This complex set of internalized object relations is established, and phantasies and anxieties concerning the state of one's internal object world become the underlying basis, Klein was later to claim, for one's behavior, moods, and sense of self. Klein conceives of the drives as more tightly bound to objects, both internal and external, than did Freud, and hence she rejected the notion of "primary narcissism."  The infant, Klein argued, has a much deeper and more immediate relation to others than previous psychoanalytic theory has credited him with.  This rejection of the concept of "primary narcissism" was no mere theoretical refinement. Narcissism had been applied, within classical psychoanalysis, as an explanatory concept with regard to many clinical phenomena, ranging from tics to schizophrenia, and as a tool for understanding rigid resistances within the psychoanalytic situation itself.  Klein and her collaborators took issue with these explanations. They argued that seemingly narcissistic manifestations like tics, schizophrenia and extreme resistances in analysis are not objectless states (i.e., with only the ego as object), but reflect intense relations to internal objects. For Klein, the content and nature of relations with objects, both real people in the outside world and phantasized images of others imagined as internal presences, are the crucial determinant of most important psychical processes, both normal and pathological.  She argued that Freud's "narcissistic libido" reflects not a cathexis of the ego itself, but of internal objects, and thus replaced Freud's distinction between narcissistic libido and object libido with the distinction between relations to internal vs. relations to external objects.

Here you can start to see why Deleuze calls these internal objects virtual.  They are the phantasies (with a "ph" to distinguish it from fantasy), images, projections that are taken back within.  And, since they relate to the self-fulfilling habits of passive local egos (aka drives), they are like hallucinations that spontaneously bring themselves into being.  In fact, for those local egos, it would make sense to talk about a "primary narcissism".

The fact that these egos should be immediately narcissistic is readily explained if we consider narcissism to be not a contemplation of oneself but the fulfilment of a self-image through the contemplation of something else: the eye or the seeing ego is filled with an image of itself in contemplating the excitation that it binds. It produces itself or 'draws itself' from what it contemplates (and from what it contracts and invests by contemplation). This is why the satisfaction which flows from binding is necessarily a 'hallucinatory' satisfaction of the ego itself, even though hallucination here in no way contradicts the effectivity of the binding.

It seems like one way to describe these internal objects is as symbols that indicate how parts of the world makes parts of us feel.  Of course, the relationship does not have to be one-to-one; our parts can feel differently about a given part-object, which goes towards explaining our global ego's unconscious ambivalence to many things.  One real object can contain many virtual objects, and one virtual object can appear as part of many different real objects.    

Klein posits the view that the real others in the infant's external world are constantly internalized, established as internal objects, and projected out onto external figures once again. Klein does not seem to consider such internalization to be a defense mechanism per se, but rather a mode of relating to the outside world. "The ego is constantly absorbing into itself  the whole external world".  Internal objects are established corresponding to real external others, as "doubles." Not just people, but all experiences and situations are internalized. The child's internal world "… consists of innumerable objects taken into the ego, corresponding partly to the multitude of varying aspects, good and bad, in which the parents appeared to the child's unconscious mind … they also represent all the real people who are continually being internalized".

Deleuze's description of the relationship between virtual and real objects quickly gets more complicated than what Klein appears to be saying here.  From the little I've read it does not seem that she focused on the mirroring aspect implicit in her definition of internal objects -- the way they internalize an aspect of something only to project it back onto that thing in an endlessly circulating self-fulfilling prophecy.  To better understand that we will need to discuss Lacan next time.  

But we can end here by examining what Deleuze says about the way virtual objects are incorporated into real ones, since it specifically refers to Klein.

Conversely, these virtual objects are incorporated in the real objects. In this sense they can correspond to parts of the subject's body, to another person, or even to very special objects such as toys or fetishes. This incorporation is in no way an identification, or even an introjection, since it exceeds the limits of the subject. Far from opposing itself to the process of isolation, it complements it. Whatever the reality in which the virtual object is incorporated, it does not become integrated: it remains planted or stuck there, and does not find in the real object the half which completes it, but rather testifies to the other virtual half which the real continues to lack. When Melanie Klein shows how many virtual objects the maternal body contains, it must not be thought that it totalises or englobes them, or possesses them, but rather that they are planted in it like trees from another world, like Gogol's nose or Deucalion's stones.

[First off, Gogol's nose is a reference to Nikolai Gogol's short story The Nose, which you can read in full here.  It's an amusing little satire of a low level bureaucrat who wakes discover that his nose is missing.  Horrified, he goes in search of it, only to discover that it has taken on a life of its own and is evading him.  The point is that the nose can show up anywhere.  It's an autonomous part of the world, a piece of Major Kovalyov (the noseless hero) that he finds wandering around outside himself.  

And Deucalion is some forgotten Greek hero (being the son of a God just ain't what it used to be) who repopulated the earth after the flood:

Once the deluge was over and the couple had given thanks to Zeus, Deucalion (said in several of the sources to have been aged 82 at the time) consulted an oracle of Themis about how to repopulate the earth. He was told to "cover your head and throw the bones of your mother behind your shoulder". Deucalion and Pyrrha understood that "mother" is Gaia, the mother of all living things, and the "bones" to be rocks. They threw the rocks behind their shoulders and the stones formed people. Pyrrha's became women; Deucalion's became men.

Again, the theme is of mobile objects sprouting out of a fixed one.]

So the overarching point here seems to be that the virtual objects are parts, but, strangely, they are not really parts of a whole.  Added together they don't sum up to the full real object.  Likewise, they are parts of an ego that doesn't exist yet; a collection of drives is not a person.  In both cases, they are inherently incomplete, they lack the whole to which they would be the corresponding parts.  I believe this is what Deleuze is trying to get at when he says that the incorporation of the virtual object into the real complements, instead of opposing the isolation, or the carving out, of the virtual object from from the real one.  

We see both that the virtuals are deducted from the series of reals and that they are incorporated in the series of reals. This derivation implies, first, an isolation or suspension which freezes the real in order to extract a pose, an aspect or a part. This isolation, however, is qualitative: it does not consist simply in subtracting a part of the real object, since the subtracted part acquires a new nature in functioning as a virtual object.

The isolated part can't be identified with the real object, and even when it is reincorporated into it, it doesn't complete the circle and reform the whole.  Conversely, the whole real object cannot be internalized (or introjected) because it would require an already formed ego to serve as its representative, and all we have so far is a collection of local passive egos.  Hence an introjection would, "exceed the limits of the subject".  The virtual object is split between two sides -- an internal and an external one -- and each of these sides is itself partial or fragmentary.

The virtual object is a partial object - not simply because it lacks a part which remains in the real, but in itself and for itself because it is cleaved or doubled into two virtual parts, one of which is always missing from the other. In short, the virtual is never subject to the global character. which affects real objects. It is - not only by its origin but by its own nature - a fragment, a shred or a remainder. It lacks its own identity.  The good and the bad mother - or, in terms of the paternal duality, the serious and the playful father - are not two partial objects but the same object in so far as it has lost its identity in the double. 

The Virtual is Sexy Because It's Us

After giving us a general introduction to the virtual object in a childhood development context, Deleuze goes on to relate the concept to psychoanalytic theory much more specifically.  He starts by describing how the real/virtual distinction is parallel to the distinction Freud makes between the self-preservative and sexual instincts.  Then he goes on to relate the virtual object to Melanie Klein's partial objects, and Lacan's object a.  At this point I've read a bit about each of these ideas, and while they are really still above my pay grade, even a rudimentary understanding of them can help illuminate what Deleuze means by the virtual object.  So let's take these references in order.

I don't know when Freud first started distinguishing between instincts for self-preservation and sexual instincts.  It probably doesn't matter though since it's clear that his theory evolved very substantially.  In fact, he gives a synopsis of this evolution in Beyond the Pleasure Principle that I found very interesting:

Here then is an opportunity for looking back over the slow development of our libido theory. In the first instance the analysis of the transference neuroses forced upon our notice the opposition between the 'sexual instincts', which are directed towards an object, and certain other instincts, with which we were very insufficiently acquainted and which we described provisionally as the 'ego-instincts'.  A foremost place among these was necessarily given to the instincts serving the self-preservation of the individual. It was impossible to say what other distinctions were to be drawn among them. No knowledge would have been more valuable as a foundation for true psychological science than an approximate grasp of the common characteristics and possible distinctive features of the instincts. But in no region of psychology were we groping more in the dark. Everyone assumed the existence of as many instincts or 'basic instincts' as he chose, and juggled with them like the ancient Greek natural philosophers with their four elements earth, air, fire and water. Psycho-analysis, which could not escape making some assumption about the instincts, kept at first to the popular division of instincts typified in the phrase 'hunger and love'. At least there was nothing arbitrary in this; and by quite a distance. The concept of 'sexuality', and at the same time of the sexual instinct, had, it is true, to be extended so as to cover many things which could not be classed under the reproductive function; and this caused no little hubbub in an austere, respectable or merely hypocritical world.

The next step was taken when psycho-analysis felt its way closer towards the psychological ego, which it had first come toknow only as a repressive, censoring agency, capable of erecting protective structures and reactive formations. Critical and far-seeing minds had, it is true, long since objected to the concept of libido being restricted to the energy of the sexual instincts directed towards an object. But they failed to explain how they had arrived at their better knowledge or to derive from it anything of which analysis could make use. Advancing more cautiously, psycho-analysis observed the regularity with which libido is withdrawn from the object and directed on to the ego (the process of introversion); and, by studying the libidinal development of children in its earliest phases, came to the conclusion that the ego is the true and original reservoir of libido, and that it is only from that reservoir that libido is extended on to objects. The ego now found its position among sexual objects and was at once given the foremost place among them. Libido which was in this way lodged in the ego was described as 'narcissistic'. This narcissistic libido was of course also a manifestation of the force of the sexual instinct in the analytical sense of those words, and it had necessarily to be identified with the 'self-preservative instincts' whose existence had been recognized from the first. Thus the original opposition between the ego-instincts and the sexual instincts proved to be inadequate. A portion of the ego-instincts was seen to be libidinal; sexual instincts -- probably alongside others -- operated in the ego. Nevertheless we are justified in saying that the old formula which lays it down that psychoneuroses are based on a conflict between ego-instincts and sexual instincts contains nothing that we need reject today.  It is merely that the distinction between the two kinds of instinct, which was originally regarded as in some sort of way qualitative must now be characterized differently namely as being topographical. And in particular it is still true that the transference neuroses, the essential subject of psycho-analytic study, are the result of a conflict between the ego and the libidinal cathexis of objects.

I highlighted some of the stuff there that's particularly relevant to our discussion.  First off, for Freud, what made an instinct "sexual" was that it had an object.  Like, say, a breast.  Later on, Freud decided that the first object of our sexual desire was actually our self, experienced as an object.  Calling this "Narcissism" was a stroke of genius, since we are precisely falling in love with the reflection of something we find in the world that we don't know to be ourself, just as in the myth.  This radically changed how he thought of the ego-instincts related to self-preservation. These are now not defined as lacking an object (other than the rather vague one of keeping oneself around) but have a very precise object, namely the projection of our self onto some part of the world and its reflection back to us.  This is why he concludes that the distinction between the two types of instinct is no longer qualitative (since they both have an object they are both of the same type) but topographical.  Presumably Freud meant to convey a higher and a lower mind, a tip of an iceberg and a part underwater, with the ego being the part of the bigger erotic (libidinal, sexual) self visible in the world.   Deleuze actually changes this term to "topological", which makes a lot more sense to me.  The distinction really seems to be more about an inside and an outside, about an inside that's been discovered outside, making the ego into a sort of surface level of the mind, or a screen on which we see ourselves projected.  

Deleuze takes most these same ideas and organizes them in a new way since his question is slightly different.  Despite all the talk of childhood development Freud actually seems to almost take the ego, or at least the normal form of unity and identification that the ego should take on if it's not neurotic, for granted.  He doesn't really seem to discuss how this narcissistic ego could form itself, and skips directly to it recognizing itself or identifying itself in an object.  The form of the ego and the form of the object both presume a prior concept of identity that allows for their identification.  Since Deleuze is trying to not take for granted any concept of prior identity as a starting point, his question becomes roughly, "how does the ego know that it's seeing itself in the object if there's neither unified ego nor unified object to begin with"?  It's in trying to answer this that he brings in real objects, virtual objects, and their interaction.

The differenciation between self-preservative and sexual drives must be related to this duality between two correlative series. The self-preservative drives are, after all, inseparable from the constitution of the reality principle, from the foundation of active synthesis and the active global ego, and from the relations with the real object perceived as satisfying or menacing. The sexual drives are no less inseparable from the constitution of virtual centres, or the extension of passive syntheses and the passive egos which correspond to them: in pre-genital sexuality, actions are always observations or contemplations, but it is always the virtual which is contemplated or observed.

Following Freud, Deleuze sees both types of drives as pursing an object (ie. they are "objectal").  The difference is in what type of object they seek.  Freud imagined our sexual drives pursuing ordinary objects like tits and ass and antenna.  What he initially saw as the self-preservative drives were the ones he later discovered covertly pursuing a strange projected object which turned out to be our ego.  

For Deleuze, this order is almost reversed.  The sexual drives pursue strange virtual objects which correspond to what each of the passive local egos has a passion for repeating (as we saw this is really just what they are, and what brings them pleasure).  Because these objects arise locally, so to speak, they don't initially reflect us as a totality, but just part of (what is a yet to be defined) us, which is why next time we'll talk about Klein and her partial objects.  A full "extended passive ego" -- essentially, our selves as a subject, Freud and Kant's starting point -- still has to be synthesized as a contemplation-contraction of these virtual objects.  We'll see later that the contemplation involved in this synthesis is still passive, but it's built on memory, rather than the habit of the first passive synthesis.  

By contrast, in Deleuze's scheme, the self-preservative drives are the ones that aim at what we would call ordinary real objects like apples and water.  These drives seek objects capable of (temporarily) satisfying the biological realities that are needed to hold all the passive local egos together as a single overall unit capable of coherent action.  The objects are constructed as wholes (instead of parts) and motivate the action of a whole active ego (whereas with the sexual drives, we are thinking with our 'parts' so to speak).  

The fact that these two types of drives are distinct doesn't mean that either can exist on its own.  The only way to answer our initial question of how the full ego got constructed is to notice how the two objects reciprocally presuppose one another.  Each sexual drive finds its virtual object -- its reflection of itself -- contained in some aspect or part of the real object.  In fact, the real object is nothing but whatever contains enough of these virtual objects to satisfy the various sexual drives well enough that they cohere together and preserve themselves (reproduce through time).  These sexual drives though just started off as a motley collection of different parts and not an integrated whole.  This whole is only brought together through the passive contemplation of the virtual objects that compose the real.  

The fact that the two series cannot exist without each other indicates not only that they are complementary, but that by virtue of their dissimilarity and their difference in kind they borrow from and feed into one another. We see both that the virtuals are deducted from the series of reals and that they are incorporated in the series of reals.

This explanation is still too tentative and abstract.  We'll discuss aspects of this same structure again as we go through Klein and Lacan, and then return to see if we can manage to sum it all up more concretely.  For now, I think the key takeaways are that the real is associated with an active and total whole composed of unknown parts, the virtual is associated with parts in the context of a whole which has not yet been defined, and the two sets of objects only stand up by drawing on one another.