Which one is autonomous?

Next up on the bookshelf in the philosophy of technology project is Langdon Winner's Autonomous Technology: Technics-out-of-control as a Theme in Political Thought.  Since Winner is a professor of political science, the tone of this book differs markedly from the frankly speculative efforts of Simondon and Arthur.  While it's both accessible and fairly well written, it still suffers from the long windedness of someone who constantly feels the need to justify their academic existence.  So, for example, we only come to understand the overall design of the book at the very end.

We began our inquiry with the simple recognition that ideas and images of technology-out-of-control have been a persistent obsession in modern thought. Rather than dismiss this notion out of hand, I asked the reader to think through some ways in which the idea could be given reasonable form. The hope was that such an enterprise could help us reexamine and revise our conceptions about the place of technology in the world. In offering this perspective, I have tried to indicate that many of our present conceptions about technics are highly questionable, misleading, and sometimes positively destructive. I have also tried to lay some of the early groundwork for a new philosophy of technology, one that begins in criticism of existing forms but aspires to the eventual articu­lation of genuine, practical alternatives. (AT, 306)

This is a nice summary, and explains the rationale for what was an otherwise puzzling duplication of material between the first third of the book and its remaining two parts.  Winner spends quite a lot of time establishing that 1) we're worried about technology and 2) while some of these worries don't make a lot of rational sense, there is some real reason to worry.  I think both were important points to make, especially at the time of publication in 1977; it's clear that even today not everyone would agree with them.  And in making them, Winner addresses some interesting questions that often go unnoticed in our thinking about technology.  For example, asking if we should be worried about technology allows us to back up and suspend judgement for a moment so that we can see what's at stake in the question itself.  If we're afraid that technology is now somehow "out of control", then we clearly need to interrogate our notions of control and power.  With this question we immediately enter a political realm. 

But in the end, Winner's answer to this question is sufficiently not-earth-shattering (we thought we were the goal-setting master and technology the goal-obeying slave), and our perspective sufficiently jaded (by the century long failure of the modern, technocratic, or internet revolutions to bring about anything resembling utopia), that this first third of the book now reads as a elaborate and unnecessary justification of why you should think about these issues at all.  The attempt to maintain a facade of neutrality throughout this section -- "some people have said that modern technology qualitatively changes human experience"-- when it will later become clear that Winner is himself almost entirely a techno-critic who agrees with 'some people', introduces a lot of duplicate material.  So, for example, we see Jacques Ellul presented first as potentially just a wild-eyed prophet of techno-doom, only to find these same basic ideas re-presented, in the chapter on "Technological Politics", as compelling observations on the changing nature of human autonomy in a technological society.  It's enough to make us wonder whether Winner devoted a third of the book to shadow boxing a straw man just to keep in top academic condition, or whether political science was backwards enough (in 1977) that this approach was necessary to keep your job.

As the summary quoted above indicates though, once Winner finally legitimizes criticism of technology, he goes on to articulate two basic approaches to the problem of improving things.  He dedicates one chapter to "Technocracy" as illustrated by the theories of Don Price and John Kenneth Galbraith.  While the details differ, these theories hold that any political problems caused by the unintended side effects of technology can be resolved with better political solutions.  Both solutions may require an elite new interest group to be implemented, but they still fit squarely within our understanding of politics as the process of human self-determination.  A second chapter illustrates a much more radical take on the problem that Winner calls "Technological Politics".  This is the view associated with figures like Ellul, Mumford, and Marcuse.  They argue that modern technology so permeates human experience that to imagine we are free to direct it as we see fit is pure hubristic fantasy.  Technology now has a logic of its own, together with its own autonomous momentum.  And we can neither control it in any straightforward sense, nor even get off the train.  Winner summarizes two positions here:

   The first, the utili­tarian-pluralist approach, sees that technology is problematic in the sense that it now requires legislation. An ever increasing array of rules, regulations, and administrative personnel is needed to maximize the benefits of technological practice while limiting its unwanted maladies. Politics is seen as the process in representative government and interest group interplay whereby such legislation takes shape.
   The second approach, disjointed and feeble though it still may be, begins with the crucial awareness that technology in a true sense is legislation. It recognizes that technical forms do, to a large extent, shape the basic pattern and content of human activity in our time. Thus, politics becomes (among other things) an active encounter with the specific forms and processes contained in technology.
   Along several lines of analysis this book has tried to advance the idea central to all thinking in the second domain - that technology is itself a political phenomenon. A crucial turning point comes when one is able to acknowledge that modern technics, much more than politics as conventionally understood, now legislates the conditions of human existence. (AT, 323)

As you can see, Winner himself falls squarely into the latter, "technological politics", approach.  But because he conceives of his work as overcoming the objections of some putative technocratically optimistic Chad, he spends a great deal of time trying to specify exactly how the (assumed to be muddle-headed) idea that technology already is politics can be made to fit with a naively dualistic view of human agency.  This is a very useful exercise.  However, it's also an exercise that tends to dissolve its own starting point in a way that Winner doesn't quite appreciate.  Fortunately, this undermining actually takes us to the heart of the issue, and carries us right back into the terrain that Simondon explored  -- what does it even mean to be an 'autonomous' individual?  So in what remains, I'll trace Winner's careful dualistic reconstruction of how the sixties critics of Technological Politics might well have been onto something, and then look at how the means of this reconstruction paint him into a corner that he could only have gotten out of with the help of Simondon's concept of reciprocal causality.  

---------

Winner's central contention is that technology has changed how power functions in human society.  Whereas previously the power to steer society towards particular values may have rested with kings, or capitalists, or even 'the people', now this power increasingly rests with 'technology'.  This is the conclusion of the school of Technological Politics that Winner wants to reconstruct.

In twentieth-century social philosophy the conception of a self-maintaining technological society has recurred in a number of interesting and disturbing books - Oswald Spengler's Man and Technics, Friedrich Georg Juenger's The Failure of Tech­nology, Karl Jasper's Man in the Modern Age, Lewis Mumford's The Myth of the Machine, Herbert Marcuse's One-Dimensional Man, Sieg­fried Giedion's Mechanization Takes Command, and Jacques Ellul's The Technological Society. These works contain a widely diverse collec­tion of arguments and conclusions, but in them one finds a roughly shared notion of society and politics, a common set of observations, assumptions, modes of thinking and sense of the whole, which, I be­lieve, unites them as an identifiable tradition. Taken together they ex­ press an inchoate theory which takes modern technology as its do­ main. (AT, 174)

But what is 'technology'?  Isn't it just a neutral collection of means for achieving what are always ultimately human ends?  How could these means, when we loosely group them together as a whole, themselves gain the power to determine the ends of human society?  Aren't we just anthropomorphizing 'technology' by treating it as an entity or organism or subject in its own right, and then turning this fiction into a political actor?  Does this represent progress in political theory, or its mystification?  To reconstruct what it might mean for technology to be a political actor, Winner first has to rescue the idea from those hard-headed critics who still believe that only human individuals can have political agency.  In other words, he has to rephrase what sounds like the description of an organism -- "a self-maintaining technological society" -- as nothing more than some peculiar thing that happens when humans have a whole lot of tools at their disposal.  

Are there certain conditions, constraints, necessities, requirements, or imperatives effectively governing how an advanced technological society operates? Do such conditions predominate regardless of the specific character of the men who ostensibly hold power? This, it seems to me, is the most crucial problem raised by the conjunction of politics and technics. It is certainly the point at which the idea of autonomous technology has its broadest significance. (AT, 173)

What characteristics allow us to confuse technology with a self-maintaining organism?  Well, one is Winner's very broad use of the term 'technology'.  He doesn't simply mean the various apparatus that we commonly refer to as machines.  He also includes the entire body of human "technical activities", that is, all goal seeking rational behavior, which he refers to as "technique" (following the French use of the term).  Finally, he also means to address the various organizations, or networks, that function like machines, regardless of whether they consist of entirely human or artificial parts, or some mix of the two.  So it's clear from the outset (AT, 11-12) that there is no clear dividing line between human society and technology.  The latter is being construed not only to include much of human social behavior, but even the parts of individual behavior that conform to a 'rational' logic of means and ends.  Already, we can see that the question of which one is an autonomous organism is complex and messy.  We might be talking about technology, or society, or even, if we stop taking it for granted for a moment, a human person.  In fact, most of the time Winner's term "autonomous technology" means the complicated systemic interaction of these three.  One of the secrets of thinking about technology, which to me was not apparent when I began with Simondon, is the way it reveals how badly we need a better definition of 'organism' or 'individual', along with the closely related notions of 'control' and 'autonomy'.

Winner, however is a social scientist, not a philosopher, so he doesn't try to give us a definition of an organism.  Instead, he appeals again and again to the various ways that technology appears to acquire a life of its own by calling into question human's ability to control and limit it.  In other words, the proof of technology's autonomy lies in the way we lose our own.  For example, modern technology seems to constantly expand its reach, and at an ever accelerating rate that hairless chimps find difficult to keep up with.  In the process, it also seems to suck in everything in its environment and reshape it as raw material for technological growth.  And this logic applies not only to natural resources, but even tends to reshape human beings into suitably standardized and machined parts, as if it were technology were incorporating or disciplining us.  Technical systems seem to inevitably become megatechnical systems, to use Mumford's term.  The scale of an individual apparatus or organization tends to increase in an endless quest for efficiency, and somehow each tends towards a complex interaction with others.  In addition, technical development doesn't always seem to respond to human goals. If it's really just working for us, then why are there so many unintended consequences that don't seem to benefit any human?  We wanted universally accessible knowledge and instead we got TikTok.  We want to get from A to B, and somehow we got an endless parade of SUVs.  Technological means seem to become ends in themselves with surprising regularity.  Finally, technology doesn't ever seem to want to be turned off.  A decade ago we might have illustrated this with the mushrooming of highly redundant data center bunkers.  Today we can imagine adding an AI whose job is to motivate those notoriously fickle humans to continue supplying fuel and replacing servers.  But Winner has in mind something more systemic that has long involved the interaction of human desire and technical means, which he calls the threat of apraxia.  Imagine a politician who suggests shutting down the internet and you'll immediately comprehend the idea.  Even if technology seems to require human participation in it as both as efficient means and final justifying end, and even if we could theoretically abandon it and survive just fine, somehow the idea of stopping is always made unthinkably apocalyptic.  For Winner, all of these examples serve to illustrate that we are not in control of our modern technology the way we imagine ourselves in control of, say, a hammer. 

In its centrality to the daily activity and consciousness of the "employee," the function-serving human com­ponent, the technical order is more properly thought of as a way of life. Whatever else it may be, a way of life is certainly not neutral. Oppor­tunities for "use" or "control" that the human components have within this system are minimal, for what kind of "control" is it that at every step requires strict obedience to technique or the necessities of techni­cal organization? One can say that the "control" is exercised from the center or apex of the system; this is true, although we shall soon see that even this has a paradoxical character. But in terms of the function­ing of individual components and the complex social interconnections, "control" in the sense of autonomous individuals directing technical means to predetermined ends has virtually no significance. "Control" and "use" simply do not describe anything about relationships of this kind. The direction of governance flows from the technical conditions to people and their social arrangements, not the other way around. What we find, then, is not a tool waiting passively to be used but a technical ensemble that demands routinized behavior.
   In this way of seeing, therefore, the tool-use model is a source of illusions and misleading cues. We do not use technologies so much as live them. (AT, 201)

But if humans aren't in charge of technology, then what is?  Has Winner given us any reason to think that technology is in charge of itself, as the word autonomy suggests?  And if modern technology constitutes a new organism, when was it born and above all, what does it want?  Here, Winner's methodology begins to create problems.  On the one hand, he's sought to be hard-headed and ignore the idea that there's anything mystical or vital involved here.  'Technology' is really just a short-hand for the systemic interaction of mechanical apparatus, human technique, and organizational networks that we see in the world around us. Humans are the only real agents in this system, and the problem is simply that we have lost this agency, this control. The idea that it has thereby passed to some other entity is just an expression of our Frankenstinian fear.  On the other hand, Winner has actually described something that walks and talks like an organism in its own right.  He hasn't emphasized this emergence, but as I suggested earlier, as the book progresses he increasingly talks as if there is a real "self-maintaining technological society".  He has even sketched what this system or organism wants.  It seems to want to expand and incorporate everything, and it seems to want to turn its ever-evolving means into ends in themselves. 

In fact, Winner will even go on to talk about how these tendencies to expansion ("technical imperative") and self justification ("reverse adaptation") actually serve to meld together the two perspectives of Technocracy and Technological Politics we discussed earlier.  Does technology require legislation or is it legislation?  Apparently the answer is yes.

  On the basis of the revised notions of technological politics pre­sented here, however, I want to suggest an alternative conception of technocracy. It is one which avoids the pitfalls of the centralist-elitist notion and which, I believe, comes closer to defining the crucial problem that writers on this question have been aiming at for some time. The conception, although difficult to formulate in terms of pre­cise indexes of measurement, could certainly be tested.
   I offer it as follows. Technocracy is a manifestation of two influen­ces upon public life, which we have dealt with at some length: the technological imperative and reverse adaptation as they appear to a whole society with the force of overwhelming necessity. From this point of view it matters little who in specific obeys the imperative or enacts the adaptation. It is of little consequence what the nature of the education, technical training, or specialized position of such persons may be; indeed, they need not be technically trained at all. Similarly, the issue has little to do with whether there is a "center" of decisions, who occupies it, or what their personal, professional, or class interests may be. (AT, 258)

Here Winner reinterprets the technocrats themselves as mere dispensable tools of technology.  Technocracy becomes a description of how the system works, not who controls it.  Technology gives itself the legislation it requires, though this legislation only serves to unleash more technology rather than allow some human elite to limit it.  This view seems perfectly accurate, and is very close to the critique of Marxism that we saw in Simondon.  The elites are slaves of the machines just as much as the workers, notwithstanding their more comfortable yachts.  Placed in the same situation, anyone would act as they do, because the system has made any other behavior unthinkable.

However, this insistence that the system runs itself, that it doesn't require an elite who understands it, or even a center from which power is constructed poses a tricky question.  What leads us to call it a single system at all?  Winner has described something that we can only recognize because it has a coherent logic of its own that resists human efforts at control.  But then he claims that this same thing is so decentralized that each megatechnical system charts its own course.

Finally, the view I am taking here does not find the essence of tech­nocracy at the center of all centers. If what we have seen is correct, then one would expect a dispersion of power into the functionally specific large-scale systems of the technological order. The systems do on occasion appeal to the central decision-making organs of the state for support and assistance, but it is incorrect to say that in so doing they necessarily yield control of their affairs. Their tendency is, in fact, to resist the final centralization Ellul predicted. The direction of gov­ernance by technological imperatives and reverse adaptation runs from megatechnical systems to the state. (AT, 261)

I think Winner sees this idea of a decentralized technocracy as his crucial contribution to the theory of Technological Politics.He thinks his more rigorous materialistic methodology allows him to reach the same type of conclusions as Mumford and Ellul, but without the whiff of vitalism or human essentialism that (supposedly) clings to those authors.  This is why he emphasizes here that his theory requires none of the state centralization that the others apparently see as inevitable. 

But this idea won't work as it stands.  If each megatechnical system has its own "affairs" to control, and these are not organized by a central state, then we really don't have a system at all.  Why would all these individual megatechnical systems all share common goals, and why would the resulting overall behavior have any coherent logic that we can recognize?  And if megatechnical systems resist final integration by the mega-mega-technical state, then why doesn't this logic apply inductively, and so prevent technical systems from ever forming mega-technical systems in the first place?  Where's the secret sauce that allows certain technical systems to come together so that they act like distinct centers that can control their affairs and resist centralization?  Conversely, if these systems are the only real agents, then how is it that we've identified a characteristic expansion and self-justification of 'technology' as a whole?  Was this just a sloppy shorthand for describing the combined effect of distinct megatechnical systems?  But then why do they all seem to have the same combined effect where they don't just individually expand but also constantly produce new megatechnical systems?  Winner appears to be arguing that we are slaves to many machines which are not necessarily compatible.  This is probably true, but doesn't help us see how the individual machines formed nor why they seem to come together to form a system. 

What we need here is the idea of causal feedback loops.  This was the core of Simondon's theory of the technical individual.  We have to distinguish the individual from the element and ensemble levels of organization, and the key to that is examining where self-reinforcing causal loops get formed.  At this point, it feels like an anti-climax to admit that I don't completely understand how to construct the theory we need here.  It  seems that Simondon developed the idea in his book on individuation, and merely applied it in On the Mode of Existence of Technical Objects.  So presumably we'll return to this point.  For now, I can only sketch a few things that our discussion of Winner has gotten me thinking about.  I think we need to think about something like an ecology of feedback loops.  On the one hand, the existence of individuals clearly requires a positive feedback loop.  I'm imagining something like an autocatalytic set of causal relations.  If the internal functioning of the individual doesn't cause the individual to continue functioning as it does then we're not going to see that individual around for very long.  On the other hand, it seems individuals also require negative feedback loops to maintain a separate identity.  We need a condition of closure to create a distinct self that can be maintained.  Left to itself, a self-reinforcing autocatalytic set wouldn't maintain the equilibrium we associate with individuals, but would simply expand without limit.  It wouldn't make a copy of itself but would simply more of itself.  This is the kind of behavior we associate with crystals, not living individuals. 

This is about as far as I've gotten with the idea so far.  It's not obvious to me that one of these loops should operate exclusively internally and the other externally, between the individual and the environment.  This was one of the interesting, though confusing, points in Simondon.  The individual and its associated milieu are held together like two poles of a teepee.  The individual can't be what it is without a suitable associated milieu, but this in turn can't be as it is without the actions of the individual.  Winner describes a similar phenomenon when he talks about the way technology reshapes both the natural and human environment in a way that leads to more technology.  At the same time, the individual is not just the same thing as its associated milieu, so there has to be some mechanism that creates this distinctness. 

The Nature of Technology

My complication had a little complication.  Loyal readers recall that we are following the thread left by Deleuze's reading of Leibniz and attempting to deepen that question of what it means to be an individual.  Roughly speaking the question is why there appear to be distinct individuals if the world is ultimately all the same stuff.  Or, conversely, how there can be a world at all, if all that really exists are distinct individuals.  In short, individuals and the world seem as simultaneously necessary and incompatibly dual as form and emptiness.  

Simondon's On the Mode of Existence of Technical Objects was our first stop on this journey largely because it was already on the shelf.  While the book definitely dealt with the question of individuality, I found the perspective Simondon opened up on technology so timely and intriguing that I now find myself on a tangent to my tangent.  Simondon is of course not the only one who considers technology as a frozen or objectified form of human goal seeking behavior.  Nor is his the only framework for examining it, despite how useful I think it is.  In fact, it turns out that I have quite a few books about technology that I've been meaning to read.  So permit me a digression.  And if it seems too far afield at first, consider the next few posts in light of Simondon's most fundamental idea -- goal seeking behavior is only ever part of what it means to be an individual, and examining our technology lets us examine this part of ourselves.  

The first technology related book I grabbed off the shelf was W. Brian Arthur's The Nature of Technology.  I've been meaning to read this for 10 years, probably as a result of Kevin Kelley's What Technology Wants.  Arthur, it turns out, is one of those rare contemporary economists who have something useful to say.  In this book he tries to channel the spirit of Schumpeter and develop a theory of technology that would help us adequately understand its relationship to the modern economy.  While it's pretty sketchy and high level as a theory, at least it's very clearly stated.  And surprisingly it has quite a lot of overlap with Simondon.  So in what follows I'd like to summarize's Arthur's theory as a means of further exploring the ideas we've already been dealing with.

The Nature of Technology breaks down into three sections that deal with three distinct though related questions: what is technology in general? how do new individual technologies arise? and how does technology as a whole evolve?  While, interestingly, Arthur doesn't actually begin with this point, he claims that the essence of technology is the harnessing of some natural phenomenon for a purpose.  In other words, technologies are actualized or operationalized forms of goal seeking behavior.  They are real teleological systems.  This comes surprisingly close to Simondon's idea that a machine is a (partially) concretized human thought of a mechanism that accomplishes a goal.  Perhaps because it sounds either too vague and philosophical or simply too commonplace, Arthur chooses not to lead with this observation.  He also tends to place the focus more on the phenomenon side of things.  That is, he talks about how technologies capture physical phenomena like the interference of waves, or biological phenomena like the self-replication of DNA, or even human behavioral phenomena like greed (if we consider something like 'the market' as a technology).  He's clear though, that these phenomena have to be captured in order or us to speak of a technology.  While philosophical, this point is fundamental.  It's really this baked in teleology that separates technology from other natural phenomenon; without a goal, there is no technical object, but simply a functioning natural system.  Arthur, however, does not specify that this goal needs to be a human goal.  As we'll see, his theory is that technology, as it evolves, can create its own needs that serve as new goals for further technologies.  While we might at first be tempted to contrast this with Simondon's emphasis on the way technology requires life, let's hold off on that till we unfold his full theory. 

To its essential definition as a teleological organization of phenomena, Arthur adds an empirical observation about how technology often seems to be organized hierarchically.  Technologies exhibit a fractal structure, where a main technology contains elements that are themselves smaller technologies, that is, systems which harness a phenomenon to a purpose, in this case the overall purpose of the main technology.  Especially these days, a given technology can be a complicated assembly of sub and sub-sub technologies whose functions are integrated to serve the purpose of the highest level technology.  Imagine a jet aircraft with its incredible complexity and redundancy of power, navigational, and safety systems, all trying to keep you in the air.  While Arthur emphasizes the hierarchical aspect of this organization, I think this is more properly seen as a particular (though very common) type of a more general notion of combination.  Technologies can be combined to produce other technologies.  Technology is thus recursive and modular.  Together, teleology and combination define the technical world.  Goals can be made into objects that capture and organize natural phenomena, and then these objectified goals can be combined to accomplish new goals.  This obviously correlates pretty well with Simondon's idea that the technical individual integrates technical elements into a teleological mechanism.  Here again, we might be tempted to contrast Simondon's careful causal analysis that lets him decide which technical groupings count as elements, and which have the sort of reciprocal causality that turn them into true individuals, with Arthur's rather casual use of the notion of sub-assembly.   But, again, let's put this question off till we finish with the whole theory. 

Finally, Arthur describes one more feature of the structure of technology as a whole.  Technologies can be broken down into smaller modules that are integrated to produce an overall functioning.  But technologies also tend to cluster into "domains" of related functionalities at the level above the integrated technical individuals.  Arthur has in mind things like the digital domain, or the domain of electrical (as opposed to say, steam) power, with all its related technologies like amplifiers, alternators, rectifiers, load limiters, etc ...  While he describes these domains as technical "languages", I'm not sure this analogy offers anything beyond the observation that there are rules of thumb or habits for effectively combining individual technologies to create new ones.  A domain provides a set of ready-made possibilities for combination, which allows new combinations to be produced relatively easily.  These elements and their 'grammar' however, do not form a tightly integrated functional unit like an individual technology and its modular sub-components, but constitute a fuzzy cloud of more or less related technologies.  At this point, it's perhaps unsurprisingly that Arthur's domains correspond well to Simondon's technical ensembles. These are collections of technical individuals held as distinct rather than integrated into a single unit.  The ensemble as a whole is not itself a technology because it is not a teleological mechanism, not a concrete solution to a specific problem.   Instead, the domain functions as a background of possible relations between technologies, which allows it to be the ground of new technologies -- in Simondon's scheme it's the technical ensemble that creates new technical elements.  What it creates are not solutions to particular problems, but objects that contain specific tendencies that provide for possible solutions to all kinds of problems.  

We can already anticipate how Arthur's definition of technology will answer his question about how new technologies are produced -- combination and recombination.  In a sense, we might say that this modularity follows from the definition of technology as a teleological mechanism.  Defined goals, after all, can always be be broken down into sub-goals like steps or stages, so perhaps its no surprise that you can break a whole down into a combination of technical parts.  More surprising is the fact that you can combine existing parts to create a completely new whole.  It's not obvious that accomplishing two separate goals should serve do any single thing that someone wants.  One of the strong points of Arthur's theory is that it deals with both of these cases, which we might respectively think of as top-down and bottom-up creation.  Both of these call a new technology into existence by linking a problem to a concrete solution.

New technologies frequently come into being in the prosaic manner that he calls "standard engineering".  Some off the shelf components are combined according to a known pattern in order to solve a specific problem.  My cat fountain needs a water pump.  This requires an electric motor, which in turn requires a power system.  These are readily available, well understood components that simply need to be combined with some economic and design constraints to lead to the production of the concrete new technical object that now sits in my bathroom.  Similarly, once an object like this has been constructed, new versions of it can be made by changing out the components for improved ones, or adding safety systems or lights or what have you; these accrue to the basic functionality without substantially altering it.  Arthur calls this process of descent with variation "structural deepening".  Together, these two plug and play approaches to 'innovation' account for most of our production of concrete technical objects.  Certainly these forces are more than adequate to account for the iPhone ... 13.

However, once upon a time, there was a first iPhone, and before that the many qualitatively new technologies that got their start somewhere.  These truly novel types of technologies are still combinations of existing elements, for the simple reason that you can't make a real something out of nothing.  But neither of the two forces we have described so far is capable of producing a completely novel solution to a problem.  This is because both standard engineering and structural deepening operate in a top-down manner.  They both begin with an already known solution to a problem, an already understood principle that effectively harnesses certain natural phenomenon and provides the structure of the technology at the outset.  However, because technology is modular and combinatorial, it's possible to hit on completely novel combinations that carry out a hitherto impossible task or achieve an existing goal using a completely new overall principle.  This bottom-up approach is what we mean by true "invention".  Arthur persuasively argues that this is not some spark of genius that comes from nowhere but simply a new combination of technical elements, usually due either to inventive thinking by analogy, or to using a completely different set of building blocks to construct a new solution to an old problem (which Arthur calls "redomainings").  This latter way of creating new technologies holds particular fascination for Arthur, and for good reason.  He sees that when an old problem is posed in the language of a new domain -- eg. power production shifts from wind and water to fossil fuel combustion, or calculation shifts from 'calculators' to computers -- veritable revolutions can follow.  In Simondon's language, we would say that the new ensemble is capable of producing elements with completely different possibilities for connection baked into them, which leads to the construction of qualitatively novel technical individuals.  

The closest biological analogy for these bottom-up types of innovation would be the horizontal gene transfer in bacteria.  Together with some mechanisms for moving genes around, the bacterial gene pool as a whole forms a sort of technical ensemble from which individual bacteria can draw on newly produced elements.  Of course, this examples falls short of what Arthur has in mind, since there is a single bacterial gene pool where he sees many distinct domains whose elements are easy to recombine, and any of which might serve as the basis for a novel solution to an existing problem.  Still, the biological analogy brings us to the final question Arthur tries to tackle, which is really the one that motivates the whole book -- how does technology as a whole evolve?  The answer, at this point fairly obvious, is combinatorial evolution.  This form of evolution, while not unheard of in biology, is clearly far from the Darwinian norm.  And it leads to a pretty startling conclusion.  If we define life by the ability to produce more of itself, then we are forced to acknowledge that technology as a whole is alive.  This is less vague and vitalist than it sounds.  Just like Simondon's, Arthur's scheme is not driven by the individual technology's ability to produce identical copies of itself.  And yet technology is still in a sense self-producing, or at least self-extending -- new technologies are produced as combinations of existing ones, and these combinations can go on to become elements in further combinations, ad infinitum.  Arthur even invokes Maturana and Varela's concept of autopoesis (though I would agree with Haraway's criticism of the use of this term here).  In short, technology creates more, and different, technology out of itself -- it is alive.  

Naturally, however, these new technologies are not created in a vacuum, but respond to the goals of their 'users'.  Remember, technology is essentially an objective goal directed system.  So there's immediately some question of what we mean by 'self' producing, if the goal has to come from outside the technology.  Arthur, however, offers a very interesting idea in this context.  He argues that technology not only produces a supply of goal directed systems, but also gradually bootstraps a demand for them into existence.  The first technologies were produced when humans captured simple phenomena like combustion (fire), elasticity (bows), or mineral cleavage (lithic reduction, aka flint knapping) and employed these to human ends.  Gradually though, the combined use of the bow, arrow, and fire themselves call into being new possible goals such as the hammer that improves the arrowhead and the smoking that preserves the meat.  These do not cease to respond to human goals, at least in the final analysis, but as technology develops through combinatorial evolution, it posits many subsidiary goals that demand technical solutions, as well as opens up many final goals that we didn't even know we had at first.  Arthur's culminating idea is that the economy is what mediates this supply of and demand for technical solutions.  It's at this point that he works his way back around to renewing Schumpeter's thoughts about creative destruction.  The structure of the economy reflects not only the technology available for our means of production, but the human and technical needs created by the operation of this very technology.  Thus when a revolutionary new technology comes into being, it has the capacity to shift the entire landscape of both production and consumption.  We stop wanting a faster horse and start wanting a horseless carriage.  The result of his theory is a much more realistic view of the economy as a complex system that evolves through a mechanism more like punctuated equilibria than the classical couldn't-be-a-real-dollar equilibrium fantasy.  

Now that we've covered the entirety of The Nature of Technology we're finally in a position to come back to the loose threads I mentioned earlier and consider Simondon and Arthur in light of one another.  While there's a lot of overlap, we can see that there's also some confusion around the question of what it means to be alive and how this relates to teleology as the defining feature of technology.  

For Simondon, teleology is associated with homeostasis (the primary goal is always to exist), which is only one aspect of life.  Life, by his definition, requires a creative exploration and production of aims that is impossible for a technology by itself; at bottom the technical object always embodies a fixed purpose given to it by life.  This is why his model of the technical individual begins with the artisan and progresses through machines that only reconverge with the natural world at the end of a long process of concretization.  These are systems whose teleology arises from, on the one hand, an reciprocally causal internal interaction of parts, and on the other hand, an interaction with the environment where the individual couldn't exist as such without the corresponding environmental structure being what it is, and vice versa.  As always, the image is of the teeppee or vortex that defines the 'self' which lies behind the teleology of homeostasis.  So what's really alive here is not the homeostatic or self-regulating individual, but the organic ground that contains the possibility of such an individual.  Once they come into existence, the technical individual is a machine like any other -- it simply executes its aim, in this case to stay in existence.  With this main goal in mind, a host of subsidiary goals falls into place.  And this newly created self-reinforcing feedback loop also becomes a unit in interaction with similar units and with the environment as a whole.  In the end, Simondon's idea of life boils down to this individual's openness to its milieu, to the possibility of having their teleology altered or re-constructed in some way.  The 'self' involved in individual homeostasis is actually a sort of dead husk.  We might say that technology is a means to a purpose, but the purpose is simply living -- the process of creation of new purposes.

Initially, Arthur's scheme may seem at odds with this.  He doesn't think technology needs life to give it goals since he considers it an autopoetic system all on its own.  But much of this difference becomes superficial when we realize that for Arthur it's only technology as a whole that begins to resemble the organic.  And even when he speaks of the self-producing aspect of technology, he is immediately forced to bring in its continual mediation by the human economy.  So what's alive here is not that different from Simondon's creative ground.  In both cases there's a life that flows through individual technologies, that can't be reduced to the execution of single purpose but consists in the generation of purposes.  We can follow Simondon and call this creativity life, or we can keep Arthur's terms and attribute it to technology as a whole -- in either case, we're talking about a life of technology.  All of the interesting stuff is happening at the intersection of life and technologies.  Technology evolves because the technical individual is always immersed in both a technical ensemble and a living milieu.

The one remaining point of difference lies in the fact that Arthur's analysis contains nothing that corresponds to the reciprocal causal closure that defines the individual for Simondon (and thus allows him to rigorously distinguish it from the element and the ensemble).  We are left with the impression that Arthur's sees any combination of technologies as a new technical individual.  While this may be true in principle, when he begins to discuss his toy model (pg. 181) of technology's combinatorial evolution it becomes clear that not all combination are actually viable in all environments.  Arthur's model actually selects for certain technologies that 'come alive' as a function the other technologies in the environment.  In this case, the top level of interesting goals are defined in advance and fixed at the outset of the model, so the only open question is what subgoals are used to sustain these once they appear.  In real life, these top level goals would themselves need to appear as the outcome of the way a complete techno-economic system functions.  Only some new combinations of technologies will result in an achievable goal that interests anyone.  And these might be extremely difficult to predict in advance.  As we observed earlier, sub-assemblies are easy to see once you know the main assembly, but a new main assembly is hard to construct from given sub-assemblies.  So here as well, the differences between Arthur and Simondon may not be as great as they first appear.

Concretizing the Nondual

The third and final part of On the Mode of Existence of Technical Objects represents a substantial departure in style and and content from the first two.  All the detailed examples of the inner workings of concrete technical objects disappear, and instead we get a much more abstract treatise about the genetic relationship between technical thought and other types of thought.  In fact, generally I found Simondon's complicated schema for organizing the various types of thought a bit too abstract.  Nevertheless, there are some really interesting ideas here, and I think the places he accords to philosophical and aesthetic thought in particular exerted a substantial influence on Deleuze.  So while I don't think I'm going to delve too deeply into Simondon's schematization of how thought develops, I do want to reflect a little on the motor that powers this development.  By understanding this motor, we'll see how it relates to the earlier description of the development of technics, and thus grasp the fundamental concept of the book as a whole.  And we'll also glimpse an outline of an early chapter in my forthcoming study: The Nondual Delueze.

The motor that powers Simondon's whole philosophy is the nonduality of figure and ground.  A simple concrete image of this would be something like Rubin's Vase or any of the other images beloved of Gestalt Psychology.  In all these images, figure and ground are clearly distinct, but reciprocally determined; the figure is only a figure on the basis of a specific ground, but the ground only appears as the ground of a specific figure.  They are co-adapted to the point of inseparability, and yet we seem to be able to perceive only one at a time.  So figure and ground are not the same, but they're also not different.  They form a sort of unity, but not one composed of independent pieces.  In short, figure and ground are nondual.  

I've called this nonduality a motor because Simondon conceives of it in dynamic terms.  An original nondual splits into figure and ground.  And then each side of this division split again into its own relative figure and ground.  Simondon conceives of each of these splits by analogy to the idea of a physical phase transition.  The same 'stuff' (in this case some nondual ur-ground of pure potential) can behave very differently depending on the circumstances.  However, while this schema is dynamic in the sense that each split generates a new figure-ground pair (and the method is hence is "genetic"), Simondon immediately distinguishes it from that other great philosophy of movement -- the Hegelian dialectic.

This schema is very different from the dialectical schema, because it implies neither necessary succession, nor the intervention of negativity as a motor of progress; furthermore, opposition, within the schema of phases, only exists in the particular case of a two-phased structure. (METO, 173)

[To these comments, we might add some related thoughts Deleuze articulated in Chapter 5 of his Nietzsche book.  The Hegelian dialectic is the sterile adventure of a pre-determined self-identical unity.  It's crucial not to confuse monism with nondualism, not to denature immanence by putting it inside anything other than itself.]

In fact, while Simondon will also go on to describe various ways that the two sides of each respective split can be rejoined or at least reconciled, his rigorous use of the phase transition metaphor gives this synthesis a completely different character than the dialectic.  For example, the sentence just prior to the citation about the dialectic refers explicitly to the idea of the triple point.  

The existence of a plurality of phases finally defines the reality of a neutral center of equilibrium in relation to which there is a phase shift. (METO, 173)

We have to keep this image in mind when Simondon frequently slips into talking about the types of synthesis capable of bridging two phases as if they were "seeking unity".  The phase equilibrium point doesn't negate the existence of the phases, or unify them in some final moment of aufheben.  On the contrary, it's the point where both phases, both figure and ground, can coexist as distinct.    

Simondon describes this nondual phase transition schema as genetic, because it generates the whole variety of human experience, including, of particular interest in this book, technical experience.  While he talks about the schema in terms of the generation of novel types of thought, it's more fundamentally a question of constructing the individuals who can have these thoughts as an experience, as a being-in-the-world.  This means that, in the final analysis, METO should be seen as an application of Simondon's theory of individuation to technology (a theory I presume is fully developed in Individuation in Light of Notions of Form and Information).  The splits into figure and ground, together with the equilibrium points that put the split terms back into relation, are all "modes of existence" that define an individual at a given instant. 

Now, having criticized Simondon's scheme as too abstract, I've given an even more abstract summary of it.  Perhaps a diagram representing the specific splits and reunifications would help bring us down to earth.

As I observed at the outset, the scheme is complicated.  But the basic idea is that every split is a split between figure and ground, between object and subject, and every double sided arrow represents a new nondual equilibrium of what has been put asunder.  So the magical splits into technics and religion before aesthetics intervenes as a mediator that reconstructs the lost "unity" (Simondon never uses the more accurate term "nonduality").  The technical mode of existence approaches the world as a set of figural objects available for human use, while the religious mode emphasizes the grounding totality of the divine subject.  Then, technics and religion each split again into figure and ground, though this time Simondon talks about a split between each domain's theory and praxis, rather than subject and object.  Technics splits into engineering, which deals with the figural forms or functional technical schemas we've seen elsewhere in the book, and physics (in the Greek sense), which deals with inductively investigating the grounding nature that powers all these schemas.  Likewise, religion splits into theology, which attempts to deduce the logical ground of being, and morality, which judges the figures of human action.  In an interesting twist, Simondon regroups these second order splits into ethics, which contains their practical, figural components, and science, which groups their theoretical, ground components.  Science thus has both a technical and inductive side, as well as a religious and deductive side (which Simondon associates with mathematics).   In other words, both science and ethics contain a mix of figure and ground elements -- eg. science is an interaction between the inductive ground of the figure (physics) and the deductive figure of the ground (theology).  However, it's only philosophy, by relating science and ethics that finally establishes a new equilibrium which aims to recover the aesthetic and magical unity.

This kind of schema raises a bunch of questions, and Simondon doesn't do much to address them (at least here).  My biggest general objection is simply how abstract this all is.  Does it really help us understand theology, say, or aesthetic experience?  And should we really treat either of those as a monolithic entity?  I might argue that various aspects of this schema -- like the idea of a phase transition, and the nondual reciprocality of figure and ground -- found their way into Deleuze's philosophy.  If that's correct though, they appear as the outcome of a concrete investigation of a particular philosopher or subject matter.  Starting with a specific situation seems to me a much better method of approaching the relationship between various modes of thought or human existence.  Then there are particular issues that cry out for clarification.  For example, the magical universe, which Simondon describes as a network of related special points (he does not use the term "singularities"), is a world in which figure and ground are already distinct, but nevertheless reciprocally interwoven so as to be inseparable.  He calls it the first "structuration" of the world.  Which only begs the question of what the world, the ground, was like before it was structured.  No doubt, this is a difficult issue, but some take on it seems essential.  And then there is the squirrely question of what sort of subjectivity is associated with the equilibrium types of thought.  Thought with a clearly distinct object, even if this object is the human subject, implicitly defines a relative subjective position.  But what of a thought that holds subject and object as simultaneous aspects of some larger experience?  Where does the thinker of such a thought stand, so to speak?  Simondon talks about "philosophical intuition" as a mode of knowledge between the operational mode of ethics and the contemplative mode of science, one that integrates those modes by following the trajectory of their development (diagrammed above).  In addition, he makes clear that such a thought would be "analogical", in the sense that it sees each of these modes (in fact, every mode) as an analogous manifestation of the figure-ground split that power everything and brings anything into existence.  In a moment reminiscent of the way each of Leibniz's monads contains the whole world, we see that every process of individuation is ultimately an analogy for every other.  in a sense, there's only one process.  However, just saying that there should in general be such an intermediary type of thought, doesn't sufficiently explain how it works.  And in particular, calling this 'analog' thinking seems to invite the obvious question of what, exactly each process of coming-into-being is an analogy for?  While I'll touch on this problem below, I'm curious whether Simondon addresses these issues more fully somewhere else.

At this point, you may be wondering what any of this has to do with the titular technical objects.  There are actually multiple levels of connection between Simondon's genetic analysis of thought and the theory of the technical object he presented in the first two parts of the book.  But the most important and interesting one is the way both theories revolve around a process of recurrent causality or feedback.  From the outset, Simondon defined the technical individual as one one whose elements come together in such a tight mutual conditioning that none of them would function as they should without all of them functioning that way.  This recurrent feedback created an individuated unit that concretized an abstract technical schema as a real technical object. In other words, the process of concretization we discussed earlier is actually a process by which the two aspects of the technical object (engineering and physics in the diagram) are synthesized into a single functional unity.  Technical concretization is a analogy for how a neutral equilibrium can be established between the phases of figure and ground.   Figure and ground become involved in a feedback loop that ends up blurring their distinction in the same way that technical concretization ended up blurring the distinction between machine and associated milieu.  And this feedback between figure and ground is actually genetic.  It creates a new type of thought or technical object.  

While I haven't explored all its implications, I think that adding the idea of recurrent causality to our understanding of nonduality has the potential to open up a deeper understanding of the latter.  For example, people often speak of a "unitive" meditative state.  However, this word hasn't described my experience very accurately (so far).  To me, its feels much more like an experience of non-difference that centers around an interaction with the world so fluid that there's simply no room left for a fixed identity.  What seems to have disappeared here is not the distinctness of objects or experiences, but any distinct supplementary dimension that would be required to represent those experiences as belonging to a subject.  The sense is of becoming flush or level with experience, being experience rather than having it.  Does the concept of recurrent causality help us conceptualize something like that?

But I digress. I mentioned that there were several levels of connection between part 3 and the earlier parts.  Now you can probably see how this is actually a result of the way Simondon scheme makes everything an analogy for figure-ground feedback.  So it appears at the level of the split between technical figure and technical ground as technical concretization.  But it also appears at the level of the relationship between ethics and science.  Recall that one of Simondon's deepest points was that cybernetics has the potential to help us see how technical finality (goal directed utility) is just one mode of living.  Modern machines are really just frozen forms of human intention.  They can't be understood without reference to their concrete inner workings, but they also cannot be understood without reference to the life of the humans who use them.  In this context, Simondon's goal was to create a technical culture that functioned as a higher level, virtual, analog of the artisan.  This culture would investigate how machines worked, what they tended to be used for, and, crucially, would other possible uses we might have for them, and thus for ourselves.  In short, he was encouraging us to become involved in a feedback loop with our machines at a level between the technical and what we might colloquially call the philosophical or even religious.  What are all these machines for?  What are they doing to us?  What are we doing to ourselves?  What is the meaning of life?  

Of course, the point is not to pronounce judgement on this question, but just the opposite.  We want to open it up.  And one obvious but often overlooked ways to do that is by examining the ever more salient fact that we surround ourselves with machines.  They are clearly a part of how we become who we are today.  Likewise, these machines only become what they are, that is, become fully technical, when we restore their evolving human context alongside a deep understanding of their inner workings.  In this sense, no object can be technical in and of itself.  Technicity is a phase of development that can only be fully understood through Simondon's more abstract scheme.  Thus we could describe the recurrent causality that creates a neutral point amongst the phases as a sort of becoming-technical that can, perhaps counter-intuitively, serve to restore our broader integration with the world.  

The technicity of technical objects can thus exist at two different levels: original and primitive technical objects, which appeared as soon as magical thought ceased having an important functional signification, are indeed the real depositaries of technicity, as tools and instruments; but they are objects only to the extent that they can be put into action by a user; the users gestures also belong to techni­ cal reality, even if they are contained in a living being that places its perceptive power, its functions of elaboration and invention, at the service of the technical task; the real unity is that of the task rather than that of the tool, but the task cannot be objectivated and can only be lived, experienced, accomplished, and not strictly speaking, reflected upon [réfléchie]. At the second level, technical objects are part of technical ensembles. Consequently, at the first level or at the second level, technical objects cannot be considered as absolute realities and as existing by themselves, even after having been constructed. Their technicity can be understood only through the integration of the activity of a human user or the functioning o f a technical ensemble. It would thus not be legitimate to seek to understand the tech­nicity of an object on the basis of an induction comparable to that which one can apply to natural beings; the technical object, which never harbors all of technicity on its own, either because it is a tool or because it is the element in an ensemble, must be known by philosophical thought, i.e., by a thought that has the intuition of the coming-into-being of the modes of relation between man and the world. (METO, 245)

The Secret Life of Machines

What exactly does it mean to live amongst the machines, to be involved in the creation of technical finality even as we recreate our own human finality along the way?  And how are these finalities related?  On the one hand, Simondon seems to want us to reprise our earlier, artisanal, role as the technical individual.  We are not to let ourselves be reduced to the level of elements or elevated to the level of ensembles.  We are meant to insert ourselves into the workings of the machines, join them together as elements in a concrete new technical individual, perhaps creating some sort of new informational mega-machine.  In this case it appears that we once again machine ourselves.  On the other hand, our new regulation (not control) of the machines seems to occur at a level removed from them.  Simondon sees our interaction with them as a means of discovering or rediscovering the construction of our own finality, but also as a way to appreciate that finality itself is merely one aspect of living things.  In that case, it would seem that the new technical individual would just be one aspect of a larger sense of what it means to be a human individual.  So then, which is it?  Are humans themselves machines, or somehow essentially different?  

In retrospect, I see that this ambiguity runs through the remainder of Part 2 (2.2.2-2.2.4) and is what forced me to read it a couple of times before I felt like I had some grip on his argument.  Here, Simondon attempts to explain in some detail how the new "self-determination" we ended with last time is supposed to work.  It's a fairly complicated story that will take us to the foundations of information theory.  

First, though, the short version -- the new cybernetic technical individual that Simondon hopes to help create is like a virtual version of what was previously the actual craftsman or artisan as technical individual.  It's as if there's another phase shift in the definition of the individual, which moves it from begin an integrator of elements to an integrator of other individuals.  This new role differs from the capitalist's management of ensembles, which assumed that each machine had a distinct and fixed purpose that served his invariant end, an arrangement which effectively turns the old "man of ensembles" into a "machine of machines".  Instead, the new cybernetic individual is investigating what tendencies machines have and asking what they might want to do and what problems they could be made to solve.  Such an investigation explores the individual's self direction at the same time as it virtually integrates the machines.  In a sense, this takes us back to our earlier role as craftsman or tool user, but our tools are now directly aimed at fabricating new human possibilities and not simply some specific material object.

Man as witness to machines is responsible for their relation; the individual machine represents man, but man represents the ensemble of machines, for there is not one machine of all machines, whereas there can be a thought that encompasses all machines. (METO, 157)

The long version is considerably longer and departs from some observations on the difference between human and machine memory that at first might seem irrelevant.  Simondon notes that humans only remember certain things, whereas machines remember everything.  A human who remembered everything would (famously) be pathological.  Likewise, a computer's omnivorous memory is, without any classification system, completely useless (for example, without a functioning file system to interface with humans, my external hard drive is just a sea of varying magnetic charge).  Memory is only useful when it has both context and content.  The computer keeps these two aspects of memory as separate as file name and file data.  Simondon is way ahead of his time in observing that human memory blurs these.  We don't remember what happened shortly after noon on November 22, 1963.  We remember the content only together with the context: we were trading cow lips for alien technology when Kennedy was shot.  

Noting this mixing of context and content in human memory is just the tip of a profound iceberg of reexamining how the living mind operates.  Simondon is ahead of his time here because he's groping towards a predictive processing Bayesian inference model of cognition.  He even seems to think of this as a sort of feedback in time (mirroring the feedback in neural processing that was discovered much later).

Content introduced into human memory will superimpose itself on prior content and take form on it: the living is that in which the a posteriori becomes a priori; memory is the function by which a posteriori matters become a priori. (METO, 138)

Normally of course, we think of memory as the faculty by which the prior (the a priori) comes back as subsequent experience (the a posteriori).  In reversing this direction, Simondon means to suggest that perception is a form of memory, in which our formed experience of what we perceive becomes what we are 'given' (the a priori).  Our experience of the world is not given as raw data, but as a series of models interpreting how this data might fit together based on our past experience.  It's this 'perceiving the past' that makes life possible, that allows us to select what's relevant from the world and act on it without ending up as paralyzed as Funes.  Which means that it's not just the past that gets involved in the present, but the future of possible actions which also get implicated in this passing present.  The error correcting feedback of predictive processing leads us to construct and reconstruct 'what happened' on the basis of what we're going to do about it.  The temporality of all this is interesting to think about but hard to describe.  The present is actually the past (statistical likelihood of certain forms).  But the past is actually the future (likelihood of the forms relevant to our organism).  And now what in God's holy name am I blathering about?  The point here is just that for Simondon, life is inherently virtual.  Living is the art of possibility, where the 'not real' somehow intervenes in the real and opens it up.  Time is the perfect symbol of this fact, and this is why he begins with a discussion of memory.

Machines, by contrast, live only in the present.  No matter how sophisticated we make them, no matter how much of the past they store up or how many simulations of the future they incorporate into their functioning, all of their goals are laid out in advance by their creator.  They simply execute.  Their execution may be enormously complex and impossible to predict.  It may even contain elements of randomness (like Ashby's homeostat) or some other limited ability to reprogram itself.  Try as it might though, the machine cannot ask itself, "what should I do?" in general, but only with reference to the specific task assigned to it.  This general direction is always supplied to the machine from outside, and never generated from within.  While an intelligent machine may solve many subsidiary problems along the way, it is incapable of grappling with the problem of self definition. This is precisely what makes something like Bostrom's paperclip maximizer so scary.  The problem with a completely autonomous machine is that there is no way to interact with it at all.  It just does what it was programmed to do, regardless.  Simondon explicitly likens the situation to Leibniz's monad, whose actions are entirely determined in advance from its internal wiring, so to speak, so that it never communicates with the outside world (Proposition 7: monads have no windows).  If you think about it, you'll see that this makes our fully autonomous and self-perpetuating dream machine kinda useless.  By hypothesis, you can't turn it off, nor can you modify it.  So unless its programmed goal happens to be your own unique goal for all of eternity, you two will inevitably part ways at some point.

At this point I'm sure it's obvious that Simondon has set up this contrast between the machine and the living in order to point out that it's the living, and most relevantly the human, that gives the machine its goal and alters it when need be.  Ultimately though, this is not so much an observation as a definition.  Having a temporality, having access to the virtual, having a capacity for self definition -- these are just what it means to be 'living'.  Likewise, executing operations in the present and lacking a capacity for self redefinition are what it means to be a 'dead' machine.  Seeing these as definitions rather than descriptions goes a long way to clearing up the ambiguity I mentioned at the outset.  

Recall that the problem with capitalism was that it took the finality of the machines for granted.  By contrast, the craftsman was directly involved in the recurrent causality that created the finality of the technical individual he represented.  The point of Simondon's Techno-Marxist history was to show us that there is a 'living' and a 'dead' mode of interacting with our technology.  In the first case we interact with our technology on its own level, by providing the ongoing feedback between different elements and regulating their interaction as an artisan working alongside a material that has its own structure and potentialities, and not according to some goal fixed forever in advance.   An analogy might be the manner that a sculptor interacts with her stone -- beginning with a general plan, but letting the outcome depend on the reaction of the material, thus allowing the combined system of sculptor and stone together to access possibilities that were not apparent at the outset.  In the second, capitalist, mode, we interact with technology either from above or below, forcing it to serve our pre-determined agenda or vice versa.  Here there is no creative feedback loop between ourselves and our technology.  The analogy would no longer be sculpting a recalcitrant material with 'a life of its own', but casting or moulding a pre-determined shape using a material presumed to be perfectly malleable.  It's all just a bunch of dead machines that execute their function, despite the fact that we claim in this instance that (certain) humans are in charge of it all.  This is to say that the role of the same biological individual completely changes depending on the mode of interaction with our technology.  Despite the fact we earlier described the craftsman as a sort of machine and the capitalist as a purely human overseer, it's actually the former that is alive and the latter dead in Simondon's sense.  In short, life is not an inherent property of biological organisms that is forever denied to other causal networks like machines -- it's a mode of interaction between something and its environment.  When we interact with technology by integrating the virtual into its operation, we literally bring it to life.

There is something alive in a technical ensemble, and the integrative function of life can be ensured only by human beings; the human being has the capacity to understand the func­tioning of the machine, on the one hand, and the capacity to live, on the other: one can speak of technical life as being that which actualizes this relation between these two functions in man. (METO, 140)

Hopefully, I've elucidated the root cause of the Simondon's tendency to flip between seeing opposition and seeing continuity between man and machine.  But this really only sets the stage for a more detailed discussion of how the living mode operates.   For instance, at this point you might still feel there's a vague whiff of vitalism about Simondon's notion of technical life.  Just how exactly does the living thing incorporate the virtual into the technical ensemble?  Or for that matter how does the organism, interpreted as a physical entity, incorporate the virtual into itself?  How is the 'not real' made real?  Fortunately, Simondon's discussion of memory has already given us the key to understanding this.  We can actually just restate those idea in information theoretic terms to see how humans could give meaning and life to the machine -- it's all a matter of expectations.

Simondon begins by distinguishing between thermodynamic and informational coupling of machines, and seems to consistently use the term "regulation" for the latter.  Of course, Maxwell's Demon assures us that these two can't be completely divorced, and the example he starts with blurs them.  Consider how a centrifugal governor works.  It is designed to maintain a constant engine speed under a load.   When the engine speed starts to slow down, the rotating balls coupled to it also slow down, and thus drop, which opens the throttle and increases the speed.  In short, the governor couples to the engine to create a more self-regulating engine.  In this case, we could say that the position of the balls provides 'information' which regulates the throttle.  This sounds kinda funny though, since the governor just seems to be part of the machine; for instance, we wouldn't really say that the pressure in the combustion chamber 'regulates' the position of the piston, though it's hard ot see much physical difference between these situations.  Simondon clarifies this confusion when he points out that this governor's regulation is not as perfect as it could be. In this scheme, the engine has to actually start to slow down appreciably before it can speed back up.  In other words, the regulation operates on the same time scale as the machine it regulates.  It would be better (ie. the engine RPM would remain more constant) if the feedback from the regulation happened much faster than the variation in the engine's thermodynamic variables.  At the limit, it would be nice if feedback from internal operating variables were made instantaneous.  Though actually, since instantaneity is a limit, it would be best if the regulator knew what was about to happen to the engine.  For example, if we had a sensor that looked ahead and saw we were about to go up a hill, it could rev the engine just in time to compensate for this.  This would enable us to track our target RPM much closely.  In that case, we would intuitively say that the sensor provides 'information' used to regulate the engine.  This makes sense for two reasons.  First, the time scale of this regulation is much faster than the time scale of relevant fluctuations in the thing we're trying to regulate.  Second, true information actually comes from outside the machine.  It coordinates the machine's potential interactions with its environment.  

So when Simondon talks about the cybernetic individual living amongst the machines and regulating them, he's talking about creating an informational coupling that goes beyond the thermodynamic coupling.  In fact, he points out that the difference between these types leads directly to differences in our technology.  Thermodynamic machines tend to be very large and very isolated from their environment, since this leads to the highest thermodynamic efficiency (defined as energy-out/ energy-in).  On the other hand, informational machines tend to get smaller and smaller, since we'd like the information to be as fast and consume as little energy as possible (a pretty prescient observation to make in 1958, seven years ahead of Moore's Law).  

Ultimately, we dream of making the information instantaneous and completely disembodied.  Alas, this is not possible.  If we send the information down an electrical wire, we run into the limits of the speed of light and quantum fluctuations (aka noise).  Even if we consider the problem in general thermodynamic terms (irrespective of the physical substrate used to transmit the information) we run into Maxwell's Demon.  And Simondon adds a philosophical reason -- information can never be 'pure' because it is parasitic upon the forms that it in-forms.  All of these arguments converge on the same idea -- if we make information too disembodied, we end up confusing it with random variation.  We see the consequences of this everyday.  To transmit information more efficiently in the thermodynamic sense (faster and with less use of energy) we compress it.  And the better we compress it, the more it ends up looking like a random string of numbers because we've rung all the redundancy out of it.  At first, there's something paradoxical about the fact that information, when it is perfected, seems to turn into its exact opposite -- randomness.  But then you realize that this almost follows from its definition as "surprisal".  For data to be informative, you have to have some expectation about what the data could have been.  Information depends on the existence of a set of background possibilities.  Without knowing what might happen, and what we might do about it, data we take in from the environment doesn't constitute information.  Which, as Gregory Bateson famously said, is a difference that makes a difference.  Consider all the data we take in about the speed of air molecules colliding with out skin; it isn't information until we decide the room is too hot or too cold, meanings which are obviously relative to our bodies attempt to maintain an equilibrium form.  Apart from the attempt to maintain this form, we could either see the stream of collision data as utterly meaningless thermal noise, or as a wealth of information about the speed of individual air molecules (if we adopt the demonic perspective).  The point is simply that information is variability with respect to a set of possible forms.    

Information is thus halfway between pure chance and absolute regularity. One can say that form, conceived as absolute spatial as well as temporal regularity, is not information but a condition of information; it is what receives information, the a priori that receives information. Form has a function of selectivity. But informa­tion is not form, nor is it a collection \ensemble] of forms; it is the variability of forms, the influx of variation with respect to a form. It is the unpredictability of a variation of form, not pure unpredictability of all variation. (METO, 150)

Just as information cannot be totally disembodied, the regulation of machines cannot dispense with the living.  Of course we can imagine a thermodynamic coupling of machines into some mega-machine (though Simondon's careful causal analysis would distinguish between whether this larger entity is actually a machine per se, or better described as an ensemble of machines, or even jsut a collection of elements).  And we can even imagine a direct 'informational' coupling of machines where the operations of one machine change the behavior of another.  In this case though, we are simply seeing the goals of life baked directly into the machine ensemble.  Because a machine's functioning can only truly be information for another machine when we have some notion of what the first machine should do.  It's only against this background of human derived expectations that the variations in the operation of a machine can become meaningful for other machines.  What the machine actually does needs to be compared to what it should do.  Humans program these informational evaluations into machines and humans re-program them all the time.  Life (really by definition) is the only thing that can invent and evaluate the new possibilities for what a set of machines could do, which means that it's the only force that can change the inter-machine wiring diagram in order to make it carry out a new goal.  

For Simondon, the informational regulation of machines means incorporating the living, the virtual, into their operation.  And with this piece, we can finally bring the whole scope of his METO project into view.  His goal is for our culture to incorporate the technical knowledge of how machines and humans are actually causally coupled.  We need to create people who understand how the machines are hooked up, how they influence us and how we influence them, what tendencies they have and what goals they can serve -- a sociologist of machines, a mechanologist, a cybernetician.  As I mentioned at the beginning, these would be new "technical individuals" in Simondon's causal definition.  Their evaluation of the possibilities afforded by different configurations of the ensemble of machines makes them a necessary element in machine function, one that can close the causal loop by which a new mega-machine defines its own finality.  Their thought becomes an essential node in the exact type of feedback system that defines a technical individual.  This makes them the virtual, informational, counterpart of the traditional artisan.  

However, the role of this new technical individual is not, as Weiner dreamed, to serve as technocratic philosopher king.  Simondon wants many of these individuals to grow from a general culture.  Ultimately, he'd like to see a technical understanding so ubiquitous that everyone can think about how technology shapes their aims and how they might, in turn, repurpose that technology to new ends.  Which is to say that he's not interested in creating a new technocratic class of cybernetician dedicated to optimizing society according to some fixed, homeostatic, metric.  Instead, he's interested in giving us the knowledge to effectively govern ourselves, that is to define what's important, and change our minds about it.  So finally, the goal of Simondon's new culture is not a better realization of a particular homeostasis, but a realization that all goals are constructed, and that finality is just one mode of approaching life, a technical one.  

A Sort of ... Ugh ... Techno-Marxism

Simondon's ideas are getting more interesting and complicated as the book progresses, so I'm going to have to start breaking it into smaller chunks.  Section 2.2.1 uses the history of technology we outlined in the first two posts as a means of reworking the Marxist notion of class warfare.  A quick elaboration of this history suffices to let us see how much more relevant Simondon's definition of "class" is for our contemporary world.  How do you say Autobahn in French? 

Once upon a time, there was traditional artisanal production.  At this stage of technical development man was the machine -- the human individual coincided with the technical individual.  The craftsmen (technical individual) is the means by which all the tools in the workshop (technical elements) come together in a concrete relation in order to accomplish some task.  

As we move through the Enlightenment and into the era of early modern production, the output of these various technical individuals starts to become diverse enough, yet causally elated enough, to begin constituting a system in its own right (technical ensemble).  We've seen how the Encyclopédie represents the spirit of this transition and gives birth to a new, more abstract, level subjectivity we might call the "man of ensembles".  While conceived by an individual human, this subjectivity lodges itself at the level of the technical ensemble by considering the abstract scientific and functional principles common to all the machines.  What's happening concretely, however, is the emergence of industrial production, which represents a significant improvement in the cost and quality of what each technical individual can produce.  At first, the coming together of this new ensemble of production doesn't change the level at which the machine is made concrete; its still human individuals integrating the effects of distinct tools into the machine (ie. the technical individual).  But because these individuals are now also coupled in an ensemble, the quality and variety of the tools available to them begins to rapidly increase.  Recall that this is how Simondon's dialectic of technical progress functions -- ensembles of individuals produce new elements that are incorporated into new individuals.  Because of the Enlightenment era's technical circumstances, this progress is felt as an increase in power at the level of the individual human.  Man preserves his role as the technical individual at the center of his workshop, but each new and improved technical element is a tool that lets him do the same job faster and more accurately.  In short, the Enlightenment naturally adopts a positive view of technical progress because the corresponding ensemble doesn't change the nature of the concrete technical individual.  We might say that while the Enlightenment invents a totally new abstract subjectivity, this largely remains confined to a bunch of Encyclopedists, while the concrete subjectivity of the average individual doesn't shift that radically.   

By the time of the Romantic reaction, however, the valence of 'progress' has begun to change.  This results from a gap opening up between the machine's notion of progress (as concretization) and the individual's feeling of progress as an increase in their personal power.  As we saw, once the industrial ensemble really gets going, its effect is to create concrete machines that mechanize everything the human individuals used to do -- machine replaces man.  The new elements produced by the industrial ensemble eventually get so precise and standardized that they can be coupled together directly to form a concrete new technical individual (which we commonly refer to as a machine), thus displacing the human individual from their role as tool user.  It's almost as if there's a phase transition to a new sense of 'the individual' (Simondon seems to think of this along the lines of a relaxation oscillator).  Suddenly 'progress' starts to feel pretty scary and disorienting.  There may be progress in general, or progress on average and over longer time periods, but for the average human individual, progress starts to feel threateningly discontinuous and revolutionary.  Progress for the machine can now feel like regress for the human.  It starts to make you understand how the Luddites saw 'progress'.  Or how Stephen King sees it today (or perhaps tomorrow).  

So the progress of technology has pushed the human individual out of their old role as technical individual, and left us with two possibilities.  Either we are reduced to elements in the new machine, made subservient to its finality, or we are reduced to managing the ensemble of machines by treating each of them as an abstract functional black box that should be dominated by our finality.  The Luddites combat the former tendency while the Encyclopedists embrace the latter.  In either case, the human individual is no longer in amongst the machines at the level of a technical individual.  We are no longer an integral link in how the machine comes together as a machine -- a concrete individuated unit defined by the network of recurrent causality.  Human action no longer provides the continuous feedback interactions between elements that concretizes the machine, which results in an alienation of the human individual from the technical individual.  

By now you can probably see how we've managed to retell the story of the class split at the heart of Marxism as a technical, rather than economic or political problem.  Workers become elements and capitalists manage ensembles.  But does reframing capitalism as a 'technical' system help us to say anything new?  I can think of a few advantages to this point of view.  

First, it seems to me more mechanistic and less ideological.  Simondon's technical description class-ifys people on the basis of how they related to concrete, material, machines.  Marx's description is (deliberately) abstracted from this level in order to focus on the political and jurisprudential question of who owns the machines.  But do we really believe that control over the means of production ultimately derives solely from such an arbitrary and purely human institution, and not, at least in part, from our everyday habits of action and thought?  The magic of the Marxist revolution lies in precisely this idealistic conceit that ownership completely determines use.  Some day the workers will wake up and see that the capitalists don't do anything productive.  So they'll redefine who the real owners are with a wave of their abstract legal wand.  Poof, we're all prime ministers!  And suddenly everything will be awesome, even though (by hypothesis) nothing in the concrete mechanistic world of cause and effect is different.  We've seen how these sorts of thought experiments work out.  In fairness to Marx, I'm sure he did imagine that all sorts of concrete things would need to change after the revolution.  But by the same token, we have to admit that Marxism has always been rather vague on just what those changes should be; the assumption has mostly been that they will just work themselves out once the place is under new management.  By contrast, Simondon's scheme implicitly makes these details central because it runs the opposite direction -- from causal control to an abstraction like ownership, and not vice versa.  Who interacts with which machines in what ways in order to accomplish which goals?  

The collectivization of the means of production cannot achieve a reduction of alien­ation on its own; it can only achieve this reduction if it is the precondition for the acquisition of the intelligence of the individuated technical object by the human individual. This relation between the human individual and the technical individual is the most difficult to form. It presupposes a technical culture, which introduces the capacity o f different attitudes rather than that o f work and o f action (work corresponding to the intelligence of the elements and action to the intelli­gence of ensembles). (METO, 134)

How will worker's interactions with the machines that constitute the means of production actually change when they become owners?  While there's clearly a feedback loop between the technical and legal system (which still means that we need, at a minimum, to account for both sides), what we're interested in at the end of the day is changing the daily experience of people in particular ways.  If we focus on changing everyday actions and attitudes, maybe the revolution will follow, instead of lead.

Second, I find that Simondon's scheme is more philosophical than Marx's (or perhaps just less tied to a Hegelian philosophy I find distasteful). This makes it less prone to simplistic moralizing.  For example, Marx constantly speaks of labor's alienation from its production, giving us the impression that it is a simple moral question of capital having stolen something.  The implicit idea is that capital possesses what rightfully belongs to labor.  Simondon, however, explicitly tells us that both labor and capital are equally alienated from their full potential as human individuals.  The problem is that both accept the finality of the machine as already given, though from opposite directions.  By being reduced to the level of the element, the worker has become a literal cog in a machine whose purposes lie beyond them.  On the other hand, as manager of the ensemble of machines, the capitalist comes to see each as a functional black box that is just supposed to spit out money.  While he theoretically controls the finality of the machine, in practice, the capitalist is no more self-determining than the worker.  All machines are useful only insofar as they produce dollar dollar bills y'all.  How they work and what happens inside them is irrelevant to the capitalist; their point is to carry out his function, and it turns out that this function is always the same.  

Alienation does indeed emerge the moment the worker is no longer the owner o f his means of production, but it does not emerge solely because of this rupture in the link of property. It also emerges outside of all collective relation to the means of production, at the physiological and psychological level of the individual properly speaking. The alienation of man in relation to the machine does not only have a socio-economic sense; it also has a physio-psychological sense; the machine no longer prolongs the corporeal schema, neither for workers, nor for those who possess the machines.  Bankers whose social role has been exalted by mathematicians such as the Saint-Simonians and Auguste Comte are as alienated in their relation to the machine as the members of the pro­letariat. (METO, 133)

This strikes me as a much better description of how the actual capitalists I know operate.  They aren't the blood sucking rapacious vampires and moral monsters of Marxist myth.  They aren't even in control in the way we usually use the term to refer to the ability to change directions or decide one's own unique course.  Their situation with respect to the machine is simply too abstract.  They're actually more machine than man, not because they're twisted and evil, but simply because they have only one stereotypical reaction to every situation.  They have become prisoners of their own fixed and limited finality, which alienates them from the continuous self-determination that Simondon considers the hallmark of the human.  

What work and action have in common is the predominance of finality over causality; in both cases, the effort is directed at a certain result to be obtained; the employment of means finds itself in the position of minority with respect to the result: the schema of action matters less than the result of the action. In the technical individual, however, this disequilibrium between causality and finality disappears; viewed from the outside the machine is made in order to obtain a certain result; but, the more the technical being becomes individualized, the more this external finality effaces itself for the benefit o f the internal coherence of functioning; the functioning is finalized with respect to itself before being so in its relation with the external world. Such is the automatism of the machine, and such is its self-regulation: there is, at the level of regulations, a functioning, and not only a causality or finality; in self-regulated functioning, all causality has a sense of finality, and all finality a sense of causality. (METO, 143)

But where did this finality of 'ensemble man' come from?  We take it for granted that everybody wants to become a capitalist, especially in our "ownership society".  That way, you get to do ... whatever you want, right?  Strangely though, all the ownership society seems to want to own is more money.  So how are our actual wants constructed, and how are they so frequently perverted by this abstraction?   This question reflects the third way I find Simondon's technical explanation of class conflict superior.  With Marx, the nature of the socialist utopia seems rather ... conservative.  People continue to want more or less the same stuff they want now, it's just that after the rapture, everyone gets what they want.  By contrast, as METO unfolds, the more radical implications of an innocuous sounding "theory of the technical object" become clearer.  When Simondon talks about integrating technical knowledge into our culture so as to reduce our alienation from machines that are, after all, products of that culture, he is not primarily seeking a pedagodgical reform.  He is really looking to restore the freedom of human self-definition.  The problem with being relegated to either the element or ensemble level is that in the individual ends up having their finality assigned to them.  Our own goals come to seem like forces outside ourselves and take on an almost unquestionable sacred aspect.  For the worker, this finality might come from corporation or nation or society.  Perhaps for us modern globalized capitalists it shows up as an almost solipsistic belief in a purely subjective 'inner' freedom.   Our little man knows what we want because, well ... he's us.  Since the goals are fixed for us from the outside (in this case by an essential self paradoxically within us) we sacralize the efficient pursuit of these allegedly individual goals.  In either case, we lose touch with how finality doesn't come from some mysterious pre-formed unit, regardless of whether we think of it as existing internally or externally.  In short, we lose touch with the fact that all finality -- of both machines and ourselves -- is constructed.  Simondon wants cybernetics to renew our understanding of what it means to be a technical individual that creates its own finality as it functions.   This is not so that we can control the function of the machines from the outside, but so that we can live amongst them, form part of them as they form part of us, in a sort of co-evolution of self-determination.  Demystifying the concept of finality by seeing it as a form of recurrent causality completely changes the meaning of the people's revolution.  Instead of history's crowning moment and a gateway to a new utopia, it becomes a dynamic path of perpetual self reinvention.  Simondon is actually aiming at a new, and much larger, notion of what it means to be an individual.  If we make ourselves large enough we will feel totally free, because we will feel completely determined -- from within.