What are the classic texts for constituting a critical theory for the twenty-first century? How would one go about choosing them? Perhaps it would be a matter of making a fresh selection from the past, even the recent past, according to the agenda for thought and action than confronts us today.

That agenda seems to me to have at least three major features. The first is the anthropocene. One can no longer bracket off nature from the social, and construct a theory exclusively on the terrain of the social. The second is the role of information in both production and reproduction. One can no longer just assume that either capital marches on as in the age of the steam engine, or that the superstructures too are not affected by such vulgar questions. The third would be a shift away from Eurocentric concerns. World history appears to be made elsewhere now.

If the first of these agenda items steered us towards Paul Burkett’s work, the second points towards some very prescient attempts to think the age of information. I want to start with Tiziana Terranova’s Network Culture: Politics for the Information Age (Pluto Press, London, 2004), which brings together the resources of the Italian autonomist thinkers with Deleuze and Guattari. It is a position I used to be close to myself but have move away from. So this appreciation of Terranova’s classic text is also in some respects an inquiry into both what reading Marx with Deleuze enabled but also where the limits of such a conjunction might lie.

Terranova starts by taking a certain distance from the ‘postmodern’ habits of thought characteristic of late twentieth century writing. This was a time when texts like Paul Virilio’s Information Bomb were popular, and encouraged a certain delirious end-of-the-world-as-we-know-it talk. If the industrial era tech of mechanically reproducible media were on the way out, then everything was supposedly becoming unreal and out of control. Welcome to what Jean-François Lyotard called the ‘immaterial’.

And that was the more respectable end of that talk. Neo-gnostic sects such as the extropians really through they could leave their bodies behind and upload consciousness into computers. The cyberpunks wanted to ‘jack-in’ and at least temporarily leave ‘meatspace’. Honestly, people really did lose their shit over the beginnings of the end of mechanical and broadcast media, as tends to happen in all such transitions in media form, as historians of media culture well know.

Contrary to certain now popular narratives by latecomers, not everybody went gaga over ‘new media’ in that period, which stretches perhaps from the popularization of cyberpunk in 1984 to the death of the internet as a purely scientific and military media in 1995. There were plenty of experimental, critical and constructivist minds at work on it. I would count Terranova as a fine exponent of the constructivist approach, a sober builder of useful theory that might open spaces for new practices in the emerging world of post-broadcast media flux.

Terranova: “I do not believe that such information dynamics simply expresses the coming hegemony of the ‘immaterial’ over the material. On the contrary, I believe that if there is an acceleration of history and an annihilation of distances within an information milieu, it is a creative destruction, that is a productive movement that releases (rather than simply inhibits) social potentials for transformation.” (2-3) It became a question, then, of a level-headed analysis of the tendencies at work.

It helps not to make a fetish of just one aspect of media form, whether one is talking about the ‘internet’ back then, or ‘big data’ now. Sometimes these are aspects of more pervasive technological phyla. Terranova: “Here I take the internet to be not simply a specific medium but a kind of active implementation of a design technique able to deal with the openness of systems.” (3) This might be a useful interpretive key for thinking how certain now-dominant approaches to tech arose. Her approach is not limited to tech, however, but studies concepts and milieu as well as techniques.

Lyotard sent everyone off on a bum steer with the idea of the immateriality of information, a problem compounded by Jameson’s famous assertion that the technics of late capitalism could not be directly represented, as if the physics of heat engines was somehow clearer in people’s heads than the physics of electrical conductivity. Terranova usefully begins again with a concrete image of information as something that happens in material systems, and thinks them through the image of a space of fluid motion rather than just as an end-to-end line from sender to receiver.

Anybody who studied communication late last century would have encountered some version of the sender -> code-> channel-> receiver model, with its mysterious vestigial term of ‘context’. Stuart Hall complicated this by adding a possible difference between the encoding and the decoding, thus making the non-identity of the message at either end a function not of noise as something negative but of culture as a positive system of differences. But even so, this way of thinking tended to make a fetish of the single, unilinear act of communication. It ended up as an endless argument over whether the sender’s power dominated the receiver’s, or if the receiver had an independent power of interpretation. That was what the difference between the Frankfurt school’s followers and the Birmingham school of Hall et al boiled down to.

Terranova usefully brackets off the whole language of domination and hegemony, moving the discussion away from the privileged questions of meaning and representation that the humanities-trained love so much. She insists we really take seriously the breakthrough of Claude Shannon’s purely mathematical theory of information of 1948.

Information actually means three different things in Shannon. It is (1) a ratio of signal to noise, (2) a statistical measure of uncertainty, and (3) a non-deterministic theory of causation. He developed his theory in close contact with engineers working on communication problems in telephony at Bell Labs, one of the key sites where our twenty-first century world was made.

The signal to noise problem arose out of attempts to amplify telephony signals for long distance calls, where the additional energy used to amplify the signals also shows up as additional noise. This, incidentally, is where one sees how the experience of information as ‘immaterial’ is actually an effect produced by decades of difficult engineering. It takes energy to make a signal pass through a copper wire, making the electrons dance in their predictable but non-deterministic way. The energy leaks into the signal as a source of noise.

What was crucial about Shannon’s approach to this problem was to separate out the concept of information from having anything to do with ‘meaning’. Information is not ‘text’ or ‘language’. It is just a ratio of novelty and redundancy. “From an informational perspective, communication is neither a rational argument nor an antagonistic experience…” (15) It has nothing to with communication as domination or resistance, and it has nothing to do either with Habermasian communicative action or Lyotard’s language games. Before one even gets to any such hermeneutic theories, one has to deal with the specific materiality of information, which was in effect Shannon’s unique contribution.

For information to be transmitted, it has to confront the demon of noise. In this approach sender and receiver appear as nodes as cooperating against noise rather than as dialectical opposites. But Terranova does not adopt Shannon without modification. Rather she follows Gilbert Simondon’s critique of information theory. Simondon points out that in Shannon, the individual sender and receiver are pre-constituted. They just appear as such, prior to the act of communication. Simondon’s approach picks up the vestigial concept of ‘context’. For him, the act of communication is also what constitutes the sender and receiver as such. His approach is to think the context as the site where information produces individuations out of a collective, undifferentiated context.

For Terranova, this is a step toward thinking the space of information as a more turbulent, metastable systems that can be disturbed by very small events: “the informational dimension of communication seems to imply an unfolding process of material constitution that neither the liberal ethics of journalism nor the cynicism of public relations officers really address.” (19) This touches on the problem of causation. Neither the liberal-rationalist nor the cynical-manipulable approach to communication really holds up to much scrutiny. Which reminds me of the advertising guru David Ogilvy, quoting one of his clients: “I know only half the advertising I pay for works, but I just don’t know which half.”

Terranova points the way to thinking information in a broader context to struggles over meaning. It could lead to problems in the organization of perception and the construction of bodily habits. It could be a way of framing a problem of the information redesign of the whole field of media and culture. Information design could be about more than messages defeating noise, but rather designing fields of possibility.

The grand obsession of the first wave of information researchers and engineers was the elimination not only of noise, but of ambiguity. As Paul Edwards has shown in The Closed World (1988), even when they did not actually work any better, closed, digital systems took preference over analog ones, particularly in military-funded projects.

For Terrannova, what might be a constructive project in the wake of that is some kind of information culture that does not enforce a cut in advance in the fabric of the world, and then reduce its manipulation to a set of predictable and calculable alternatives. “Informational cultures challenge the coincidence of the real with the possible.” (20) Information systems reduce material processed to closed systems defined by the relation between selection (the actual) and the field of possibilities (the virtual), but where that field appears in an impoverished form. “Information thus operates as a form of probabilistic containment and resolution of the instability, uncertainty and virtuality of a process.” (24)

Interestingly, Terranova’s approach is less about information as a way of producing copies, and more about the reduction of events to probabilities, thus sidestepping the language of simulation, although perhaps also neglecting somewhat the question of how information challenged regimes of private property. The emphasis is much more on information as a form of control for managing and reproducing closed systems. This then appears as a closure of the horizon of radical transformation. Instead of future societies we have futures markets. “A cultural politics of information thus also implies a renewed and intense struggle around the definition of the limits and alternatives that identify the potential for change and transformation.” (25)

This would be a cultural politics of the probable, the possible and real. “What lies beyond the possible and the real is thus the openness of the virtual, of the invention and the fluctuation, of what cannot be planned or even thought in advance, of what has no real permanence but only reverberations… The cultural politics of information involves a stab at the fabric of possibility.” (27) It does not arise out of negation, out of a confrontation with techno-power as an other. It is rather a positive feedback effect.

There was a time when I would have shared some of this language and this project. But I think what became of an engagement with Deleuze was in the end an extension of a certain kind of romanticism, even if it is one that finds its magical other domain now immanent to the mundane world of things. Deleuze was enabling at the time, with his constructivist approach to building concepts that work across different domains. But he was a bit too quick to impose a philosophical authority on top of those other domains, as if everything was grist to the mill of a still-universalizing discourse of concept-making. It fell short of a genuine epistemological pluralism, where those other domains of knowledge and practice could push back and insist on their own protocols. Nowhere is this clearer than in Donna Haraway’s demolition job in When Species Meet (2007) of Deleuze and Guattari’s metafictional writing about wolf packs. Sometimes one has to cede the ground to people who actually know something about wolves. Perhaps I’ll pick this up again in a separate post on Terrannova’s very interesting chapter on biological computing.

One of the more powerful features of the theory of information in Shannon and since is the way it linked together information and entropy. Thermodynamics, that key to the scientific worldview in Marx’s era, offered the breakthrough of an irreversible concept of time, and one which appeared as a powerful metaphor for the era of the combustion engine. In short: heat leaks, energy dissipates. Any system based on a heat differential eventually ‘runs out of stram.’

Hence the figure of Maxwell’s Demon, which could magically sort the hot particles out from the cool ones, and prevent an energy system from entropic decline into disorder. But that, in a sense, is exactly what information systems do. The tendency of things might still be entropic: systems dissipate and break down. But there might still be neg-entropic counter-systems that can sort and order and organize. Such might be an information system. Such might also, as Joseph Needham among many others started to think, might be what is distinctive about living systems.

Needham’s organicism borrowed from the systems-theory of Bertalanffy which pre-dates Shannon, and was based a lot more on analog thinking, particularly the powerful image of the organizing field. Much more influential was the transposition of the thought-image of the digital to the question of how life is organized as neg-entropic system, resulting in what for Haraway in Modest_Witness (1997) is a kind of code fetishism. What is appealing to Terranova in the confluence of biological and information thinking is the way it bypassed the humanistic subject, and thought instead toward populations at macro and micro scales.

But in some ways Terranova is not that far from Haraway, even though Haraway makes almost no appearance in this text. Where they intersect is in the project of understanding how scientific knowledge is both real knowledge and shot through with ideological residues at the same time: “An engagement with the technical and scientific genealogy of a concept such as information… can be actively critical without dis-acknowledging its power to give expression and visibility to social and physical processes… Information is neither simply a physical domain nor a social construction, nor the content of a communication act, nor an immaterial entity set to take over the real, but a specific reorientation of forms of power and modes of resistance.” (37) While I would want to pause over the word ‘resistance’, this seems to me a usefully nuanced approach.

I am a bit more skeptical these days about the will to impute a domain of otherness as a sort of immanent plane. Thus, while Terranova acknowledges the power of Manuel Castell’s figure of the network as a space of flows coming to dominate a space of places, she wants to retain a strong sense of radical possibility. One way she does so is by appealing to Bergson’s distinction between a quantified and a qualified sense of time, where time as quality, as duration, retains primacy, offering the promise of a “virtuality of duration.” (51)

But is this not yet another offshoot of romanticism? And what if it was really quite the other way around? What if the figure of time as quality actually depended on measurable, quantitative time? I’m thinking here of Peter Gallison’s demonstration of how the engineering feat of electrically synchronized time, so useful to the railways, enabled Einstein to question the metaphysical time and space that was the backdrop to Newton’s mechanics. As Gallison shows, it is only after you can actually distribute a measure of clock time pretty accurately between distant locations that you can even think about how time might be relative to mass and motion.

It is certainly useful that Terranova offers a language within which to think a more elastic relation between the information in a network and the topology of that network itself. It isn’t always the case that, as with Shannon’s sender and receiver, that the nodes are fixed and pre-constituted. “A piece of information spreading throughout the open space of the network is not only a vector in search of a target, it is also a potential transformation of the space crossed that always leaves something behind.” (51)

This more elastic space, incidentally, is how I had proposed thinking the category of vector in Virtual Geography (1995). In geometry a vector is a line of fixed length but of no fixed position. Thus one could think it as a channel that has certain affordances, but which could actually be deployed not only to connect different nodes, but sometimes to even call those nodes into being. Hence I thought vector as part of a vector-field, which might have a certain malleable geometry, but where what might matter is not some elusive ‘virtual’ dimension, but the tactics and experiments of finding what it actually affords.

Terranova stresses the way the internet functions as an open system, with distributed command functions. It was in this sense not quite the same as the attempts to build closed systems of an early generation of communication engineers: “resilience needs decentralization; decentralization brings localization and autonomy; localization and autonomy produce differentiation and divergence.” (57) The network, like empire, is tolerant of differences, and inclusive up to a point – but also expansionist. As Terranova notes, rather presciently, “There is nothing to stop every object from being given an internet address that makes it locatable in electronic space.” (62)

In short, the internet starts to acquire the properties of a fully-realized vector-field: “Unlike telegraphy and telephony… the communication of information in computer networks does not start with a sender, a receiver and a line, but with an overall information space, constituted by a tangle of possible directions and routes, where information propagates by autonomously finding the lines of least resistance.” (65) “In a packet-switched network… there is no simple vector or route between A… and B…” (67) But I think its still helpful to think of it as a vector-field, in that each of those routes still has fairly fixed affordances.

Terranova was a pioneer in understanding that the build-out of an apparatus of which information theory was the concept had significant implications for rethinking the work of culture and politics. “There is no cultural experimentation with aesthetic forms or political organization, no building of alliances or elaboration of tactics that does not have to confront the turbulence of electronic space. The politics of network culture are thus not only about competing viewpoints, anarchistic self-regulation and barriers to access, but also about the pragmatic production of viable topological formations able to persist within an open and fluid milieu.” (68)

She notes in passing some of the experiments of the late twentieth century in “network hydrodynamics” (69) such as the Communitree BBS, Andreas Broeckmann’s Syndicate list-serv, Amsterdam’s Digital City, and the rhizome list-serv. All of these fell apart one way or another, even if many others lived on and even mutated. Much of the functionality of today’s social media derives from these early experiments. Geert Lovink has devoted several books now to documenting what is living and what is dead in the experimental culture and politics of networks.

Terranova was also prescient in asking questions about the ‘free labor’ that was just starting to become a visible feature of network cultures at the time she was writing. She reads this through the autonomist-Marxist figure of the shift of work processes from the factory to society, or ‘the social factory.’ I sometimes wonder if this image might be a bit too limiting. A lot of free labor in the ‘nets looks more like the social office, or even like a social boudoir. Rather than the figure of the social as factory, it might be more helpful to think of a dismantling and repartitioning of all institutionalized divisions of labor under the impact of networked communication.

Still, it was useful at the time to insist on the category of labor, at a time when it was tending towards invisibility. One has to remember that ten years ago there was a lot more celebration of the ‘playful’ contributions of things like fan cultures to the net. Henry Jenkins’ repurposing of something like the Birmingham school’s insistence on popular agency would be a signal instance of this. Terranova: “The internet does not automatically turn every user into an active producer, and every worker into a creative subject.” (75)

In a 1998 nettime.org post, Richard Barbrook suggested that the internet of that era had become the site for a kind of post-situationist practice of détournement, of which nettime itself might not have been a bad example. Before anybody had figured out how to really commodify the internet, it was a space for a “high tech gift economy.” Terranova thinks Barbrook put too much emphasis on the difference between this high tech gift economy and old fashioned capitalism. But perhaps it might be helpful to ask whether, at its commanding heights, this still is old fashioned capitalism, or whether the ruling class itself may not have mutated.

Certainly, the internet became a vector along which the desires that were not recognizable under old-style capitalism chose to flee. Terranova: “Is the end of Marxist alienation wished for by the management gurus the same thing as the gift economy heralded by leftist discourse?” (79) Not so much. Those desires were recaptured again. I don’t know who exactly is supposed to have fallen for “naïve technological utopianism” (80) back in the 90s, apart from the extropians and their fellow travellers. In the main I think a kind of radical pragmatism of the kind advocated by Geert Lovink reigned, in practice at least. We were on the internet to do with it what what we wanted, what we could, for as long as it lasted, or as long as we could make it last, before somebody shut the party down.

For a long time now there’s been a tension over how to regard what the internet has done to labor. Even in the 90s, it was not uncommon to see attacks on the elitism and rabid libertarianism of hacker culture – as if there weren’t complexities and internal divisions within that culture. Such a view renders certain less glamorous kinds of new work less visible, and also shuts down thinking about other kinds of agency that new kinds of labor might give rise to. Terranova: “it matters whether these are seen as the owners of elitist cultural and economic power or the avant-garde of new configurations of labor which do not automatically guarantee elite status.” (81)

Terranova’s Network Culture provided an early introduction in the Anglopohone world to the work of Mauritzio Lazzarato, but I always thought that his category of immaterial labor was less than helpful. Since I agree with Terranova’s earlier dismissal of the notion of information as ‘immaterial’, I am surprised to see her reintroduce the term to refer to labor, which if anything is even more clearly embedded in material systems.

For Lazzarato and Terranova, immaterial labor refers to two aspects of labor: the rise of the information content of the commodity, and the activity that produces its affective and cultural content. Terranova: “immaterial labor involves a series of activities that are not normally recognized as ‘work’ – in other words, the kinds of activities involved in defining and fixing cultural and artistic standards, fashions tastes, consumer norms, and more strategically, public opinion.” (82) It is the form of activity of “every productive subject within postindustrial societies.” (83)

Knowledge is inherently collaborative, hence there are tensions in immaterial labor (but other kinds of labor are collaborative too). “The internet highlights the existence of networks of immaterial labor and speeds up their accretion into a collective entity.” (84) An observation that would prove to be quite prescient. Immaterial labor includes activities that fall outside the concept of abstract labor, meaning time used for the production of exchange value, or socially necessary labor time. Immaterial labor imbues the production process with desire. “Capital wants to retain control over the unfolding of these vitualities.” (84)

Terranova follows those autonomist Marxists who have been interested in the mutations of labor after the classic factory form, and like them her central text is Marx’s ‘Fragment on Machines’ from the Grundrisse. The autonomists base themselves on the idea that the general intellect, or ensemble of knowledge constitutes the center of social production, but with some modification. “They claim that Marx completely identified the general intellect (or knowledge as the principle productive force) with fixed capital (the machine) and thus neglected to account for the fact that the general intellect cannot exist independently of the concrete subjects who mediate the articulation of the machines with each other.” (87) For the autonomists, living labor is always the determining factor, here recast as a mass intellectuality. (See here for a different reading of the ‘Fragment on Machines’)

The autonomists think that taking the labor point of view means to think labor as subjectivity. Living labor alone acts as a kind of vitalist essence, of vast and virtual capacities, against which capital is always a reactive and recuperative force. This is in contrast to what the labor point of view meant, for example, to Bogdanov, which is that labor’s task is not just to think its collective self-interest, but to think about how to acquire the means to manage the whole of the social and natural world, but using the forms of organizing specific to it as a class.

From that point of view, it might be instructive to look to the internet for baby steps in self-organization, or at what Terranova calls free labor, and of how it was exploited in quite novel ways. “Free labor is a desire of labor immanent to late capitalism, and late capitalism is the field which both sustains free labor and exhausts it. It exhausts it by undermining the means through which that labor can sustain itself: from the burn-out syndromes of internet start-ups to under-compensation and exploitation in the cultural economy at large.” (94)

Here I think it is helpful not to just assume that this is the same ‘capitalism’ as in Marx’s era. The internet was the most public aspect of a whole modification of the forces of production, which enabled users to break with private property in information, to start creating both new code and new culture outside such constraints. But those forces of production drove not just popular strategies from below, but also enabled the formation of a new kind of ruling class from above. One based on extracting not so much surplus labor as surplus information.