"The Limits of Networking"

A Reply to Lovink and Schneider's "Notes on the State of Networking"

Alexander Galloway and Eugene Thacker

The question we aim to explore here is: what is the principle of political
organization or control that stitches a network together? Writers like
Michael Hardt and Antonio Negri have helped answer this question in the
socio-political sphere using the concept of "Empire." Like a network,
Empire is not reducible to any single state power, nor does it follow an
architecture of pyramidal hierarchy. Empire is fluid, flexible, dynamic,
and far-reaching. In that sense, the concept of Empire helps us greatly to
begin thinking about political organization in networks. But like Lovink
and Schneider, we are concerned that no one has yet adequately answered
this question for the technological sphere of bits and atoms.To this end, the principle of political control we suggest is most helpful
for thinking about technological networks is "protocol," a word derived
from computer science but which resonates in the life sciences as well.

Protocol abounds in techno-culture. It is a totalizing control apparatus
that guides both the technical and political formation of computer
networks, biological systems and other media. Put simply, protocols are
all the conventional rules and standards that govern relationships within
networks. Quite often these relationships come in the form of
communication between two or more computers, but "relationships within
networks" can also refer to purely biological processes as in the systemic
phenomenon of gene expression. Thus by "networks" we want to refer to any
system of interrelationality, whether biological or informatic, organic or
inorganic, technical or natural--with the ultimate goal of undoing the
polar restrictiveness of these pairings.

In computer networks, science professionals have, over the years, drafted
hundreds of protocols to govern email, web pages, and so on, plus many
other standards for technologies rarely seen by human eyes. The first
protocols for computer networks were written in 1969 by Steve Crocker and
others. If networks are the structures that connect people, then protocols
are the rules that make sure the connections actually work.

Likewise, molecular biotechnology research frequently makes use of
protocol to configure biological life as a network phenomenon, be it in
gene expression networks, metabolic networks, or the circuitry of cell
signaling pathways. In such instances, the biological and the informatic
become increasingly enmeshed in hybrid systems that are more than
biological: proprietary genome databases, DNA chips for medical
diagnostics, and real-time detection systems for biowarfare agents.
Protocol is twofold; it is both an apparatus that facilitates networks and
also a logic that governs how things are done within that apparatus.

From the large technological discourse of white papers, memos, and
manuals, we can derive some of the basic qualities of the apparatus of
organization which we here call protocol:

+ protocol facilitates relationships between interconnected, but
autonomous, entities;

+ protocol's virtues include robustness, contingency, interoperability,
flexibility, and heterogeneity;

+ a goal of protocol is to accommodate everything, no matter what source
or destination, no matter what originary definition or identity;

+ while protocol is universal, it is always achieved through negotiation
(meaning that in the future protocol can and will be different).

+ protocol is a system for maintaining organization and control in
networks;

We agree wholeheartedly with Lovink and Schneider's observation that
"networks are the emerging form of organization of our time." And we agree
that, due to this emerging form of organization, "networking has lost its
mysterious and subversive character."

Yet they also note that, despite being the site of control and
organization, networks are also the very medium of freedom, if only a
provisional or piecemeal liberation. They write that networking is able
"to free the user from the bonds of locality and identity." And later they
describe networking as "a syncope of power."

In this sense, Lovink and Schneider posit power as the opposite of
networking, as the force that restricts networking and thus restricts
individual freedom:

"Power responds to the pressure of increasing mobility and
communications of the multitudes with attempts to regulate them in
the framework of traditional regimes that cannot be abandoned, but
need to be reconfigured from scratch and recompiled against the
networking paradigm: borders and property, labour and recreation,
education and entertainment industries undergo radical
transformations."

Our point of departure is this: Lovink and Schneider's "Info-Empire"
should not be defined in terms of either corporate or state power, what
they call "the corruption of state sovereignty." Instead it must be
defined at the level of the medium itself. (Otherwise we are no longer
talking about Info-Empire but about the more familiar topics of corporate
greed, fascism, or what have you.) Informatic control is something
different and thus it must be defined differently. It must be defined via
the actual technologies of control that are contained within networks, not
the content carried by those networks, or the intentionality of the people
using them. This position resonates with the "media archaeology" approach
mentioned in Lovink's recent nettime interview with Wolfgang Ernst. This
is why we propose the basic principles of protocol above.

Networks are often seen to be advantageous in political struggles, for
there is presumed to be something about the structure of networks that
enables forms of resistance to take place against more centralized power
structures. The characteristics of multiple sites of locality,
many-to-many communications channels, and a self-organizing capacity
(local actions, global results) are some of the aspects that are cited as
part of the network structure. Indeed, analysis of computer virus attacks,
distributed political protests, and other forms of what John Arquilla and
David Ronfeldt call "netwar" all mention these aspects of networks.

But we find it curious that networks in this characterization are rarely
contextualized--or rendered historical, archaeological. On the one hand,
the centralized structure of "Empire" is assumed to emerge out of a long
history of economically-driven imperialism and colonialism. On the other
hand, the various "networks" which resist Empire seem to suddenly appear
out of nowhere, despite the fact that the technologies which constitute
these networks are themselves rooted in governmental, military, and
commercial developments. We need only remind ourselves of the military
backdrop of WWII mainframe computing and the Cold War context of ARPAnet,
to suggest that networks are not ahistorical entities.

Thus, in many current political discussions, networks are seen as the new
paradigm of social and political organization. The reason is that networks
exhibit a set of properties that distinguishes them from more centralized
power structures. These properties are often taken to be merely abstract,
formal aspects of the network--which is itself characterized as a kind of
meta-structure. We see this in "pop science" books discussing complexity
and network science, as well as in the political discourse of "netwars"
and so forth. What we end up with is a *metaphysics of networks*. The
network, then, appears as a universal signifier of political resistance,
be it in Chiapas, Seattle, Geneva, or online. What we question is not the
network concept itself, for, as a number of network examples show, they
can indeed be effective modes of political struggle. What we do question
is the undue and exclusive reliance on the metaphysics of the network, as
if this ahistorical concept legitimizes itself merely by existing.

An engaged, political understanding of networks will not only pay
attention to networks generally, but to networks specifically. If there
are no networks in general, then there are also no general networks.
(Marx: "If there is no production in general, then there is no general
production.") Networks can be engaged with at the general level, but they
always need to be qualified--and we mean this in technical as well as
socio-political terms. The discourse surrounding "Empire" has been very
good at contextualizing globalization; it has not done so well at
contextualizing "the movement," "the multitude," or "networks" (which are
arguably, three different concepts).

Biological or computational, the network is always configured by its
protocols. We stress this integrative approach because we cannot afford to
view "information" naively as solely immaterial. Negri notes that "all
politics is biopolitics," and to this, we would add that all networks are
not only biopolitical but biotechnical networks. Protocological control in
networks is as much about networks as *living networks* as it is about the
materiality of informatics.

Thus we are quite interested in a understanding of political change within
networks. What follows might be thought of as a series of challenges for
"counterprotocological practice," designed for anyone wishing progressive
change inside of biotechnical networks.

First, oppositional practices will have to focus not on a static map of
one-to-one relationships, but a dynamic diagram of many-to-many
relationships. This is a nearly insurmountable task. These practices will
have to attend to many-to-many relationships without making the dangerous
mistake of thinking that many-to-many means total or universal. There will
be no universals for life. This means that the counterprotocols of current
networks will be pliant and vigorous where existing protocols are flexible
and robust. They will attend to the tensions and contradictions within
such systems, such as the contradiction between rigid control implicit in
network protocols and the liberal ideologies that underpin them.
Counterprotocological practice will not avoid downtime. It will restart
often.

The second point is about tactics. In reality, counterprotocological
practice is not "counter" anything! Saying that politics is an act of
"resistance" was never true, except for the most literal interpretation of
conservatism. We must search-and-replace all occurrences of "resistance"
with "impulsion" or perhaps "thrust." Thus the concept of resistance in
politics should be superceded by the concept of hypertrophy. Resistance is
a Clausewitzian mentality; the strategy of maneuvers teaches us instead
that the best way to beat an enemy is to become a better enemy. One must
push through to the other side, rather than drag one's heels. There are
two directions for political change: resistance implies a desire for
stasis or retrograde motion, but hypertrophy is the desire for pushing
beyond. The goal is not to destroy technology in some neoluddite delusion,
but to push technology into a hypertrophic state, further than it is meant
to go. We must scale up, not unplug. Then, during the passage of
technology into this injured, engorged, and unguarded condition, it will
be sculpted anew into something better, something in closer agreement with
the real wants and desires of its users.

The third point has to do with structure. Because networks are
(technically) predicated on creating possible communications between
nodes, oppositional practices will have to focus less on the
characteristics of the nodes, and more on the quality of the interactions
between nodes. In this sense the node-edge distinction will break down.
Nodes will be constructed as a byproduct of the creation of edges, and
edges will be a precondition for the inclusion of nodes in the network.
Conveyances are key. From the oppositional perspective, nodes are nothing
but dilated or relaxed edges, while edges are constricted, hyper-kinetic
nodes. Nodes may be composed of clustering edges, while edges may be
extended nodes.

Using various protocols as their operational standards, networks tend to
combine large masses of different elements under a single umbrella. The
fourth point we offer, then, deals with motion: counterprotocol practices
can capitalize on the homogeneity found in networks to resonate far and
wide with little effort. Again, the point is not to do away with standards
or the process of standardization altogether, for there is no imaginary
zone of non-standardization, no zero-place where there is a ghostly, pure
flow of only edges. Protocological control works through inherent
tensions, and as such, counterprotocol practices can be understood as
tactical implementations and intensifications of protocological control.