Also in this section

a144b - 12/7/2004
Bush nie istnieje: George W. Bush jako prezydenckie symulakrum
Carol V. Hamilton (Polish translation by Jakub Maciejczyk)
a142a - 5/18/2004
Networks, Swarms, Multitudes (Part One)
Eugene Thacker


Email this link to a friend HOME : ARTICLES
Printer Friendly Version Articles: a142b
Date Published: 5/18/2004
Arthur and Marilouise Kroker, Editors

Networks, Swarms, Multitudes

Part Two

Eugene Thacker

Are networks, swarms, and multitudes indications of mutations in the body politic? If so, how do they articulate the concepts of connectivity and collectivity, which have been at the heart of political theory for some time? This essay attempts to sort out the conceptual underpinnings of network thinking, and in doing so it brings together technology, biology, and politics into a set of relationships.

While networks may seem ubiquitous these days, we are still far from a thorough understanding of them as political ontologies. Much of this has to do with the way in which time and dynamic change is configured in fields such as graph theory and network science. In these fields, as in much network thinking generally, a network is taken as a static, meta-temporal pattern -- a Eulerian and Kantian view. Bergson's view of duration critiques this notion of dynamic change in networks, but this leaves us with the strange sense that networks do not exist (at least not in their traditional, static forms).

Where can we look, then, for theoretical models which understand the dynamic aspects of networks and other group phenomena? Can the biological paradigm of "swarms" provide one way of complicating the relatively static notion of networks? The concept of a swarm itself has a particular history, and its own set of problematics. The first step will therefore be to consider swarms in the context of biocomplexity studies, molecular biology, and ethology. Swarms can then be considered philosophically in terms of a potential contribution to a political ontology of networks.

Swarms (First Run)

The concept of the swarm is predominantly a biological concept -- its roots are to be found in the study of ethology (animal behavior), and can be traced back to 19th century natural science.[1] Ethology is a curious field, because it is situated between zoology (classification), comparative anatomy (structure), and the study of ecosystems (context).[2] Ethology brings together these areas of study in order to understand the ways in which living organisms interact. Organisms are never just individuals, and never just groups; the "behavior" of an organism is at the intersection of individual, group, and environment. In ethological studies of particular species (such as insects, birds, or predator-prey relationships), the locale of agency is never clear-cut. Rather, it emerges out of the interactions within groups, between individuals, and in response to environmental constraints.

Nowhere is this highlighted more than in the ethological study of "social insects," such as ants, bees and fireflies. The study of social insects provides an example of "dumb" animals forming complicated, "intelligent" social structures, such as the division of labor, hierarchies of control, and coordinated task fulfillment. Entomologist William Morton Wheeler, writing in the late-19th century, referred to social insects as a "superorganism," in which the whole had a vitality which the individual parts did not possess.[3]

Earlier studies assumed a form of centralized control -- the ant queen or queen bee, for instance -- as well as a top-down management of task allocation and social function -- as in army ants or worker bees.[4] Through the ghosts of Leninist-Stalinist communism in such models, a tension arises. On one hand, ethological studies imply that sociality is natural. On the other hand, the type of sociality that develops is rigidly hierarchical, totalitarian, and highly centralized.[5] Studies like these helped to naturalize the social, for it appeared that even insects exhibited highly organized and efficient social behavior. The comparison to human social institutions, manufacturing processes, and industrialization formed the ideological backdrop for such studies. From a political and philosophical view, these early studies of social insects are noteworthy because, in identifying similar social structures across different species -- dumb and intelligent, non-conscious and conscious -- the implication was that sociality and social structure are "natural." The Hobbesian "state of nature" was actually a "State" of nature, in nature.

Although early ethological studies viewed social insects as examples of the universality and naturalness of social behavior, later studies painted a different picture. In particular, the discovery that social insects had no central controller -- no queen bee -- led to a new understanding of the ethology of social insects, focusing more on understanding the systems-wide activities of social insects.[6] Studies of ant foraging, army ant swarming, wasp nest-building, and coordinated flashing among fireflies, are examples of a shift in ethological focus from the discrete actions of individuals, to the collective processes constituted by large groups.[7]

A key lesson of these studies is that global patterns emerge not from a central control, but from aggregates of localized interactions and decisions. In the case of ant foraging (one of the most studied areas), a group of ants sets out in search of a food source. Each ant leaves behind a "pheromone trail," or a chemical path that the other ants can detect. If one ant finds a food source, it will return sooner than the others, thereby reinforcing its specific trail. Subsequent ants will tend to follow the path with the strongest pheromone trail, creating a condition of positive feedback that directs the majority of the ants to the strongest path. In this process, there is no centralized direction of the foraging, and the ants begin their search through random movement. However, over time, the ant colony establishes "highways" to one or more nearby food sources.[8]

Likewise, the study of flocking in birds has been another example of how organization happens without central control. It was the combination of ethology and computer science that has contributed the most to an understanding of flocks.[9] Computer research and simulations often suggest that bird flocks in the real world follow a set of simple rules, which are carried out at a local level between individual birds. These include aggregation (moving towards the flock), steering (avoiding collision with other birds), and rate variability (adjusting speed to match nearest neighbors). These loosely defined "rules" are carried out not at the level of the flock as a whole, but at the level of individual birds and their awareness of their immediate neighbors.[10] Each bird, carrying out these simple rules, produces patterns in the flock as a whole, a pattern not determined by any single bird.

Such studies have challenged conventional notions in the biological sciences concerning the study of "life itself." Increasingly there is a shift from a science of "life" to a technoscience of "living systems." Contemporary research into "swarm intelligence" combines computer science, ethology, and elements of physics (fuzzy logic) to study how swarm particles interact with each other in different contexts ("particle swarm optimization").[11] Nowhere is this merger of biology and computer science more evident than in molecular biotechnology and its related fields (genetics, genomics, proteomics). In biotechnology we have a life science at its most reductive, in the sense that it breaks down life into its smallest component parts and processes. The DNA molecule, which for over 50 years has been the scientific icon of life, represents the culmination of this reductionism -- at once blueprint, dictionary, and central command.

However, recent studies dealing with the molecular level have painted a much more horizontal, distributed picture. Studies dealing with "biological control" systems -- gene expression, cellular metabolism, membrane signaling -- have shown that processes taking place at the molecular level are not linear, cause-effect mechanisms between a master molecule and its subsequent commands.[12] A simple, common process -- such as a cell's intake of sugar molecules, and its conversion of sugar into a usable energy form (ATP) -- requires a network of many interacting units (DNA, amino acids, enzymes, lipids, organelles). The findings of the mapping of the human genome (that the human being only has some 30,000 genes, not much more than mice) has reinforced this view of the organism, at the molecular level, as a system that is much more than the sum of its parts.

Biotechnology research can be broken down into three network topologies. First, the centralized model of DNA as the master molecule, which is exemplified in the famous "central dogma" of DNA formulated by Francis Crick (DNA makes RNA makes proteins, and proteins make us). All instructions and products radiate from the center, which is DNA. However, recent diversifications of genetics and biotechnology, in part brought about by new computer technologies, have enabled more sophisticated forms of analysis (as in the fields of genomics and proteomics). This leads many researchers to adopt a "systems biology" approach, in which DNA, or genes, are simply seen as one component among many that participate in a network of pathways and interactions. The emphasis on genes is still there, but the scope of the research is much broader and more flexible than central-dogma based theories. This second view is more decentralized; it is a cluster of more local, centralized nodes (DNA or genes) that radiates in a local context, thereby interacting with other like clusters (proteins). Finally, the broad body of research known as biocomplexity takes the idea of a biological network further. Contemporary research into gene expression networks, autocatalysis, antibody production, and protein-protein interactions in cells, all point to a distributed model.[13] This is the model that describes swarming behavior at the molecular level. Interactions -- not molecules, compounds, or substances -- are the basis of this model, taking place in a localized context (without external control or a top-down plan). In this way, they create an aggregate of local interactions that produce a global effect. The coordination of protein synthesis, the energy-transformation of molecules in the cell, and the processes by which molecules are selectively brought into and out of cell membranes, are all basic processes that broadly follow swarming behavior.[14]

With this wide range of examples in mind, we can characterize swarms in the following ways:

  • A swarm is an organization of multiple, individuated units with some relation to one another. That is, a swarm is a particular kind of collectivity or group phenomenon that may be dependent upon a condition of connectivity.
  • A swarm is a collectivity that is defined by relationality. This pertains as much to the level of the individual unit as it does to the overall organization of the swarm. Relation is the rule in swarms.
  • A swarm is a dynamic phenomenon (following from its relationality). This differentiates it from the concept of a "network," which has its roots in graph theory and spatial modes of mathematically understanding "things" (or nodes) and "relations" (or edges). A swarm always exists in time and, as such, is always acting, interacting, interrelating, and self-transforming. At some level "living networks" and "swarms" overlap.
  • A swarm is a whole that is more than the sum of its parts, but it is also a heterogeneous whole. This is not to identify a unified, homogeneous group that serves the heterogeneous needs and desires of individuals. Rather, the principles of self-organization require that the group only arises from the localized, singular, heterogeneous actions of individual units.
  • Swarms come with ambiguous politics. The swarm is not the newest term for the concept of the "masses," the "people," or the "proletariat"; the parts are not subservient to the whole -- both exist simultaneously and because of each other. A swarm may exhibit a discernible global pattern, but this does not mean that a swarm prioritizes the group over the individual. Because of this, a swarm does not exist at a local or global level, but at a third level, where multiplicity and relation intersect.

In political terms, swarm intelligence is significantly different from its earlier ethological predecessors. While there are still stratifications within a group (an ant colony, a wasp nest), what is most significant about swarm intelligence studies is the emphasis on collective behavior instead of the static relationship between individual and group. This emphasis on collectivity also means that the global pattern which emerges is not simply one "thing." A flock of birds, a school of fish, or a swarm of insects is not one, homogenous thing, but rather a dynamic and highly differentiated collectivity of interacting agents.

Swarm studies have also been done on human systems, such as the formation and movement of crowds and traffic systems (people, planes, cars).[15] The model of army ant swarming has been a model for the U.S. military for some time, and there are many resonances with the Trotskyist notion of the communist "cell."[16] Most recently the swarming phenomenon has been observed in a new type of protest (distributed dissent, netwars, and smart mobs) that combines physical and technological swarming, movements of bodies (affinity groups) and movements of data (wireless networks).[17]

To summarize this diverse research is to suggest that the study of swarms analyzes the pragmatic dimensions of self-organization, or how an emergent collectivity is able to carry out complicated tasks. Indirectly or directly, swarm studies raise the issue of intentionality and teleology -- how and why does a swarm or a flock form the pattern it forms? If there is no central control, how are tasks created, initiated, allocated? Often, this question is the most difficult to ascertain in swarm research -- researchers often make reference to natural selection or even genetic determinism to account for the question of teleology. But it is the most interesting aspect of swarms, because it bears forth the most basic political tension: in a distributed organization, how is anything accomplished? In the decisively non-human models of swarms, flocks, and so forth, how are the all-too-human characters of desires, intentions, and agencies accommodated?

Swarms (Second Run)

The response to this question by researchers in network science, swarm intelligence, and biocomplexity generally has been less than satisfying. While the self-organizing and complex nature of living systems is understood in great detail, there is inevitably the assumption that these systems are already in progress. How such systems develop, how initiatives within systems are initiated, and how the criteria for self-organization itself "emerges" are questions which existing studies cannot adequately answer.

The issue is not just one of origins or ends. Studies in network science, swarm intelligence, and biocomplexity all define self-organization as the emergence of a global pattern from localized interactions. This paradoxical definition is what makes swarms interesting -- politically, technologically, and biologically -- for it imputes an intentionality-without-intention, an act-without-actor, and a heterogeneous whole. In swarms there is no central command, no unit or agent which is able to survey, oversee and control the entire swarm. Yet the actions of the swarm are directed, the movement motivated, and the pattern has a purpose. This is the paradox of swarms.

In fact, the tension within swarms, as both political and biological entities, is a tension between pattern and purpose. Organization does not necessarily imply a reason for its own existence, unless organization itself is the reason. At one pole is a highly directed, purposeful collectivity, such as a crowd of demonstrators (whose purpose may be to block city streets or obtain visibility), or, on purely biological terms, a swarm of army ants (whose purpose is to look for a food resource). Such collectivities may be called swarms, in that they fulfill two basic components of swarms: they exhibit global patterns from local interactions, and they exhibit a directional force with intention that is without centralized control. At another pole are collectivities which are also highly ordered and dynamically organized, but which do not display any overt "purpose" or goal, other than to maintain themselves as such. Examples may include a large crowd at a festival or concert, or on a biological level, flocks of birds and schools of fish. While researchers interpret such examples as driven by an evolutionary necessity (and therefore dictated by the purpose of survival), the kind of teleology this exhibits is remote, indirect, and ultimately relies on the explanatory capacity of evolutionary theory.

This may simply be splitting hairs; the question of teleology is often secondary in the face of the full presence of a swarm, crowd, or flock. But what is needed in the political domain is precisely a critique of such notions -- what would amount to a critique of self-organization in the political realm. The sciences of biocomplexity that study self-organization (of which swarming is a particular case) want it both ways: on the one hand the appearance of pattern without central control; on the other hand, the use of evolution and natural selection to account for the "rules" and criteria for self-organization. Complex biological phenomena such as swarms therefore appear to be both internally organized and externally constituted, both with full agency and within external constraints.

The case of political swarming is similar, where swarms confront forms of political bunkering-in (e.g. WTO protests in Genoa). More optimistic descriptions of smart mobs and netwars emphasize their self-organized character -- once you press "go." Often, what is not accounted for are the myriad forces and relations that enable such self-organizations to exist in the first place; external to the netwar or the smartmob is the conditions for their existence. Some scientific studies of such phenomena often move towards a sociobiological explanation (and it is no accident that many sociological studies take Edward Wilson's writings on social insects as a reference point). These studies suggest that explicitly non-human forces, such as natural selection or pre-social physical laws, govern what appear on the surface to be ideological concerns.

While in biology it may be adequate to leave the study of swarms at the level of organization and pattern, in the political domain this is only a beginning. Simple identification of patterns does not imply a politics, nor does it show how contingency and potentiality are affected in such contexts. Identifying a crowd of people at a subway station as being organized in a particular way does not say anything about the significance of such a pattern, in this context, at this social and cultural moment. Identifying swarms at a concert, on a crowded city street, and at a large peaceful protest are three qualitatively different phenomena, though from the view of network sciences they may fall into the same category.

There are, then, two axes, two different types of tension, and two sets of concepts. On one axis is the tension between collectivity and connectivity. While connectivity may be a prerequisite for collectivity, collectivity is not necessarily a prerequisite for connectivity as such. Complications arise when a combination of technological euphoria and new social practices lead to an over-optimistic view of connectivity as immediately implying a collectivity. At the extreme point of technological determinism, political forms such as democracy are rendered inherent in both nature and technology.

On the other axis is the tension between pattern and purpose. While pattern may be implicit in the purpose of any swarm, defining the tricky term "purpose" for any self-organizing system is a difficult task. Opinions range as to the degree that self-organizing systems contain a teleology, aim, or goal. In some instances (such as ant foraging), the aim is clear, while in other instances (such as bird flocking), mere existence qualifies teleology. In human groups the issue takes on a more contentious tone, in part because the desires, intentions, and agencies of individual subjects are at stake.

How then, do we begin to sort out the issues raised in these new group formations (networks, swarms, living networks)? How do the two axes of collectivity-connectivity and pattern-purpose play out on the level of political ontology? What is required is a final consideration of "multitudes" (the mode of political aggregation) in addition to the discussion of networks (the mode of technological aggregation) and swarms (the mode of biological aggregation).


The concept of the multitude, which has been given currency recently by the work of political theorists such as Antonio Negri and Michael Hardt, is one that can be found in the emergence of modern political thought. "Multitude" is a term that has been used by Machiavelli, Hobbes, Spinoza, and later political theorists such as Locke and Rousseau. But there is almost no agreement among such thinkers as to what the term signifies. At times the multitude is used as a synonym for what we might call the "masses" or even the "people," while at other times the multitude is given a very specific political charge as the constitutive force of social and political life itself. Pointing to the central tension in the concept of the multitude, Spinoza notes that:

the right of the commonwealth is determined by the power of the multitude, which is led, as it were, by one mind. But this unity of mind can in no way be conceived, unless the commonwealth pursues chiefly the very end, which sound reason teaches is to the interest of all men.[18]

While Hobbes and Machiavelli (in The Prince) come down on the multitude as that which must be forcefully governed, Spinoza and Machiavelli (in The Discourses) create openings for the immanent and essential role of the multitude in making social and political life possible.[19] Indeed, even in a single author's thought -- compare Spinoza's Theological-Political Treatise to the unfinished Political Treatise -- the concept of the multitude changes, from a volatile, unstable, social energy, to a socially constructive and politically radical collective voice.[20] Whatever the case, the multitude always seems to be "at odds" with its conditions; at once constituting and resisting social and political order.

In current political theory, the key issue is a shift in the way the question of the multitude is posed. The question is no longer that of choosing between the priority of the individual and the priority of the group. Such a position asks: is the multitude governable? Instead, the key question now is: how can the multitude self-govern?[21] Understanding the relevance of this shift requires an elaboration of the concept of the multitude itself.

As a political concept, the multitude is defined in terms that are both ontological and political. The term is ontological because, at its root, the multitude articulates a set of affects and experiences which serve as the ground for political affiliation and action. As Hardt and Negri propose, "[o]ntology is not a theory of foundation. It is a theory about the immersion in being and being's continuous construction." On a basic level, then, the multitude is centrally concerned with its own constitution, and from this, with its ability to accommodate both consensus and dissent.

On the ontological plane, the first important observation is that the multitude is neither the individual nor the group. It is positioned somewhere in between, or somewhere else entirely.[22] The multitude is related to the concept of multiplicity, which has been developed by Bergson and, subsequently, Deleuze. The multitude is multiple, in that, like the notion of "the group" and "the people," it implies a significantly large number of individual bodies that are grouped together in some way so as to make them one. In this sense the multitude is singular.

But the multiple is not a multiplicity, and the singular is not a singularity. The multitude is not just a large number that has been unified, homogenized, rendered a body-above-the-body (a body politic); it is also defined by its composition, which is labile, permeable, and morphological. What keeps the multitude from being "one" is that it is defined by a set of diverse interests, affects, and relations. The multitude has no social contract, and is not predicated on a single moment of unity-in-consensus; its diverse interests may not even overlap, except in the most tangential, networked ways.[23]

This means that the multitude constantly confronts the question of what Negri calls "the common:" that set of common interests, concerns, or conditions which enable the multitude to self-organize, if only momentarily, and thus to exercise self-governance.[24] The common is both material, rooted in the ongoing contestation over the production of "life," and (because of this) affective and experiential. But the multitude's self-organization does not automatically imply self-governance.[25] What underlies both is something that Spinoza has pointed to: the fundamental relationality -- or connectivity -- of bodies, affects, and subjects.

Again and again, the multitude is described in an almost paradoxical manner: a multiplicity of singularities, a difference that is repetition, a contraction and an expansion, a diastole and systole.

  • "The multitude is a whole of singularities."[26]
  • "Unlike the people, the multitude is not a unity, but as opposed to the masses and plebs, we can see it as something organized. In fact, it is an agent of self-organization."[27]
  • "These days we refer to the multitude as a group of singularities who have reappropriated the instruments of production, the tools of work for themselves."[28]
  • "[O]ne movement pushes with great force towards absoluteness in the strict sense, toward the unity and indivisibility of government, towards its representation as one soul and one mind...But the other movement is plural...The life of absolute government is endowed in Spinoza with a systole and diastole, with a movement towards unity and a movement of expansion."[29]
  • "In effect, by working, the multitude produces itself as singularity...a singularity that is a reality produced by cooperation..."[30]

It is for this reason that the ontological questions of the multitude are indissociable from the political questions of the multitude. The extensive tradition of social contract theory in modern political thought shows a bias towards juridical forms of authority and legitimation (from Machiavelli to Grotius and Hobbes). At the same time however, there is a counter-thread in contract theory which opens up, if only briefly, the possibility of considering democracy as "the most natural form of government" (from Althusius to Spinoza). It is from this second, minoritarian thread of thinking that the contemporary concept of the multitude gains weight.

As Hardt and Negri point out, the multitude opposes the social contract, and thus implicitly opposes the concept of "the people." Whether the social contract is understood formally as an actual contract, or whether it is taken as a necessary fiction of the state, the contract presupposes the question: is the multitude governable? Such a question inevitably leads to the a posteriori legitimation of modern forms of sovereignty. Hardt and Negri, as is well known, suggest that, while monarchical and imperialistic regimes may be historically waning, this does not at all mean that sovereignty itself has disappeared. The central challenge to rethinking the concept of the multitude is therefore to differentiate it from its ties to modern sovereignty and absolutism.[31]

This process requires a reconsideration of group phenomena or aggregations. It requires a refusal of the fiction of the pre-social "state of nature" as that which needs to be tamed, domesticated, and disciplined. In this way, the multitude should not be confused with its related terms the plebs, the mob, the masses, or even the people. But this does not mean that the multitude becomes an institution. It retains Negri's "savage" element, because of its ontological composition. It is both an aggregation and a differentiation, both an emanation and a contraction; yet, neither a group nor a set of individuals. Etienne Balibar suggests that this ongoing, dynamic tension is in fact what defines the multitude in Spinoza's political writings.[32] Indeed, it was Spinoza who highlighted two important aspects of the multitude: its instability or volatility, and also its capacity to self-organize and form a collective, constitutive power.

Does the multitude exist today, or is it a political vision of something "to-come"? Democracy does not imply multitudes, and in fact, one of the major targets of the multitude is the form of representative democracy (the difference between "constituted" and "constitutive" power that Negri points to). Some arguments suggest that multitudes do in fact exist, in global AIDS activism, in the anti-globalization movements, in the "no border" and digital commons campaigns, and in the notion of an international citizenship.[33] Despite this, it would be a mistake to forget the tensions and contradictions that define the multitude, as well as the ongoing unresolvability of the multitude that Balibar posits.

At this point, we can summarize the contemporary concept of the multitude as follows:

  • The multitude is positioned between the individual and the group; it is a "multiplicity of singularities"
  • The multitude operates through relationality and cooperation, which establishes "the common" or a set of partially-overlapping common affects, issues, and experiences
  • The multitude positions itself against the social contract tradition, and therefore against the inevitability of modern sovereignty and the "state of exception"
  • The central problematic of the multitude is the "problem of the political decision," or how the common can be constituted while fostering difference
  • The question the multitude asks of itself is "can the multitude self-govern?" rather than the question asked of the multitude -- "is the multitude governable?"
  • The most dangerous aspects of the multitude -- its volatility, its unpredictability, and its instability -- are also its most radical aspects -- constitutive power, collective voice, and an immanent ethics.

A Life of the Multitude?

Networks, swarms, and multitudes are not enough in themselves. While they offer alternatives to the tradition of the modern, sovereign body politic, they also contain significant limitations, and raise crucial problems. If networks, swarms, and multitudes are to be viable models for a radical political ontology, then it will be necessary to modify them as concepts.

Networks -- based on a technical Eulerian-Kantian paradigm and continued today in network science -- are unable to adequately account for the components of time, duration, and dynamic change. Networks will, above all, have to be considered as being constituted by temporal dynamics, as real networks existing in and through time -- as living networks. This shift will necessitate a rethinking of the conceptual assumptions inherent in network thinking, such as the separation between "nodes" and "edges," the nature of causality, the measurable quality of networks, and the prioritization of nodes over edges.

The concept of swarms -- based on an ethological or biological paradigm and continued today in swarm intelligence research -- is constrained by its inability to account for the teleology of swarms. Swarm intelligence and biocomplexity research has trouble explaining the constitutive conditions of swarms, or how the criteria for swarming are formed within the swarm. The concept of swarms will have to amplify the role of the Spinozist circulation of affects in swarm behavior, to emphasize the ways in which swarming is always affective swarming. This will encourage a reconsideration of swarms as a constitutive power, as an instance of a conatus that is equal to "resistance."

The concept of multitudes -- grounded in its Spinozist definition and extended in current political theory -- plays an interesting role here, for in many instances (i.e., smart mobs, netwars, electronic civil disobedience) it brings together the technically-based regime of networks and the biologically-based regime of swarms. But if the concept of the multitude layers networks and swarms on top of each other, it also brings with it a set of perhaps unresolvable questions: how can a group be constituted as both a multiplicity and a singularity, how can we materialize a "community without unity?" In such a multiplicity, where heterogeneity, diversity, and difference flourish, how is "political decision" possible, without resort to the tradition of the social contract?

In a sense, multitudes are not separate from networks and swarms; networks and swarms are what inform and transform multitudes. Networks are better suited for some situations, while swarms are better suited for others. The difference lies in what counts as an "edge" or a relation in each paradigm. In networks, an edge (or relation) is often mediated by some facilitating system that is separate from the nodes or units of the network. In the most literal example of computer networks, this would be the hardware elements (fiber, hubs, computers) and the software elements (ports, links, protocols). In computer networks, computers can exist without networks, but the reverse is not necessarily true.

By contrast, swarms are based less on exchanging data through channels, than through the continual modification of action, motion, and movement through the affective signals of local states (state of self, state of nearest neighbors, state of environment). Just any large grouping of people does not constitute a swarm. They may be crowds, masses, or mobs, but, as has been outlined above, a swarm is a particular mode of collective organization. While a single person may certainly exist without a swarm, a swarm is dependent upon a particular kind of constitutive power of individuals. Individuals are individuals of a different sort in swarms, combining localized decision-making and movement, local area consensus-building, and an affective capacity (circulation of affects) linking the individual to the swarm as a whole.

Additionally, in many case studies, networks and swarms co-exist. The "Battle for Seattle" is one case in point. On one level, there were the self-organized affinity groups of the protesters coordinated by Direct Action Network. This level involved a swarming of bodies at particular physical locations (intersections, roads, buildings). Yet, as numerous commentators have noted, this swarming would not have been as successful without the layer of networks that, in part, enabled protesters to coordinate their local movements. This layer was composed of mobile and wireless gadgets, police scanners, and even streaming video. Therefore, in this case there is a combination of swarming bodies with a network of data transmissions. Similar instances of the co-existence between swarms and networks can be seen in the biological domain. In ethology, the study of ant foraging involves not just the corporeal swarming of individual ants, but what enables the swarm to achieve its goal of finding a food source is the informational content of the pheromone trails. The laying of pheromone trails constitutes a network of data exchanges, which both communicates a message ("go this way") as well as enabling the swarm to achieve its overall goal. At an even smaller level, the study of antibody production in the immune system shows that, in addition to a swarming of antibodies and other enzyme co-factors, there is an informational network that exists through the signaling of molecules in relation to each other. The specificity of cell-surface binding receptors is the equivalent to a cipher key: the precise match to the structural surface of a cell that enables a virus to gain entry and infect the host cell.

Despite the similarities between these examples, there are also important and obvious differences. For instance, "information" is defined differently in each case. In the examples of distributed dissent, information is an immaterial, abstract entity packaged in a message (an email, a website, a voice mail message, a text message). Information, in this context, follows the classical communications model set out in information theory -- a message transmitted along a channel that is separate from the message itself. By contrast, the biological example of immunology -- and molecular biology generally -- provides us with a very different notion of information. While it is common to think of DNA as a "code," this is largely a convenient metaphor that belies the role of structural properties of DNA and other molecules in the body. In antibody production, viral infection, cellular metabolism, and a whole set of biological processes, there is no message separate from a channel. Information is wholly material, chemical, and physical. The synthesis of a protein molecule from a set of DNA sequences does not involve the transmission of an informational message over a communications channel. It is a coordinated set of material relations between molecules (DNA, RNA, amino acids, enzymes, lipids). There is no message at the molecular level, only sets of material interactions, the production of antibodies from the immune system, or the metabolizing of substances within the cell. In a sense, there is no information, only deformation and transformation.

All of this brings us back to the two axes of tension in these group phenomena: collectivity/connectivity, and pattern/purpose. On the surface, these two axes are isomorphic with each other. Just as a collectivity is constituted by its purpose, so does connectivity create patterns of links and relations. However, networks and swarms are significantly different kinds of organization. If collectivity is more than number (more than a large number of units), then it is purpose that qualifies collectivity (but a purpose that is internally formulated and initiated). Similarly, if connectivity is more than a random linking or relating, then it is pattern that qualifies connectivity. Networks, then, are those forms of distributed organization that facilitate connectivity (qualified by pattern). Similarly, swarms are those forms of distributed organization that facilitate collectivity (qualified by purpose). This in turn outlines the criteria for both networks and swarms: networks can form a collectivity, through connectivity, while swarms can initiate a connectivity, but only through collectivity.

There are a number of challenges that any political ontology of distributed organization will have to consider. The ability of a group to mobilize itself in a political way (and not just a technological or biological way) will depend on the kinds of responses formed to these challenges.

  • One fundamental challenge will be developing a "tactical" combination of connectivity and collectivity. This tension is at the heart of networks and swarms, and is most pronounced in political examples of multitudes. The tensions between individual experience, political ideology, and practices at both the individual and group levels, are all about discovering the most workable combination of collectivity and connectivity.
  • Another challenge will be the development of new forms of understanding how "control" exists within networks, swarms, and multitudes. A network or swarm is not just an arbitrary pattern, but a directed motion implying a material basis for consensus or the "common." Against the prematurely optimistic views of networks and swarms as examples of an "absence of control," the nature of control itself will have to be re-thought within the particular ontology of distributed behavior. This is perhaps the point at which to think about the differences between control, authority, and coercion, as different types of power dynamics in these group phenomena.
  • The challenge of re-thinking control also means that, in order for networks, swarms, and multitudes to operate in a distributed manner, their mode of control will have to be one that is internal. If there is one thing that the prevalence of networks, swarms, and multitudes demonstrate, it is that we require new understandings of how to conceive of a control that is immanent to and internal to any distributed behavior. This is more than thinking about self-determination for individuals and groups; it is thinking about self-determination as self-organization.
  • Finally, a political ontology of networks, swarms, and multitudes will necessitate a reconsideration of many terms that are central to political thought -- power, right, and democracy. These terms are in turn related to philosophical issues -- individuation, multiplicity, and materiality. This might be as simple as thinking about instances of distributed dissent (politics) in terms of the concepts of living systems (biology), or it might be finding ways to set political concepts (such as "the people") into a dialogue with technoscientific concepts (such as "the pack" or "the network").

But, despite these challenges, there is an equal and more anxious interest in the unpredictability, the instability, and the uncontrollable nature of networks, swarms, and multitudes. They seem to fascinate because their distributed modes of organization foster both an absence of centralized control, and an anxiety about that loss of control. There seems to be a desire to encourage the ubiquity of networks, swarms, and multitudes along with a desire to be able to control or instrumentalize them. We witness this contradiction in the popular media, where, on one hand, a network like the Internet is touted as a leading tool in globalizing democracy, and on the other, fear is fostered within the U.S. against amorphous terrorist networks. An instance of swarming can be, in one instance, a new mode of anti-globalization dissent, and in another instance, a military strategy deployed in the service of U.S. political hegemony. Perhaps it is the purpose of the multitude as a concept to constantly mediate these tensions. Networks, swarms, and multitudes seem to be everywhere in general, and yet nowhere in particular. The political fantasy that is a by-product of these contradiction is that of a "control without control," the best of both worlds: a totally distributed, self-organizing, flexible and robust model of group organization, that can be directed towards certain ends, that is able to mobilize in a particular way, and that is able to express purpose or intention on a global scale.

Thanks are due to Harry and Dot Bowers for providing an environment for research and writing. Thanks are also due to CTheory for their editorial assistance.


[1] In particular, Lorenz's study on animal aggression in predator-prey relationships, and Wheeler's concept of the "superorganism" in his study of ant colonies. See Konrad Lorenz, On Aggression, New York: Routledge, 1990; and William Morton Wheeler, Emergent Evolution and the Development of Societies, New York: Norton, 1928.

[2] For a textbook on ethology, see David McFarland, Animal Behaviour: Psychobiology, Ethology, and Evolution, New York: Longman, 1998.

[3] See Wheeler, Emergent Evolution and the Development of Societies.

[4] Studies such as Edward Wilson and Bert Holldobler's The Ants, Cambridge: Harvard, 1990, display a shift from early studies of insect societies to "hive mind" models.

[5] This tension plays itself out in popular culture, where numerous sci-fi films portray the horrific transformation of human societies into inhuman, insect/communist ones. There is a trajectory of swarm politics which runs from The Birds and Them! to the use of computer modeling in films such as Starship Troopers and Matrix Reloaded, to video games such as SimCity and State of Emergency.

[6] For popular summaries of this research, see chapter 1 of Steven Johnson's, Emergence, New York: Scribner, 2001; and chapter 2 of Kevin Kelly's, Out of Control, New York: Perseus, 1994.

[7] For an excellent overview of current research in this area, see Scott Camazine et al., Self-Organization in Biological Systems, Princeton: Princeton, 2001.

[8] See Eric Bonabeau and Guy Théraulaz, "Swarm Smarts," Scientific American, March 2000: 72-79; and Camazine et al., Self-Organization in Biological Systems, chapters 13 and 14.

[9] See Craig Reynolds, "Flocks, Herds, and Schools: A Distributed Behavioral Model," SIGGRAPH '87 Computer Graphics 21.4, July 1987: 25-34.

[10] The film Winged Migration is an instructive demonstration of these principles.

[11] See James Kennedy et al., Swarm Intelligence.

[12] See Leroy Hood, "The Human Genome Project and the Future of Biology," at, 2001; Sui Huang, "The Practical Problems of Post-Genomic Biology," Nature Biotechnology 18, May 2000: 471-2; Bernhard Palsson, "The Challenges of in silico Biology," Nature Biotechnology 18, November 2000: 1147-50. For a key research article on systems biology, see Troy Ideker et al., "Integrated Genomic and Proteomic Analyses of a Systematically Perturbed Metabolic Network," Science 292, 4 May 2001: 929-934.

[13] See Brian Goodwin and Ricard Solé, Signs of Life: How Complexity Pervades Biology, New York: Basic Books, 2000; and Stuart Kauffman, The Origins of Order: Self-Organization and Selection in Evolution, New York: Oxford, 1993.

[14] Other examples follow this pattern: local interactions, global patterns. Often this process is referred to as "self-organization," and the global pattern is said to "emerge" from the local interactions of its agents. Aside from studies of bird flocks, fish schools, and other group animals, similar studies along these lines have been extended to areas as diverse as computer graphics (seen in many films where a large number of agents must move independently), economics (in the study of multiple transactions in the stock market), and molecular biology (in the analysis of enzymatic and metabolic networks in the cell).

[15] For an overview of the breadth of complexity research, see Manuel Delanda's essay "Non-Organic Life," in Jonathan Crary and Sanford Kwinter, eds., Zone: Incorporations, New York: Zone, 1992, 128-68.

[16] See John Arquilla and David Ronfeldt, Swarming and the Future of Conflict, Santa Monica: RAND, 2000.

[17] While this would risk relativising the concept, numerous other examples of "human swarming" could be included, such as the Anarcho-Syndicalist movement, the Dutch squatter movement, European tactical media labs, international labor unions, Indymedia, global AIDS activism, and so forth.

[18] Spinoza, Political Treatise, Chapter III, Section 7, trans. R.H.M. Elwes.

[19] As Spinoza notes in the Political Treatise, "it is clear that the right of the supreme authorities is nothing else than simple natural right, limited, indeed, by the power, not of every individual, but of the multitude, which is guided, as it were, by one mind -- that is, as each individual in the state of nature, so the body and mind of a dominion have as much right as they have power" (Political Treatise, Chapter III, Section 2; trans. R.H.M. Elwes). Likewise, Machiavelli portrays a more complex, more conflicted multitude in The Discourses than he had in The Prince: "Livy writes these words: 'They [Roman citizens] were fierce when united, but when isolated the fear of each one rendered them obedient.' And, truly, nothing can demonstrate the nature of the masses better, in this respect, than what is shown in this passage. Many times the masses are bold in speaking out against the prince's decision; then, when they see the penalty is at hand, they do not trust one another and run to obey" (The Discourses, trans. Peter Bondanella and Mark Musa, The Portable Machiavelli, New York: Penguin, 1979, p.280/Chapter LVII).

[20] Both Balibar and Negri note the shift in Spinoza's thinking about the multitude. While Balibar concludes that Spinoza was able to articulate the contradictions inherent in democracy, Negri proposes that Spinoza's last writings on the multitude open the way for a new kind of radical collectivism. See Etienne Balibar, Spinoza and Politics, London: Verso, 1998; and Antonio Negri, "Reliqa Desiderantur: A Conjecture for a Definition of the Concept of Democracy in the Final Spinoza," in The New Spinoza, ed. Warren Montag and Ted Stolze, Minneapolis: Univ. of Minnesota, 1997.

[21] This is a recurring question in contemporary political thought influenced by Spinoza. Both Balibar and Negri note the centrality of the problematic of the decision of the multitude. As Balibar notes, "behind the question which arises for any regime -- is the multitude governable? -- lies another, which conditions to varying degrees this first question: to what extent is the multitude capable of governing its own passions?" (Spinoza and Politics, 58). Michael Hardt and Negri, in Empire, (Cambridge: Harvard, 2000) also add that the "multitude is not formed simply by throwing together and mixing nations and peoples indifferently" (395).

[22] "Thus the political problem no longer has two terms but three. 'Individual' and 'State' are in fact abstractions, which only have meaning in relation to one another. In the final analysis, each of them serves merely to express one modality through which the power of the multitude can be realized as such" (Balibar, Spinoza and Politics, 69)

[23] One of the recurrent features of contemporary theories of the multitude is their opposition to social contract traditions. See Michael Hardt and Antonio Negri, "Globalization and Democracy" in Implicating Empire, ed. Stanley Aronowitz and Heather Gautney, New York: Basic, 2002.

[24] See the chapter "Multitudo" in Negri, Time for Revolution, New York: Continuum, 2003.

[25] This is the source of the tension cited by Jean-Luc Nancy, a tension between "community and sovereignty": "The citizen becomes subject at the point where the community gives itself (as) an interiority, and at the point where sovereignty no longer contents itself with residing in the formal autoteleology of a 'contract'..." (The Sense of the World, trans. Jeffrey S. Librett, Minneapolis: Minnesota, 1997, 106).

[26] Negri, "Approximations: Towards an Ontological Definition of the Multitude," Interactivist Info Exchange, 12 November 2002:

[27] Ibid.

[28] Negri, "N for Negri: Antonio Negri in Conversation with Charles Guerra," Grey Room 11, Spring 2003, 99.

[29] Negri, "Reliqa Desiderantur," 229-30.

[30] Hardt and Negri, Empire, 395.

[31] Giorgio Agamben has offered the most extensive commentary on modern sovereignty, particularly as it pertains to the tension between "bare life" and sovereign power, between the political inclusion of the biological "population," and the exclusive status of this biopolitical body: "But this also means that the constitution of the human species in a political body passes through a fundamental division and that in the concept 'people' we can easily recognize the categorical pairs that we have seen to define the original political structure: bare life (people) and political existence (People), exclusion and inclusion, zoe and bios..." (Homo Sacer: Sovereign Power and Bare Life, trans. Daniel Heller-Roazen, Stanford: Stanford, 1998,177).

[32] As Balibar states, "the less sovereignty is physically identified with one fraction of society (in the limiting case, with one individual), the more it will tend to coincide with the people as a whole, and the stronger and more stable it will be. But at the same time, the more difficult it will be to imagine its unity (its unanimity) and its indivisibility (its capacity for decision), and the more complicated it will be to organize them in practice" (Spinoza and Politics, 57-58).

[33] See Hardt and Negri, Empire, 396-400.


Eugene Thacker is assistant professor in the School of Literature, Communication, and Culture at Georgia Institute of Technology. He is the author of Biomedia and the forthcoming The Global Genome: Biotechnology, Politics, and Culture.

© CTheory. All Rights Reserved