Emergence on the Web: Cyberspace and the Science of Complexity

by: Selena Sol
This paper discusses the emergent characteristics of the Web; that is, the natural, "more-than" summation of human interaction with information/telecommunications technology. As such, this paper begins by introducing the reader to both the Web and a few principle qualities of emergent systems that it manifests. Finally, it concludes by suggesting policy implications given the nature of the Web as a complex system, the nature of complex systems generically, and the Web's central role in the emerging twenty-first century society.


"...the societies that master the new science of complexity and can convert that knowledge into new products and forms of social organization will become the cultural, economic, and military superpowers of the next century." (1)
The Internet, or as it is quickly being re-metaphored, The Web, is a complex "network of networks" connecting a multitude of independent networks and individual computers globally via basic dial-up phone lines, high-speed dedicated leased lines, microwave links, satellites, and fiber optic links. (2) That is, if all the computers linked together in your office or school are considered a network, then all the networks of all the schools and all the offices in every country, linked together, is the Web.

Granted, this is a pretty vague description. However, this network of networks is far from being a quantifiable, bordered, describable thing. As LaQuey writes, "trying to describe [the Web] and how it fits together is a bit like trying to count the stars." In fact, she continues, given the number and diversity of networks, sub-networks, and sub-sub networks, it is, "impossible to show an accurate, up-to-date picture." (3)

The fact is that the Web is a jumbled hodgepodge of diverse actors using diverse hardware and software tools for a diverse set of reasons. Yet, though no centralized standards-setting body controls or defines this virtual reality, friendships are forged, products are sold, and study is conducted. By its very name, the Web, is related to ideas of interactive complexity as discussed by such theorists as Dosi, Prigogine, and the host of Santa Fe Instituters from Arthur to Kauffman. By its behavior and characteristics, the Web can be identified as both child and parent of complex adaptive systems.

Specifically, the Web provides the necessary infrastructure on which to build complex adaptive systems by creating a highly inter-related, open, decentralized architecture for time and distance transcendent communication. The Web connects. It connects like no technology has before connected. And when "people" are incorporated into this system, the resultant cybernetic system fosters a variety of emergent phenomenon from cyberbusiness to cybergovernment to cyberculture; phenomena far greater than simply the summation of a bunch of linked computers and their users tapping away at their keyboards.

Business on the Web, for example, is not simply a more efficient means of mass advertising as early critics forecasted. Rather, The Web becomes an essential tool fostering the emergence of a wholly unique form of customized, highly consumer feedback-oriented transactions in which mass production becomes mass customization in the hyper-service sector.

http://www.oracle.com/ links one to Oracle Corporation's home on the Information Superhighway. This Web page is a "one stop source for the latest information about Oracle products and services, announcements, alliances, and business partners," and provides such services as "test driving the latest products," customer support including customer specific software modifications, and customer/company feedback and discussion tools for over 100 countries worldwide (currently). Surely advertising still exists. In fact, advertising is an important component of any lean corporation; but only as an evolving characteristic in a greater emergent phenomenon. Oracle and others do not use the Web simply to pump out information more efficiently, they depend on the Web to rapidly and intimately meet specific customer needs anywhere in the world in a two way exchange.

Pick up any number of pop cyberanthropologies and you'll find ample examples of similar unexpected and emergent systems from politics to education. Do a keyword search for "paradigm transformation" and your search screen will be scrolling for days. What is important in all this is that we attempt to understand the Web regardless of its daunting level of complexity and its tendency to foster surprises. For the Web promises to be as central to the Information Revolution as the factory was for the Industrial Revolution and the farm was for the Agricultural Revolution. In the Information Age, the Web becomes not only medium of communication, but mode of production and center of community as well. In a McLuhanian sense, where the wheel is an extension of the foot, and the television an extension of the eye, the Web becomes extension of the self and we find ourselves becoming cyborgs; not the "Borg" of Star Trek, or The Terminator, but meta-humans, post-hominids, homo-cyberians, the emergent creations of our physical, mental and spiritual integration with our technology. (4)

The Characteristics of Complexity

The study of complexity attempts to bring rigor to the social sciences by applying such hard sciences as quantum physics, liquid dynamics and chaos theory to social systems. "Scientists," writes Kiel, "are growing ever more aware that the behavior of nonlinear systems parallels and explains the behavior of a variety of social and human systems." (5) Thus, as complexity science provides our best, most up-to-date metaphor for dealing with complex social systems, by categorizing the Web more rigorously as a complex adaptive system, we might understand and erect public policy upon a rapidly expanding conceptual foundation.

However, in current discussions of complexity theory, complex adaptive systems are described inductively as exhibiting a variety of generic characteristics rather than defined deductively. Such a process reflects the basic givens of complexity, which argue that truly complex systems can only be described by the systems themselves. (6) Trying to break down complex systems into sets of formulae is wrong-headed. Rather, one should distill their identities by looking at the commonalities of their patterns and behavior as a whole. Complexity theorists argue that the mistake of the Newtonians, for example, was to assume that we would learn something about the nature of a Lego car by repeatedly smashing it against a wall (a la particle accelerators) and shattering it into ever smaller pieces. However, after our Newtonian hyperactive tantrums, all we have are individual characteristics of Legos, of Lego chemicals, of Lego atoms, of subatomic Lego particles. We learn nothing of the "nature" of the original Lego car; what it's for and what makes it different from a Lego house made from the same Legos. (7)

Thus, continues the argument, since mathematical or otherwise reductionist definitions of complexity are presently unavailable (and might never be), it is most effective to understand complex systems to the degree to which they exhibit certain generic characteristics. In particular, five characteristics seem prevalent in all complex adaptive systems:

Complex Systems Self Organize and Facilitate Emergent Order - One might imagine that such a monstrosity of thousands and thousands of nodes, each with varied standards, and in the case of the millions of multi-national users, embodying varied values, would be a jumble of mixed messages and useless data. Instead, the Web is a unified and self-organized system.

At first, self-organization seems counter intuitive. "Who runs the net?" is one of the most frequently asked questions of new users. And of course, the answer, "nobody," never satisfies. But actually, nature provides ample examples of such "order for free". "The marvel of 'hive mind," writes Kelly,

"is that no one is in control, and yet an invisible hand governs, a hand that emerges from very dumb members. The marvel is that more is different. To generate a colony organism from a bug organism requires only that the bugs be multiplied so that there are many, many more of them, and that they communicate with each other. At some stage the level of complexity reaches a point where new categories like 'colony' can emerge from simple categories like 'bug'. Thus, there is nothing to be found in a beehive that is not submerged in a bee. And yet you can search a bee forever with cyclotron and fluoroscope, and you will never find the hive." (8)
Hence, in any complex system, one expects emergence to a higher level of order and complexity to be generated endogenously, that is, without outside control or guidance. The Web is indeed such a system. Christopher Davis writes that the Web is controlled by, "lot's of people, and nobody, and the National Science Foundation, kinda, sorta. In a sense the net is a cooperative endeavor, both institutional and anti-institutional at the same time, massive and intimate, organized and chaotic." (9)

Thus, although no supranational government has yet risen to set standards, regulate commerce or define law, order has emerged endogenously in the form of protocols like HTML, netiquette or the self-regulation of Usenet and "spamming". In fact, not only does the Web provide its own order, it strongly resists outside regulatory pressure. John Gilmore, co-founder of the Electronic Frontier Foundation is often cited as saying that the reason the Web can never be censored, is because it will simply "route around" censorship as if it were a disease.

Such imagery calls to mind the actions of mail daemons, programs built to deliver information anywhere in the network regardless of whether or not all the nodes in the system are functioning. Mail daemons independently decide on the best way to navigate the Web and deliver their information. In a more than anthropomorphised sense, they are free to "choose" alternative paths when faced with node failure or information traffic jams. (10) Built to maintain command/control structures in the case of nuclear war, these AI daemons are metaphors for the very nature of the larger net which resists authority and supports the autonomy of its sub- components. Thus, the hive-mind of the Web exhibits its own emergent agency beyond the will or influences of its sub-components like national governments.

A further, more profound corollary of emergence applies in the case of the Web. As Kelly writes,

"What else is packed into the bee that we haven't seen yet? Or what else is packed into the hive that has not yet appeared because there haven't been enough honeybee hives in a row all at once? And for that matter, what is contained in a human that will not emerge until we are all interconnected by wires? The most unexpected things will brew in this bionic hivelike supermind." (11)
Thus, as the Web grows to incorporate a larger, global population, we should expect profound changes to emerge, along the lines of those forecasted by McLuhan and Toffler. Using Marchetti's S-Curve metaphor, forty years into the information era, the rapid growth in technology, participation, and the shear number of emergent properties signal the beginning of the end of the "initial conditions" period in the life cycle of the Web, characterized by highly intensive investment and little transformation, and heralds the beginning of the rapid period of transformation characteristic of all technological and social S-Curves in their "maturing" period. (12)

Thus, getting policy correct now is doubly important since the present period of sensitivity to initial conditions allows us to have great effect on the future, and since an end of that period threatens to hamstring future policy due to post-bifurcation lock-in. In other words, we stand at a crossroads leading to many possible futures. But once we choose our path, it will be impossible to go back, so we had better choose carefully.

Complex Systems are Large and Intensively Interconnected - Princeton University biology professor J.T.Bonner notes that, "complexity increases with size in both plants and animals." (13) This is fairly intuitive. As any Kindergarten teacher or president will tell you, the more equations, or people, or widgets are involved in anything, the harder it is to assert control over them. So we should expect that the Web is large and includes many agents.

In fact, as LaQuey noted long ago in 1992, "It's been estimated that between 5 and 10 million people use the [Web] itself and that upwards of 25 million people can exchange online messages between the [Web] and other interconnecting networks." (14) Along the same lines, Mark Lotto notes, "in 1981, 213 computers were registered on the [Web]; by 1989 there were 80,000. In October of 1990 there were 313,000; only three months later in January of 1991, there were 376,00. And in January 1992, there were 727,000 [Web] registered computers." (15) The web certainly satisfies Bonner's law.

However, size or number of components is not enough. As Flood notes, complexity is defined not only by the number of parts, but by the number of relationships between those parts as well. (16) Without interactivity, a huge quantity of agents in one place is simply chaos. Interactivity, in the form of information exchange, or a system of feedback, provides the extropian spark. Pippenger writes, "...the most important lesson of complexity theory is the demonstration of the diversity of phenomena that can arise through the interaction of [many] simple components. Today's computer and telephone exchanges [for example] present problems beyond our understanding." (17)

Of course, as Pippenger recognizes above, as a medium of interconnectivity, the Web is not unique. The Web can certainly be seen as an incremental, evolutionary step beyond earlier networking technologies including roads, the telegraph, and the telephone which, like the Web, involved 1) rapid (virtually instantaneous in the case of the Web), distance transcendent (global in the case of the Web) communication between a great number of individuals and groups (significantly more, simultaneously in the case of the Web) and 2) a decentralized and horizontally-linked architecture.

However, the Web is revolutionary as well as evolutionary. Unlike earlier mediums of interactivity, the Web involves such a degree of hyper-feedback and hyper-interchange of information that it facilitates emergence to a higher order. More is different. Just as water "phase shifts" at a critical temperature, information phase shifts at a critical point of interactivity. The Web is qualitatively different from any technology before it in its ability to foster emergence because it fosters emergence so efficiently.

"It's fantastic," writes Mark Fletcher.

"I can carry on two or three unrelated conversations at one time. I feel like I'm on speed, my brain is working so hard. On one [IRC] channel I'm talking about my cyber-law paper with this PhD at UCSD with two web browsers open, one raiding Yahoo for links to legal resources and the other cross referencing the Library of Congress. On the other channel, I'm editing a recent article for Toposthesia [an online journal of Anthropology] with five of the other editors on the board. And on the third channel, I'm talking with some really cool girl named "enymore!" And all while my script is logging the whole thing, breaking it up into separate HTML documents and saving it to my webbed reference library. Is this power or what?" (18)
When interactive experiences like these are multiplied by millions of both biological and technological autonomous agents, we see a true complex system and the emergence of a true "cyborg reality". It is exactly because information flows are hyperactive due to the number of interconnects that such emergence is possible. In fact, such is the case with all emergence which proceeds because of the anxiety of information overload. That is, it is the very nature of that which exists, to become. As Clark writes, "[evolution] is essentially a self organizing [process]...involving information converted into more complex internal structure and entropy exported into the environment." (19)

Thus, where matter is the emergent property of energy, biology the emergent property of matter, culture that of biology, cyborgification is the emergent property of culture. Each step is simply the result of information interactivity outgrowing its shell, causing emergence of more complex structures with which to facilitate its own propagation. The Web then, combines machine and mind to create a higher order emergent complexity.

Complex Systems Complexify Relative to the Level of Autonomy and Decentralization of their subsystems - "Another theme under the aegis of complexity," writes Pagels "is the emphasis on parallel (network) rather than serial (hierarchical) systems." (20) Serial systems are hierarchical systems like pyramidical organization charts for corporations, the church, or the military with clearly visible "tops" and "bottoms" which dissolve if the "top" is cut off from the rest of the system. Parallel systems on the other hand, have no top or bottom. Rather they exist as a "plurality of connections that increase the possible interactions between components of the network." (21) There is no need for centralization of authority because autonomy assures that if a part of the network is destroyed, the whole network continues to function. However, such systems demand their own form of management for what resilience is bought is paid for in complexity.

For example, game theory has shown that even the change from a two to a three player game is a non-linear step leading to incredibly complex relationships between players. Whereas Von Neumann's Minimax Theorem could derive an "optimal pure strategy" in simple two-player cases, for non-finite, non-zero-sum, multiple player games, it did not apply. (22) Thus, for example, Von Neumann realized that though extremely complicated, Chess has an optimal solution whereas a multiple player game like Risk, is truly complex and hence, does not.

Similarly, when John Stuart Mill wrote that free speech was a cornerstone of democracy, he was really talking about emergence from interacting autonomous agents. Truth, he explained, could only arise if people were free to express, explore and discuss. (23) In other words, truth is an emergent property of autonomous anarchic interaction of memes and their human carriers. When ideas are controlled hierarchically however, freedom and human cultural evolution is stunted.

Thus, acceptance of network architecture need not be an acceptance of unmanageable chaos. As Flood explains,

"Complexity, arise[s] when there is some sort of localized transient anarchy in a system. [This] anarchy can cause such strains on a system, including positive feedback, that the system itself may go through a transition. In many cases anarchic behavior is the catalyst of system transition, not from one state to another (within predetermined state space; one-to-one mapping of state variables), but rather, from one nature to another." (24)
Multiple player games and other autonomous, decentralized systems, do converge, they just converge chaotically. Autonomous agents are necessary for innovation in all complex adaptive systems and that innovation drives emergence. Without them, a system cannot phase shift and is left stagnant within its predetermined state.

Just as Darwinian evolution needs that innovative solar flare to instill anarchy in the genetic pool, innovative action confounds the Game, and innovative expression stirs the meme pool. Without autonomy, institution petrifies, things die, empires fall and anarchy invades. For without the extropian engine of emergence, entropy reigns. Thus anarchy becomes both executioner and parent of structure.

Of course, as explained before, the Web is a network by design and the free expression of memes has been paramount to the emergent culture. In a sense, when the government created the precursors to today's Web, like NSFNET, which decentralized control and enhanced the autonomy of subnetworks, it forever let the genie out of the bottle.

Unlike television or most print media, there are no centralized truth makers on the Web. Every user is her own publisher and broadcast station. This, of course, has serious implications for public policy, and for the last few years, we have seen a barrage of public policy measures meant to reign in the Web, from last year's Clipper Chip scheme to this year's, Communications Decency Act (CDA) debacle. The CDA in particular, is a ripe example of both the creative potential of autonomous systems and the need for our public policy to adapt significantly to its new hyper-culture environment.

In short, the CDA attempts to protect children from readily available pornography on the Web. However, it seeks to do so by making it illegal for anyone to distribute such pornography (which would include steamy love notes between spouses, pornography legal in print, and common insults typical of "flames"), and threatens to send everyone and their technophobic grandmothers, to jail in the process. It's more complex than that of course, including provisions for service providers, arcane legal definitions of "indecent and filthy" language etc, but according to such groups as the ACLU and the Center for Democracy and Technology, that is the bottom line of the bill. (25)

However, as Pagels explains above, you cannot control a network like the Web from above. Despite the civil liberties issues raised by such blatant censorship, the bill is definitionally absurd and technologically impossible to enforce. For one, there are simply too many nodes on the Web. Because the Web is decentralized, it is impossible to stamp out speech without stamping out all the nodes, and since there are so many autonomous nodes, we are talking about a super-human task.

Whereas in the case of the hierarchical medium of television, in which stations could easily be shut down and "offensive" speech could be censored, on the Web, as one node is closed, a dozen more sprout. Further, of the dozen new nodes, four will flower in foreign countries or different states. Web activists have continually warned that to successfully censor the Web one would have to enforce moral standards globally. What is obscene to the Brazilians must coincide with what is obscene to those from San Francisco, Tennessee, Iran, and Amsterdam or else censorship will fail, for what is banned in one community will be easily had by jacking into nodes in another.

Finally, the anonymity built into the Web preserves and facilitates autonomous actions of the mass of Web users. Even if a world government asserted standards for the entire global Web, users would still be able to hide in the system, ghosts in the machine. In short, the Web's decentralized architecture creates a Millian utopia in which autonomy is guaranteed and free speech refuses to be quelled.

Typical of emergent systems, however, control seeps up from below. Do not think that your children are lost forever to the hoard of bomb-making, drug-dealing, child-molesting perverts on the Web as the FBI and Senator Exon would have you believe. Such emergent control includes filter services like those offered by Surfwatch, the first company to prescreen the Web for a fee, to significant standards/protocol changes being implemented by the Netscape/Microsoft web rating system to allow individual families to easily set their own level of information content. Both Surfwatch and the Netscape/Microsoft collaboration exemplify how complex systems innovate solutions as quickly as they create problems when left alone. Uninformed policy only leads to unintended, and often negative consequences elsewhere in the system. (26)

Of course, in this initial period, it is a crap shoot as to what endogenously derived solution will emerge. But we can be sure that one will. Bills like the counter-CDA Leahy Bill represent a more complexity-aware approach by calling on the government to support endogenously grown solutions rather than attempt to implement them from outside the system.

Complex Systems Exhibit Non-Linear Behavior - "Deterministic Chaos," writes Radzicki, "is characterized by nonrepetitive and nonpredictable fluctuations that arise through the interplay of a system's instability generating mechanisms (such as instrumental valuation with self- reinforcing behavior) and nonlinear confining factors (often representing human values)." (27) In other words, though one might look back at the history of complex adaptive systems and say that A caused B, one can never look forward and say that B will cause C because chaos guarantees something surprising. This is because in complex systems A, B, and C are always in flux due to the internal dynamics of their feedback systems. Autonomus agents continually mix, merge, die, divide, innovate and generally act in confounding ways. Thus, in complex systems, determinism and causality are only valid retrospectively and reality is a constant reformat.

Specifically, chaos can be understood in terms of 1) sensitivity to initial conditions and 2) the existence of chaotic attractors. Of course the study of chaos is much more dense and involved than these two characteristics suggest. However, such a discussion is beyond the scope of this study. Nevertheless, identifying sensitivity to initial conditions and the existence of attractors is a "good start" when identifying the non-linear aspects of complex adaptive systems and so it makes some sense to at least cover them.

Perhaps the mantle-piece metaphor for chaos theory is sensitive dependence which argues that during specific points in time, called bifurcation points, the actions of very small factors can have huge effects on the destiny of all the other parts in a system. Thus, it is said that the flapping wings of a butterfly in San Francisco can "cause" a hurricane in Tokyo. The "butterfly effect" shows that surprise and anarchy are definitional to complex systems. Kiel writes, "Consider the tragic event of the space Shuttle Challenger disaster of 1986. A lack of communication about the potential failure of the shuttle's O-rings led not only to the disastrous explosion, but to an erosion of confidence in NASA that still lingers today." Thus, not only do minutia greatly affect complex systems in ways that seem to correlate (ie: bad O-ring=mechanical failure) they also have non-obvious (NASA's current image problem), but equally dramatic implications. (28)

Like any complex system, the Web abounds with the products of sensitive dependence. Perhaps the most influential are the metaphor-makers of the Web. It is no coincidence, for example, that the policies adopted by the federal government in support of the growth of the Web resemble the policies applied in the creation of the interstate highway system of the 1950s. (29) Afterall, the father of Vice President Gore, the largest proponent of the "information superhighway" metaphor, was a pioneer in the 1950s interstate highway policy. Nevertheless, adoption of a metaphor is necessarily an acceptance of some, and a rejection of other, policy options and sensitive dependence on Gore's lineage turns out to have a great impact on how the Web unfolds, especially in terms of super highway inspired programs like NREN or NSFNET. Other influential metaphors include the "electronic frontier," introduced by ex-cattle rancher John Perry Barlow and the "virtual community," popularized by such pop cyberanthropologists as Howard Rheingold.

Similarly, authors William Gibson (Neuromancer) and George Orwell (1984) have both had great affect on the collective conception of what the Web will look like. Their writings have inspired and "grown" a grassroots native identification associated with the libertarianism and anarchy of the cyberpunks and hackers, and their stories are being forever retold in the popular media in film, television and print through such works as Johnny Mneumonmic or Blade Runner.

Thus, both the fears and fantasies of cyberians become interlaced with the visions of these writers, and since cyberians create and recreate the Web after their own imaginations, the web is made in the image of Chiba City and its citizens primary fear becomes insidious government and corporate intrusion of privacy and autonomy. Thus the extremely widespread "US out of Cyberspace" attitude has arisen from humble pre-Web beginnings.

Yet cybercatepillars continue to hatch. The Web for example, is fertile ground for entrepreneurs of all sorts from online activists to innovative businessmen. On the one hand, online activists have used the Web to organize community action from organizing students in Tienamen, to multinational support for the Chiapas, to fighting legislation in United States Congress. However, these huge tidal waves of activism have traditionally begun with small groups of dedicated activists. "Action Alerts" spread like wildfire across the Web and their effects on policy are often dramatic. Within 3 days, for example, the Anti-CDA petition had received thousands of signatures, adding 100/hour at some points.

Similarly, few deny the vast impact the design team of Netscape has had and will continue to have on the shape of the Web and the faces of cybercommerce, education and community. In just two years, Netscape has gone from startup to giant and, in the process, redefined the very language of the Web to HTML3.

A second aspect of chaotic systems are chaotic (Strange) Attractors which are determined by such fluctuations in initial conditions. "The set of initial conditions that yields dissipative paths to a particular attractor in a system," explains Radzicki, "defines that attractors 'basin of attraction'" (30) All systems have attractors of course. The basic Newtonian cause and effect loop is a form of attractor as is the circle traced out by a pendulum. However, though chaotic attractors attract, they do not converge as other non-strange ones do. "The crucial feature of these strange paths" continues Pagels, "is that if you examine two nearby paths in state space and follow them along, they rapidly diverge." (31)

Thus, in a complex adaptive system, the adaptability comes from the fact that the post- bifurcation period is followed not by institutionalization, but by more bifurcation. Sensitive dependence continues to affect outcomes and the system never settles. What is revolutionary about the Web is its dynamics. Culture is a good example of a chaotic attractor, but culture is always a negative feedback, an inhibiting factor just like genetics was for evolution before culture emerged to outstrip genetics.

Life on the Web, and as the Web begins infiltrating biological life, life in the real world, becomes a matter of continually navigating a changing reality. John Perry Barlow expresses the ramifications saying,

"I look at confusing circumstances as opportunity - but not everybody feels that way. That's not the standard neurotic response. We've got a culture that's based on the ability of people to control everything. Once you start to embrace confusion as a way of life, concomitant with that is the assumption that you really don't control anything. At best it's a matter of surfing the whitewater." (32)
Such a situation is fostered on the Web because the rapid change of technology keeps the landscape unstable and continual fusion and innovation by autonomous and hyper-interactive agents, continually upsets initial conditions by reformulating linkages and creating new ones. Hence the Web becomes a continually evolving system. The lesson of chaotic attractors is that in a complex world, institutions become the exception.

Complex Systems are Evolving - But what does it mean to evolve? Emergence shows that large, autonomously composed systems do evolve and chaos theory shows that they do so continually, but one must ask, "Why do they do it? What is the mechanism?"

As Dosi and Nelson explain, evolution is the process which generates change in a system over time. Such a process involves 1) some process of imperfect learning and discovery, 2) some selection mechanism and 3) a unit of selection. (33) That is, for change to occur in a system, there must be some variation from a norm and some way to choose which variation into which the system will metamorph.

With regard to the Web, we can identify the unit of selection as data. For in the virtual community, what is being exchanged in the form of bits and bytes is simply data. Further, the method of selection takes the form of coevolution which is, of course, the endogenous engine of change. And finally, learning, or novelty can be seen as generated endogenously by intensely interactive agents as discussed earlier.

In CoEvolution, Stuart Brand describes coevolution in biological terms.

"In eastern Mexico live a variety of acacia shrubs and marauding ants. Most acacias have thorns, bitter leaves, and other protections against a hungry world. One, the "swollen thorn acacia," learned to encourage a species of ant to monopolize it as a food source and kill or run off other predators. Enticements gradually included nifty water-proof swollen thorns to live in, handy nectar fountains, and special ant-food buds at the leaf tips. The ants, whose interests increasingly coincided with the acacia's, learned to inhabit the thorns, patrol the acacia day and night, [and] attack every acacia-hungry organism....The acacia gave up its bitter leaves, sharp thorns, and other devices and now requires the acacia-ant for survival. And the ant colonies can no longer live without the acacia. Together they're unbeatable." (34)
However, coevolution is not limited to the biological realm. As the emergent property of genes and their selective environment is evolution, the continual morphogenesis of bodies, the emergent property of data and their selective environment is information, the continual morphogenesis of ideas.

The structure of the Web changes intellectual property law, for example, because the net provides instantaneous, perfect, copies of anything anywhere. Thus enforcement of traditional American intellectual property becomes impossible. This new environment has significantly changed the expression of ideas as can be seen in the emerging market of shareware. Intellectual Property theorists have been discussing this dilemma for the last seven or eight years now and have realized that intellectual property law is anachronistic. More recent work by John Perry Barlow and Esther Dyson suggests that intellectual property law is heading for a transformation which will foster a new system which will; facilitate the flow of information rather than the "bottling." This metamorphosis, in turn, will have an effect on the expression of ideas. Thus, on the Web, intellectual property laws and the ideas they hope to manage are coevolving. The concept of shareware might continue, but only to facilitate the emergence of an information service sector around those products. (35)


So, what does all this complexity stuff mean for public policy which has traditionally focussed on hierarchy, predictability and control? As noted previously, the first thing to understand is the urgency of this dilemma. Focussing on the rate of change of the Web and its emergent by- products, it is clear that we are reaching a critical point in its development. If we fail to act now, we will miss the boat completely, and if we fail to act appropriately, we will sink it. Such fears are mirrored in Kirby's excellent discussions on the implications of complex society on democracy. Kirby writes, "The implications for democracy are that as system performance declines, the public will look for (and politicians will be tempted to offer) simple solutions based on purely ideological responses." (36) Recent reactions like the Exon Bill or the post-Oklahoma bombing Anti-Net- Terrorism Bill suggest that Kirby's fears are well founded.

However, Soedjatmoko notes that even should a government turn towards authoritarian- ism, such a regime will not solve the problems of complexity,

"Future scenarios remain silent on the question of freedom. The implicit assumption always seems to be that the management of the global interdependent system will ultimately have to be authoritarian. This to me is unacceptable....Complexity cannot be managed, intellectually or practically, through increased control. We have to learn to understand and manage complex systems while respecting the autonomy of the processes and the elements within these systems." (37)
In the long run, though many might be martyred in the process of transformation, authoritarianism, or other types of control from above, will fall just as hard as the Soviet Union did. information "wants" to be free and the emergence of complexity is beyond the control of any command structure. However, it would be much more preferable not to try.

Thus a second implication of complexity on policy is that complex systems cannot be manipulated with traditional policy tools if we are to apply any human agency at all to our destiny. Just as one would not use a hammer for a saws job or vice versa, one should not use Newtonian policy tools on complex, non-Newtonian problems. Instead, we must accept certain basic assumptions about complex systems as recently outlined by such theorists as Kirby, Loye and Eisler, Kiel and others and build from there.

Specifically, unlike hierarchical systems, complex systems like the Web are 1) nonoptimal in that they necessarily contain redundancy (and thus demand policy aimed not at specialization but at unification), 2) noncontrollable in that they are too complex for any hierarchical system to contain (and thus demand carefully researched and rarely used policy measures), 3) nonpredictable in that they are sensitive to initial conditions (and thus demand policy makers to "spread their risks" over many possible futures, and 4) nonunderstandable in that emergent phenomena are typically of a different nature rather than simply on a higher order of magnitude. Thus, two plus two will not equal four or five, but apple. (38)

Hence, our goal in setting policy for the Web should be to enhance these characteristics to our own benefit. further, policy has a goal in facilitating smooth transition from industrial to information society and from Newtonian to complexity philosophy. Arthur's classic "paddle recommendation" is ever-appropriate.


  1. Pagels, Heinz R. The Dreams of Reason: The Computer and the Rise of the Sciences of Complexity, New York; Simon and Schuster, 1988 (pg 53)
  2. LaQuey, Tracy and Ryer, Jeanne C. The Internet Companion, Reading; Addison-Wesley, 1992 (pg 21)
  3. Ibid (pg 22)
  4. McLuhan, Marshall. The Medium is the Message,
  5. Kiel, Douglas L. Managing Chaos and Complexity in Government: A New Paradigm for Managing Chaos, Innovation, and Organizational Renewal, San Francisco; Jossey Bass, 1994 (pg 5)
  6. Pagels, Heinz R. The Dreams of Reason: The Computer and the Rise of the Sciences of Complexity, New York; Simon and Schuster, 1988 (pg 54-59)
  7. Stewart, Ian and Cohen, Jack. Why are There Simple Rules in a Complicated Universe?; Futures, July/August 1994 (pg 654)
  8. Kelly, Kevin. Out of control: The New Biology of Machines, Social Systems, and the Economic World, Reading; Addison-Wesley, 1995 (pg 13)
  9. LaQuey, Tracy and Ryer, Jeanne C. The Internet Companion, Reading; Addison-Wesley, 1992 (pg 27)
  10. DeLanda, Manuel. War in the Age of Intelligent Machines,
  11. Kelly, Kevin. Out of control: The New Biology of Machines, Social Systems, and the Economic World, Reading; Addison-Wesley, 1995 (pg 25)
  12. Casti, John L. Complexification: Explaining a Paradoxical World Through the Science of Surprise, New York: Harper Collins, 1994 (pg 37)
  13. Cambel A. B. Applied Chaos Theory: A Paradigm for Complexity, Boston: Academic Press, 1993 (pg 19)
  14. LaQuey, Tracy and Ryer, Jeanne C. The Internet Companion, Reading; Addison-Wesley, 1992 (pg 7)
  15. Lottor Mark. Internet Growth (1981-1991); RFC 1296, January 1992
  16. Flood, R.L. Complexity: A Definition by Construction of a Conceptual Framework, Systems Research, No. 3, 1987 (pg 178)
  17. Ibid (pg 179)
  18. Log capture from Cleveland Freenet IRC.
  19. Clark, Norman. Organization and information in the Evolution of Economic Systems, Reading: Harwood Academic, 1991 (pg 98)
  20. Pagels, Heinz R. The Dreams of Reason: The Computer and the Rise of the Sciences of Complexity, New York; Simon and Schuster, 1988 (pg 50)
  21. Ibid
  22. Shapely, Lloyd S. Game Theory Hyperlink, The Software Toolworks 1994 Encyclopedia CD ROM
  23. Mill, John Stuart. On Liberty
  24. Flood, R.L. Complexity: A Definition by Construction of a Conceptual Framework, Systems Research, No. 3, 1987 (pg 181)
  25. CDA Legal and Policy commentary may be found at... http://www.eff.org/pub/Alerts/Communictions_Decency_Exon_bill/00-links.html
  26. Kirby, M. J. L. Complexity, Democracy, and Governance, Tokyo: United Nations University 1985 (pg 5)
  27. Radzicki, Michael J. Institutional Dynamics, Deterministic Chaos, and Self-Organizing Systems, Journal of Economic Issues, March 1990 (pg 63)
  28. Kiel, Douglas L. Managing Chaos and Complexity in Government: A New Paradigm for Managing Chaos, Innovation, and Organizational Renewal, San Francisco; Jossey Bass, 1994 (pg 7)
  29. Karraker, Roger. Highways of the Mind or Toll Roads Between Information Castles Sausalitto; Whole Earth Review 1991 (pg 3)
  30. Radzicki, Michael J. Institutional Dynamics, Deterministic Chaos, and Self-Organizing Systems, Journal of Economic Issues, March 1990 (pg 67)
  31. Pagels, Heinz R. The Dreams of Reason: The Computer and the Rise of the Sciences of Complexity, New York; Simon and Schuster, 1988 (pg 78)
  32. Rushkoff, Douglas. Cyberia: Life in the Trenches of Cyberspace
  33. Dosi, Giovanni and Nelson, Richard. As Introduction to Evolutionary Theories in Economics, Journal of Evolutionary Economics, No 3, 1994 (pg 155)
  34. Kelly, Kevin. Out of control: The New Biology of Machines, Social Systems, and the Economic World, Reading; Addison-Wesley, 1995 (pg 75)
  35. Dyson, Esther. Intellectual Property on the Net, http://www.eff.org/pub/Publications/Esther_Dyson/ip_on_the_net.paper
  36. Kirby, M. J. L. Complexity, Democracy, and Governance, Tokyo: United Nations University 1985 (pg 6)
  37. Ibid pg 7
  38. Kelly, Kevin. Out of control: The New Biology of Machines, Social Systems, and the Economic World, Reading; Addison-Wesley, 1995 (pg 23)

    Selena's Home | Creations | Seleves | Photo Gallery | Resume