Information Overload

IN THE INFORMATION AGE it is widely understood that there is now too much information. As the process of electronic publishing on the Web is simplified and digital storage capacity increases, it is difficult to find evidence of quality over quantity. But even efficiently managed data and well constructed documents have begun to lose value due to the sheer volume of information available. According to the estimate of one web expert, the size of the World Wide Web is to grow to 50 billion pages by 2005 (Nielsen 347). Some of this newly created information will most certainly be valuable, but despite marked improvement in search tools, finding the valuable information is a slow panhandle. In The Social Life of Information, authors John Seely Brown and Paul Deguid cite a study finding that "Digital technologies currently produce between one and two exabytes (one billion gigabytes) per year" (Seely Brown and Deguid XIII). They then go on to eloquently conclude that,

Storage does not correlate with significance, nor volume with value. Standing atop gigabytes, terabytes, and even exabytes of information will not necessarily help us see further. It may only put our heads further into the clouds. (Brown and Deguid XIII)

Arising out of this environment is a proposed framework for a more manageable and thus more useful model for the World Wide Web. Tim Berners-Lee, the inventor of the Web, and the World Wide Web Consortium (W3C), a non-profit organization led by Berners-Lee, have been pushing a recommendation for a Semantic Web in much the same grassroots fashion that began the development of the Web we know today. The difference this time around is that in addition to the small corps of PC enthusiasts and computer scientists that helped make the first Web, Berners-Lee has the support and interest of industry and big business.

What is the Semantic Web?

Today's web is a combination of hypertext technology and internet technology. The strength of the idea is in its simplicity. Hypertext Markup Language or HTML was created by Berners-Lee as a simple tool for creating Web documents. In this sense it serves its purpose well. But the strength inherent in this model is also its weakness. While HTML makes it easy to produce documents that are readable by humans, it is not practical to access the data stored in these documents with software programs. Thus as the simplicity of the web allows many people to publish information, it also means that the vast amount of unstructured and difficult to access information made available by this facility is overwhelming. Perhaps in light of this situation, the W3C under the direction of Berners-Lee has begun to build the foundation for the next phase of the web. This phase, called the Semantic Web, will make information stored with this technology much more processible by machines (Berners-Lee 177).

In HTML a document is described in terms of its structure. For example, an element of HTML called a tag tells a browser program how to display a piece of data. The tag can describe a wide range of data such as titles, paragraphs, and images. By processing or "interpreting" this tag the browser can display whatever data is contained by the tag. The data within the tag is then formatted according to specified properties such as font size, bolding and alignment. When the browser displays an HTML document, humans can read the contents of a tag such as a title and understand what it means relative to the rest of the information on the page. On the other hand, the logic coded into a software program cannot easily infer what the title means when it is coded as HTML.

The idea with the Semantic Web is to create documents that represent the information in a highly structured fashion. This representation is strictly semantic and contains no information on how to present the information. A simple analogy would be to think of the Semantic Web as a large relational database or table (Berners-Lee 180). Individual documents would be rows and the columns would represent document properties. Each of these documents would then be ordered hierarchically from left to right, or even in three dimensions, with many tables stacked on top of each other. At the same time the web-like structure is far more organic than the angles of columns and rows.

Where the World Wide Web employs HTML to create documents the Semantic Web uses another language called XML. The creation of Semantic Web documents and groups of related documents called "ontologies" are the foundation for storing information on the Semantic Web. The word ontology is borrowed from a branch of philosophy called metaphysics, and the entire structure of the Semantic Web is intended to be composed of interconnected ontologies that form what is known as a "Knowledge Representation". Knowledge Representation in the field of study called Artificial Intelligence had its roots in both Phenomenology, a kind of philosophy and Connectionism, an approach to Linguistics. Ontologies are standardized to a limited degree through the RDF or Resource Definition Framework. The RDF provides a syntax that acts as a universally understood structure. This neutral system will allow software schemata and artificially intelligent agents created by anyone to access the information stored on the Semantic Web.

Another crucial element of the Semantic Web is the "agent" technology. Agents are software programs that contain logical algorithms that process information and make intelligent decisions based on a pre-programmed set of rules. There are already many examples of agents in use today. Search engines for example perform a set of logical processes based on the search criterion you supply them. On the Semantic Web the hope is that many different agents will be able to work together to perform even more complex tasks than are performed by agents today.

Lessons Learned From the Field of Artificial Intelligence

This is just one of a number of similar statements that reveal Berners-Lee's optimistic and some might say grandiose claims about the potential for the Semantic Web. This enthusiasm is understandable given the scope of this project, but similar projects in the field of Artificial Intelligence over the last half century have failed to live up to expectations (Dreyfus; Searle). Detractors of the field of Artificial Intelligence have criticized the tendency of AI researchers to make strong claims about the potential for computers to emulate or surpass the human faculties. These arguments generally point out that the human mind is far more complex than AI researchers had anticipated, and early AI projects have led AI researchers to concede that the complexity of human thought was indeed underestimated. At the heart of this debate there is one particular dialogue that seems to capture the essence of issues in the field. This debate between one of the founders of the AI field Marvin Minsky and U.C. Berkeley philosopher Hubert Dreyfus surrounds the concept of Knowledge Representation and is particularly pertinent to the design of the Semantic Web.

In 1974 the MIT AI researcher Marvin Minsky published an article titled "A Framework for Representing Knowledge". It was an article that would influence a whole generation of AI researchers. In this article Minsky laid out a broad theory of the mind informed by earlier AI research and principles of computer science. Minsky specifically credits the psychologist Frederick Bartlett who did notable work on schemata and physicist Thomas S. Kuhn the influential science epistemology theorist as primary influences (Mind Design 113). In this theory Minsky identifies the "frame" and "frame systems" as comprising human thought in the most basic sense. "Transformations" effected by events or actions are reflected in the reordering of frames and frame systems. Many of these frames and frame systems share the same "terminals" where "default" assignments are stored. According to Minsky this structure is what allows the human mind to make important intuitive generalizations (Mind Design 112).

Minsky correlates this theory with the scientific "paradigms" of Thomas S. Kuhn. He presents Kuhn's argument as a societal version of his own theory of the mind. Minsky writes, "Now while Kuhn prefers to apply his own very effective redescription paradigm at the level of major scientific revolutions, it seems to me that the same idea applies as well to the microcosm of everyday thinking" (Mind Design 135). Thus in Kuhn's The Structure of Scientific Revolutions the practice of "normal science" is the period of practicing what Minsky describes as "default" frames. One can then conclude that according to Minsky's theory, "Transformations" in the mind like "Scientific Revolutions" occur when a high level frame system is replaced by a newly created frame system.

As a general criticism of AI and as a direct response to Minsky, Hubert Dreyfus attacks these ideas from a philosophical perspective in his well known article "From Micro-Worlds to Knowledge Representation: AI at an Impasse". Dreyfus argues that much of AI research as of 1979 had been limited to the "micro-world". That is, much of the research in replicating human intelligence depended on artificial environments that were extremely limited in scope. This basis of study led, according to Dreyfus, to narrow and impractical applications of machine intelligence. Instead of focusing on these small environments, AI researchers should instead begin with a broad, high level theory of consciousness and human intelligence that draws upon the background knowledge of philosophy.

To support a particular instance of this idea in relation to Minksy's theory, Dreyfus cites the work of the phenomenologist Edmund Husserl. Specifically Dreyfus explains that Husserl had himself attempted to propose a theory of "transcendental-phenomenology" similar to Minsky's theory of frames (Mind Design 161). Dreyfus describes the work of Husserl and points out that while Husserl began by attempting to describe his version of Knowledge Representation for simple objects called the "noema", he ended up discovering that even the most basic ideas are inevitably linked to ever broader conceptions of the outside world (Mind Design 162). In a particularly compelling fashion Dreyfus says of Husserl and his student Martin Heidegger,

He sadly concluded at the age of seventy-five that he was 'a perpetual beginner' and that phenomenology was an 'infinite task' – and even that may be too optimistic. His successor, Heidegger, pointed out that since the outer horizon or background of cultural practices was the condition of the possibility of determining relevant facts and features and thus prerequisite for structuring the inner horizon, as long as the cultural context had not been clarified, the proposed analysis of the inner horizon of the noema could not even claim progress. (Mind Design 162)

Dreyfus then concludes his acerbic criticism by completing the comparison, "There are hints in the frame paper that Minsky has embarked on the same misguided 'infinite task' that eventually overwhelmed Husserl." And again he states, "Minsky's naïveté and faith are astonishing. Philosophers from Plato to Husserl, who uncovered all these problems and more, have carried on serious epistemological research in this area for two thousand years without notable success."

The Semantic Web, to the extent that it can be considered a theory of the mind, seems to draw from both of these positions. On the one hand the Semantic Web is modeled in large part after Minsky's theory of the mind, either directly or in the sense that Minsky's ideas influenced the field of AI on the whole. On the other hand the Web publishing model addresses the problem of limited micro-worlds raised by Dreyfus more convincingly than previous Knowledge Representation efforts, because it allows the knowledge base to be continually expanded and redefined. Of course whether or not this model can match or surpass the variety of human inputs remains to be seen, but the theoretical framework is markedly improved.

The Web as Organic System

There are three important qualities that contribute to the vast success of the World Wide Web. First, the structure of the World Wide Web is decentralized. This is important because it eliminates any potential bottlenecks for the Web's propagation. As control is not hierarchical, the Web is able to spread in a more natural way. Second, the Web is not platform specific. This platform universality eliminates problems associated with allowing various types of machines to communicate. In more natural social systems this is an important factor, as for unification among groups of people to occur, it is always necessary to have some level of agreement. Linguistic differences for example often act as inhibitors to unification of social groups. The third factor is economic as usage of the Web is basically free. There are of course connection fees and one must be able to afford a computer, but there are no charges specifically associated with the Web.

Tim Berners-Lee discusses these concepts in his chapter entitled "Protocols: Simple Rules for Global Systems". He points out that natural systems like the weather, follow similar principles (Berners-Lee 36). It is important to describe these qualities of the World Wide Web because they are also the foundations for the Semantic Web. It seems obvious that Berners-Lee has modeled the World Wide Web and the Semantic Web by carefully considering the structure of natural systems. Marvin Minsky too has built a theory of the mind that is of course modeled after the natural system of the human brain, and like Berners-Lee sees the parallels between society and the individual mind. The point is not to quibble over Berners-Lee's influences but rather to point out that lessons learned in AI seem to be internalized and addressed by the model of the Semantic Web, which is itself a theory of the mind.

The strength of the Semantic Web model is that like the World Wide Web it is not overly specific. It is a framework that leaves room for change and variation among parts. This is in fact an important part of the W3C philosophy. As an organization it does not dictate policy but simply acts as shepherd in an effort to promote standardized communication in a minimal sense (Berners-Lee 189). This strategy allows for tremendous variation. Specifically the model of the Semantic Web even extends the flexibility of its standardization to allow for ambiguity. Again this is also a recognized strength of the human mind.

The Web Experiment

Much of the debate concerning what the philosopher John Searle termed "strong AI" surrounds the question of whether machines can ever become conscious. According to Searle "weak AI" constitutes intelligent technologies that employ machines as mere tools. Strong AI assumes that machines have the potential to resemble the human mind in the sense that cognitive states can be created that constitute a form of consciousness (Searle 183). The point is not to argue this utterly philosophical question. Instead the Semantic Web should be considered as the next phase of the World Wide Web and thus as a possible next step in the evolution of human communication.

The success of the World Wide Web is widely recognized and there has been much speculation towards its effect on society. As was mentioned, the Web became popular because it grew as a kind of organic social phenomenon. Credit is due to its inventor, but no one blames Tim Berners-Lee for any of the negative repercussions. Computer viruses, terrorist communication, socially harmful content, and neglect of copyright are just a few of the issues that have arisen with this new technology. It is generally understood that these problems are no more than mutations of latent social issues, simply extensions of pre-existing societal woes. If there is any blame to be placed, it would be the failure to recognize the impending changes in time to take preemptive measures. Intellectual property based businesses are still reeling from this mistake. While it is unrealistic to expect such changes to be wholly anticipated, some preparation is prudent. The fact that the Semantic Web is still young means that involvement at this early stage could help point things in a positive direction.

In terms of strong AI the Semantic Web has a long way to go. Most of the early steps involve making machines that can increase the quality of everyday life (e.g. scheduling appointments, buying airplane tickets, etc.). Nevertheless, the proven foundations of the Web as social phenomenon and a well articulated theory of the mind modeled on these same organic principles are good reasons to take the Semantic Web seriously. Even if its potential as strong AI is never realized, its possible effect on society should be considered. Thus the question is not whether or not the Semantic Web will ultimately outthink the human brain, but rather, what effects will it have on society as a communication medium. In this sense the Semantic Web has the potential to effect change as dramatically as any other new communication technology, if not more. We are, to a certain extent, still coming to terms with societal changes due to the experiment of the World Wide Web. If the Semantic Web takes off in the way the World Wide Web did, it will be more and more difficult to shape its direction as it matures. As once social phenomena gather enough momentum, it is very difficult to do anything but step back and watch events unfold.

A Short History of Communication

Until the recognition of the World Wide Web as a new communication medium, many modern theories of communication began with the telegraph. But with the realization of the Web as an electronic publishing medium many people began rethinking the introduction of the print era in Western Europe. Comparisons between Thomas Gutenburg's printing press and Tim Berners-Lee's World Wide Web are now commonplace, and it seems clear that any contemporary theory of communication must begin with the printing press. Of course this is not to say that the telegraph was not an important step in the evolution of communication. The telegraph with its ability to transmit information almost instantaneously marked the beginning of an era where the world began to seem ever smaller. This has certainly changed the structure of society but in many respects this same transformation had already occurred to a lesser extent in the literate population. There are, however, some notable differences between the effect on society of print and of the electric mass communication of radio and television that was based on the telegraph.

As the historian Elizabeth Eisenstein has pointed out, many of the early European print houses arose out of the social structure of the age of scribes (Eisenstein 273). While she does not go as far as to attempt a full theory of the effect of print on society, she describes the social structure of a medieval Europe that was fragmented along the ecclesiastical Protestant and Catholic divisions of the Reformation (Eisenstein 148). This division and the cottage industry beginnings of the print era allowed for a relatively unregulated and thus relatively uncensored early development. As much of the Lutheran rhetoric of Protestantism arose from a general desire to expose a Catholic hypocrisy, one could infer that empirical traditions of science were due in some part to this environment. Eisenstein suggests that the Enlightenment period and the rise of the natural sciences owes much to the printing press (Eisenstein 259). She also suggests that this period of new ways of thinking about the world would not have been possible in a more autocratic environment.

In a different cultural context, the same technology might have been used for different ends (as was the case in China and Korea) or it might have been unwelcome and not been used at all (as was the case in many regions outside Europe where Western missionary presses were the first to be installed). (Eisenstein 273)

Most people view the age of the radio and television as the age of a mass media. As was mentioned, while print did unify opinion and promote the creation of the sciences, it is viewed by some as a generally more divisive form of communication than the mass media of radio and television. In this view, while print fosters the wide dispersion of ideas, it also allows for varied interpretations. For media culture expert Marshall McLuhan, the age of the printing press gave rise to individualism and worked to break society apart by allowing people to think in isolation. McLuhan says of printing, "The private, fixed point of view became possible and literacy conferred the power of detachment, non-involvement" (McLuhan 50). The age of electronic media on the other hand was a positive and unifying force and thus led McLuhan to conclude that society was returning to the tribal structure of prehistory societies, hence his famous "Global Village" analogy.

For others the unification afforded by mass media is seen as less positive and has led to abuses by totalitarian political regimes. Most notably George Orwell criticized the mass media for imposing a monolithic societal perspective that was sponsored by elites and thus tended to favor a certain privileged few. The well known example is his novel 1984 in which an oppressive "Big Brother" government monitored all aspects of people's private lives and controlled public opinion through complete surveillance and "newspeek" propaganda. During the cold war this perspective was itself cultivated by the Western mass media. The "Big Brother" figure was adopted as anti-communist and time has proven this to be an accurate depiction. At the same time there are some who consider this to be a simplistic and manipulative reading of Orwell. Professor of Communication Studies James Winter explains that the true reading of Orwell includes a more subtle criticism of the West as well as the obvious anti-totalitarian message. In his words,

In the meantime, another work by Orwell has gone largely unmentioned. In fact, it was the preface Orwell wrote to Animal Farm, which was excised from the original publication and only surfaced later with his original manuscript. Rather than writing about totalitarianism in Russia, in his preface Orwell was writing about voluntary literary censorship in Britain at the time, a topic which was much less fashionable. Orwell noted that, 'Unpopular ideas can be silenced, and inconvenient facts kept dark, without the need for any official ban.' This may be accomplished, he said, 'because of a general tacit agreement that 'it wouldn't do' to mention that particular fact.'" (Winter XXVi)

Another well known political activist and Professor of Linguistics Noam Chomsky has elaborated on this aspect of Orwellian social commentary. He explains that the mainstream opinion that is 'manufactured' by mass media is so pervasive that it extends all the way to the mechanics of language itself. Instead of blatant censorship Chomsky suggests a sophisticated corruption of thought makes dissidence seem simply ill-logical. In discussing this same phenomenon he writes,

Such statements lack cognitive meaning. They are imprecations, like shouting 'F*** You' in public; they can elicit only a stream of abuse, not a rational response. We see here the ultimate achievement of thought control, well beyond what Orwell imagined. Large parts of the language are simply determined to be devoid of meaning. It all makes good sense: In a Free Society, all must goose-step on command, or keep silent. Anything else is just too dangerous."(Chomsky 317)

To be clear, the role of mass media in society is, for the current time, primarily a social issue but one that cannot be divorced from technology. The discussion of the development of human communication beginning with the printing press is not an argument that attempts to provide technological causal explanations for social phenomenon. In fact there are strong arguments that relegate technological change as subject to an overarching rate of social change that is moving at its own pace. Brian Winston, BBC journalist and Professor at the University of Westminster, School of Communication, describes a model of communication where technology plays a more subordinate role. In "Media Technology and Society" he argues that prototypes for new technologies are not necessarily adopted just because they work. For prototype to become invention there must be a demonstrated societal usefulness. To support this idea he provides several examples including the early rejection of the telegraph by the British navy (Winston 7). Less convincingly he calls the "Information Revolution", "Largely an illusion, a rhetorical gambit and an expression of technological ignorance" (Winston 2). He may be right that change is not increasing at the rate that is implied by today's High Technology rhetoric, but the historical timeline of communication is too crowded with events in the second half of the 20th century to discount an acceleration of change caused by technology. John Seely Brown and Paul Deguid offer a similar argument in "The Social Life of Information". Yet where Winston seems to dismiss the Information Revolution as a myth that is unsubstantiated by facts (Winston 2), Brown and Deguid seem to accept the reality of an Information Revolution, but argue a kind of technological determinism where society is master, the tool its slave. They write,

The ends of information, after all, are human ends. The logic of information must ultimately be the logic of humanity. For all information's independence and extent, it is people, in their communities, organizations, and institutions, who ultimately decide what it all means and why it matters." (Brown and Deguid 18)

Whatever the position on the effect of mass media most agree that while the age of print began to build general consensus in the sciences, the age of mass media beginning with the telegraph blurred the lines of public opinion.

Returning then to the discussion of the differences of print and the era of electric communication, it is apparent that the printed word allowed for a more varied and less autocratic pool of ideas. Interestingly, the World Wide Web is a technology that seems to blend the two aspects of communication. With even greater ease than during the original printing revolution, a large number of people are able to publish their ideas because the machinery involved is far more attainable than was the older print machinery. Initially this fact led some to herald the new technology as a great equalizer and the new hope for true democracy. This may still prove to be the case, but the recent realization is that with the relative ease of publishing it has become increasingly difficult to be heard. The problem now is how to go about maintaining the varied pool of ideas while simultaneously creating information access that is useful. At this point Brown and Deguid's analysis is especially relevant. Here is a technology that at first offered the opportunity to overcome some of the traditional political issues in society. But as soon as this direction was perceived, the same new technological mechanism became weighted down by the same natural forces it arose from and has consequently begun to fold back upon itself. In this way momentous societal changes are usually recursive in that they must always return to serve the same albeit slightly different society. The question now is in what direction will the next evolution of the Web, the Semantic Web, take us? Is the hope for a multivariate voice to die with the Semantic Web? Will the tradition of an autocratic mass media once again drown the individual by providing the only means to navigate a sea of overwhelming data? The logic that develops to glean the meaning from the Semantic Web already seems to be developing unilaterally and this fact threatens to limit knowledge as much, if not more than it ever has.

Building the Semantic Web

Variation in nature as it pertains to DNA information is an important ingredient in adaptation and thus the survival of species. Returning to Marvin Minsky's theory of the mind and the idea implied by his theory of frames, the strength of human intelligence lies in the ability to connect and relate information. Minsky cites Thomas S. Kuhn's model of scientific revolutions to show that social processes very closely resemble individual mental processes. This seems an entirely logical conclusion since a society is in fact a conglomeration of individual minds. Variation then, in this model correlates to the notion of informed decisions. Having more knowledge available presumably allows for stronger inferences. In this sense individual ignorance can be seen on the societal level as similar to periods of scientific stagnation where perspectives that challenged the status quo were discouraged or suppressed. The contrary concepts of tolerance and free speech can again be seen at the level of the individual mind as the ability to accept ambiguity and a more general open-mindedness. In his 1984 book "Society of the Mind", Minsky includes the concept of ambiguity and describes the important role it plays in human thought. Essentially, ambiguity is tolerated because meaning and understanding are context relative. What seems logical at one moment in time may change given future circumstances (Minsky 207). This is true at the most basic level of mental processing. Thus at both the level of society and the individual mind the success of the human species is due to an ability to make inferences and apply intuitive logic. Even if some of the information stored in record is paradoxical and seems at odds at a given point in time, this contradiction may not always exist.

The Semantic Web model for Knowledge Representation is based on similar principles. Semantic Web documents correspond to frames and frame systems and schemata or web agents are employed to manipulate the data stored in these documents. Thus the design of the Semantic Web, like the human mind, is able to store potentially ambiguous information and allows a varied group of agents access to this information. However, like a human mind that learns only a limited range of information, the Semantic Web will only be successful when there is a diverse body of information stored in it. While the model for the Semantic Web is sound, the sources do not appear sufficiently varied. At the current time the majority of W3C members are comprised of industrial interests. There are a few members such as the Society of Technical Communicators (STC) and the Web Accessibility Initiative (WAI), but even if these organizations increase their efforts, they still remain greatly outnumbered.

Much of this potential unilateralism stems from the fact that relative to the ease of publishing on the World Wide Web, it is difficult to build ontologies. But even building ontologies will not guarantee that a voice can be heard. Constructing the logic that is able to parse the Semantic Web at the level of human intuition is an enormous task that will require considerable resources. As Noam Chomsky identified the politics embedded in a given discourse and way of thinking, it is conceivable that this same narrow-mindedness become manifest in our digital record. Either by coding logic that leaves certain thoughts out, or because a voice cannot speak the language of a given discourse, a Semantic Web could become like other mass communication media. In the case of the Semantic Web, XML and the Resource Definition Framework is this language. To further explore this idea consider the ambiguity below coded in RDF. A familiarity with XML is not required as the words in bold alone illustrate the idea. Both definitions are possible, but both will only exist if each in turn is created.

Example 1

Example 2

This example is of course over simplified for clarity and the issue need not be so politically charged. Still, the dangers must be taken seriously because in the extreme sense a myopic definition that ensures public safety today can be the cause of mistakes in the future. Here we can see the potential for the Orwellian threat in a Semantic Web. As a less radical example consider an entire group of such descriptions organized in an ontology. There are more subtle ways that information stored from a limited perspective could fail to provide the desired inferences for a given scenario. It is entirely possible that an ontology that represents knowledge that seems counter-productive at one point may provide insights or even thought patterns that are the basis of future "Transitions". The flexibility of the Semantic Web is that any definition is possible. The problem is that in order to be useful, they must have been created in the first place.

Tim Berners-Lee addresses these concerns in his book "Weaving the Web" and through the policy of the W3C. For Berners-Lee the solution is to keep things simple and transparent and to ensure trust through a "web of trust" built upon encryption technology. Transparency means keeping the mechanisms visible by making the technology public via the W3C. This is a good first step and if the technology truly remains "simple" this may work, but as the complexity increases transparency becomes a relative term. Simply because everyone can look at how the Semantic Web works, it is unlikely that everyone will understand what is happening. Here again it is more likely that the technology will fall back again on the basics of social communication. People trust something or somebody they know not because they understand everything about the thing or person, but because they have assigned trust based on a track record. For similar reasons the success of the large portal sites and search engines is based largely on reputation and ethos. For agent technology to be a success, not only will there need to be agents created by numerous people with varied interests, but the concept of author agents must exist to lend credibility through a trusted name. In real life people place their trust in politicians, doctors or specialists even when they do not understand every aspect of a situation. The trust that titles and affiliations provide is an important part of communication because our society of specialization means people have to depend on each other. One would hope that as the Semantic Web evolves that the brokering of agents will give rise to a new level of cooperation in society. The art of human negotiation is highly complex and machines are very far from this kind of sophistication, but the fact that war and oppression have always existed demands an improved form of mediation. If we could equate the sum of human knowledge and experience, we might be able to solve many of our problems, but every missing piece will reduce the integrity of the whole.

Conclusion

The Semantic Web is emerging as one of the most realistic means of getting the most out of the increasingly large amount of information in society. As a system that is modeled after the human mind, it has great potential to answer questions about how we think and how we live, but with this potential comes great responsibility. In this sense it is not unlike raising a child and we must make an effort to give it a broad basis of knowledge. The more we can teach it about the world, the more useful it will be to us. As Berners-Lee himself says of the World Wide Web, "It must also transcend levels, because creative people are always crossing boundaries. That is how we solve problems and innovate" (Berners-Lee 164). Crossing boundaries implies that more than one perspective is represented and this logic must also extend to the Semantic Web. We have seen the rate of change and unexpected directions brought about by the World Wide Web. Some of the same phenomena may act to grow the Semantic Web in the same natural and unpredictable way that saw the explosion of the World Wide Web. Looking back at the history of human communication there is no doubt that each new major change in human communication technology was followed by a fundamental change in society. These changes were due to both the nature of the technology and the social conditions existing at the time of invention. The age of the printing press gave rise to a flurry of new ideas, whereas the advent of mass communication seemed to solidify a limited number of positions. What is needed now to ensure the Semantic Web's positive effect on society is an effort to create a broad base of knowledge and understanding, and the means to consider this knowledge from varied perspectives.

Works Cited

Berners-Lee, Tim, with Mark Fiscetti. Weaving the Web. New York: HarperCollins, 1999

Berners-Lee, Tim, James Hendler, and Ora Lassila. "The Semantic Web." Scientific American (May, 2001). 10 Dec. 2002 <http://www.scientificamerican.com/article.cfm?articleID=00048144-10D2-1C70-84A9809EC588EF21&pageNumber=1&catID=2>.

Brown, John Seely, and Paul Duguid. The Social Life of Information. Boston, MA Harvard Business School Press, 2000

Chomsky, Noam. Deterring Democracy. New York: Hill and Wang, 1991

Dreyfus, Hubert. "From Micro-Worlds to Knowledge Representation: AI at an Impasse." Mind Design II. Ed. John Haugeland. Cambridge: MIT Press, 1997. 143-182.

Eisenstein, Elizabeth. The Printing Revolution in Early Modern Europe. Cambridge Harvard Business School Press, 2000

Kurzweil, Ray. The Age of Spiritual Machines. New York: Viking, 1999

McLuhan, Marshall, and Quentin Fiore. The Medium is the Massage. Singapore Renewed, 1967, 1996

Minsky, Marvin. "A Framework for Representing Knowledge." Mind Design II. Ed. John Haugeland. Cambridge: MIT Press, 1997. 111-142.

Minksy, Marvin. The Society of Mind. New York Simon and Schuster, 1985

Nielsen, Jacob. Designing Web Usability. Indianapolis, IA New Riders, 1999

Searle, John. "Minds, Brains, and Programs." Mind Design II. Ed. John Haugeland. Cambridge: MIT Press, 1997. 183-204.

Winston, Brian. Media Technology and Society A History: From the Telegraph to the Internet. London and New York: Routledge, 1998

Winter, James. Media Think. Montreal, New York, London Black Rose Books, 2002