Claudio Gutierrez

Internet, one of the memes defining the new era of globalization, is an invention both technological and social. In its current form, it consists of a network of networks, resulting from the interconnection of a multiplicity of computer networks of arbitrary design. Its common factor, which was the foundation of the first network and underlies all current sub-networks, is packet switching technology. Such technology contrast sharply with telephone switching inasmuch as it does not maintain process identity for the communication lapse. In contrast, it divides information in small chunks sent consecutively but independently, duly labeled with an address and an ordinal number. Such packets take the most promising route from each visited network until reaching their destination, at which point they get reassembled according to their ordinal numbers. Such technology realizes the most preposterous fantasy of a family captured in a traffic jam, namely to magically transform their van in, let us say, five different motorcycles (one for each member of the family), squeeze them through the jam and, having reached home, let the van reconstruct itself to its former solid structure. As to the social aspects of this impressive meme –surely one of the strongest candidate to be declared some day the paramount meme of the 20^th century– they started to be explored at the beginning of the 60s, in a series of research memos on a "galactic network," by J. C. R. Licklider, an MIT professor. Interestingly enough, to this social lucubration had preceded the concoction of the technical means to practically realize such universal "network of networks."

In fact, Leonard Kleinrock, also from MIT, had published in July 1961 the first article on packet switching. His colleague, Lawrence G. Roberts, working together with Thomas Merrill, succeeded the following year at connecting a MIT computer with another one in California, experiment which demonstrated the feasibility of two computers working together at a long distance, exchanging information according to their needs. This experiment demonstrated also that the system of telephone circuits was totally inadequate for the task. By the end of 1966, Roberts joined DARPA (Defense Advanced Research Project Agency) where he swiftly developed a plan for the construction of ARPANET (Advanced Research Projects Agency Network), the working precursor of Internet. It consisted of an ample WAN (Wide-Area Network) called upon to serve as tryout field for the new networking technologies. It coordinated several universities and research centers, without any intervention of its sponsor, the U. S. Department of Defense.

The plan prepared by Roberts was published in 1967. In the conference where this paper was discussed, it came to light that MIT research had been going on in a parallel fashion with two other independent research endeavors, conducted in the United Kingdom by RAND (since 1962) and NPL (since 1964) corporations, without any of the groups knowing of the activities of the others. Starting from there, the three groups joined their efforts. With DARPA support, the development of the first electronic components specific to the project, branded as Interface Message Processors (IMP), was open to a public bid. The bid was won by Bolt Beranek and Newman (BBN) in Boston. Meanwhile, other groups joined the fray, devoting themselves to develop and optimize the topology and economics of the future network.

Given Kleinrock's germinal work on packet switching, his UCLA group was selected to constitute itself as the first ARPANET node. In September 1969 the first IMP was installed, leaving thus connected the first host computer of the pioneer network. The second one was established at SRI (Stanford Research Institute) and, a month later, the first message was sent from Kleinrock's lab to SRI. Two new nodes were added right away: the University of California's in Santa Barbara and the University of Utah. Thus, by 1969's end, ARPANET had begun its existence. Since the very first moment it served as a research instrument, in relation to its own development but also to all kinds of scientific projects. In October 1972, at the International Computer Communication Conference (ICCC), BBN's Bob Kahn presented a large convincing demonstration of the new technology. Also in 1972, the initial hottest application of ARPANET, electronic mail, was introduced. In March, BBN's Ray Tomlinson wrote the basic "send" and "receive" software, while Roberts –in July– expanded the utility to selectively read, send and reply messages (Leiner, 2000). This represented the takeoff of a tool, e-mail, which would constitute the most important application of Internet for at least two decades. In 1979 and 1980, during my first visits to MIT, The University of Edinburgh, SRI, Stanford University, and Carnegie Mellon University, I was able to observe the practitioners of artificial intelligence (AI) exchanging information among them, mutually sharing technical memos and, in general, enhancing reciprocally their research work. Such exchanges were bound to characterize the technical communication among computer-science researchers during the next decade, as a prelude for a new era of global academic interaction. A few months later, as I became affiliated to the University of Delaware –one of the immediately-following new nodes in the network–, I had the great pleasure of acquiring my own "ARPA address" which still flaunts some of my old visiting cards.

Internet rests on the principle of open architecture which leaves to each individual network connected to it the freedom of doing its internal transmission by the ways and means considered more convenient by their implementers. It only imposes a number of protocols which enable the smooth exchange of messages between each pair of independent networks.³⁵⁶ The technical sense of the word "protocol" is basically the same that the word has in the context of diplomatic encounters. Indeed, between two independent nations –with their own usages, customs and values– the exchange of ambassadors poses serious problems. How can we be sure that something done during the presentation of credentials will not offend the other party? The international tradition of ritualized procedures –approximately the same for all countries– gives relative assurance of not screwing it all up at the very beginning. This original sense of the word is immediately applicable to networking technology, even with much more accuracy and rigor. Networking protocols establish the steps computers must go through in order to open, sustain or close a communication session; if they are not complying with, communication is not established. The internal freedom of each member network has stimulated technological progress, since each network has been able to experiment with separate designs. The idea of open architecture was first introduced by Robert E. Kahn in 1972; he created the protocol which eventually came to be called "Transmission Control Protocol/Internet Protocol," the today famous TCP/IP with which we all configure our cherished personal computers to be connected to the web. This great protocol, a harbinger of the 21^th century, is based on the following four fundamental rules:

• Autonomy: Each individual network must be self-sufficient, so that no internal change is necessary to connect it to Internet.
• Good will: If a packet does not reach its final destination, it will get retransmitted a moment later.
• Privacy: Black boxes are used to interconnect the networks, the famous gateways or routers, which each supplier specifies at will. These black boxes do neither conserve any information about the packets that go through them nor worry about reconstructing messages in case of failure. As a consequence, their design is extremely simple.
• Autarchy: There exist neither global control of operations nor central command of the network. Things work because each router complies with simplicity and efficiency its mission of getting each packet a little closer to its final destination (Leiner, 2000).

These simple principles, which make today possible the communication –without any central intervention– among millions of computers distributed all around the world, operated by people of the most diverse cultures, could well serve globally as a model for other societal relationships. Their translation to a political context is trivial and could contribute to a re-writing of the libertarian ideals, appropriate for the 21^th century and beyond: respect to individual autonomy and to grass-root groups, trust on individual initiatives, no State intervention in private affairs, and reduction to a minimum of the governmental apparatus. The success of those principles in electronic communication should be used as incentive to create and maintain similar political protocols, assuring efficient and just governance in today's multicultural world. Such political protocols could encompass international trade, migration of persons, currency circulation, law enactment, justice administration, relationships between individuals and corporations as well as between public and private sectors.

The original motivation of ARPANET was the sharing of scarce resources by scientific researchers. The intention was to exchange files (mainly research memos) among colleagues, and to operate remotely very expensive supercomputers installed in a few scientific centers. Nevertheless, as noted earlier, e-mail rapidly became a major Internet application. Electronic mail swiftly changed the nature of collaboration among scientists, recently also among business people, and finally today the regular interpersonal communication in most levels of society. An excellent illustration of a fundamental characteristic of technology, inherited from an identical property of organic life, i.e. the capacity to find important applications not anticipated at its original invention. This process was facilitated on account of the creators of Internet being utterly familiar with computers, general-purpose infrastructures by their very definition.

Another confirmation of the intrinsic flexibility and fertility of the Internet meme has been the even more recent emergence of the World Wide Web. This happened at CERN, the European Organization for Nuclear Research, in Switzerland. In March 1989, the British physicist Tim Berners-Lee wrote a proposal on how information could be transferred easily over the Internet using hypertext technology, the now familiar system of navigating through web pages by means of appropriate links. The first web address was "http://cern.ch," still today the entry point to the founders' site. This great feat was joined by a second equally brilliant one, some years later: the creation of automatic searching mechanisms over the growing mass of web pages, based on simple –yet extremely powerful– algorithms; the most imposing of those searchers has been the one mounted by GOOGLE, with incredible popularity and financial success. These two inventions together have made possible the greatest explosion of written knowledge in all human history, selectively reachable in a matter of seconds, at a cost rapidly tending to zero, by –at least in principle– every member of the human species.