The Internet as we know it today is a worldwide network of interconnected computers facilitating electronic communication, be it human or electronic in origin. The Internet has spread across the face of the globe and, like the universe, is expanding all the time. However, "Internet" was not a household word before some time in the mid-1990's. Before that time, the Internet was largely governmental and academic in nature. This essay aims to explore the origins and development of the Internet in addition to highlighting the ways it has enabled and influenced communication over its lifetime.
Like most all innovations, the Internet arose to meet a perceived need. The Advanced Research Projects Agency (ARPA) was formed in 1957. One of the factors leading to the founding of ARPA was the Soviet Launch of the first artificial satellite, Sputnik. (Hauben) ARPA was founded to in order to conduct military research with the goal of keeping the US more technologically advanced than her enemies. In 1962, ARPA hired J.C.R. Licklider to head its Behavioral Sciences and Command and Control departments. (Griffin) Licklider had an academic background in physics, math, psychology, psychoacoustics, and a professional background in computers. Despite his rather informal knowledge of computers, Licklider had several innovative ideas regarding the improvement of human use of computers. He was one of the first to value computers for their potential to assist human thought and act as a communications medium rather than as algebraic number crunching machines. (Griffin) Licklider used his position at ARPA to facilitate research into better uses of computers; he dubbed his research groups the "Intergalactic Computer Network". Seeing the need for the separate research groups to be able to share information, computing resources, and build upon one another's work, in 1963 Licklider called for the formation of a computer network to enable the ARPA contractors to collaborate with one another. According to Larry Roberts, the "father of the ARPANET", Licklider had a vision of a computer network in which "everybody could use computers anywhere and get at data anywhere in the world." (Griffin)
Licklider never saw his vision to fruition; he left the following year. Licklider's position at ARPA was filled by Robert Taylor. Seeing the same duplication of effort and computing resources between different ARPA funded projects, Taylor realized that the network concept Licklider had proposed could be a viable solution. (Griffin) After receiving approval from ARPA to develop a network, Taylor recruited Larry Roberts who had previously networked the TX-2 at MIT's Lincoln Lab with the Q-32 at System Development Corporation in Santa Monica, CA in 1965 via a dedicated telephone line. (Computer; Zakon)
In his initial presentation to ARPA, Roberts proposed directly connecting all ARPA computers via dial-up telephone lines, an idea that did not meet with much favor among the other ARPA scientists. (Griffin) The scientists did not want to sacrifice their computing power to internetwork, nor were they excited about trying to interface the many different systems with one another. One of the scientists, Wes Clark, suggested to Roberts that small dedicated computers be used to interface the larger computers together. (Griffin) In addition to adopting Clark's idea, Roberts decided that the network would be packet-switched instead of circuit-switched. In a circuit-switched network, each host has a dedicated connection to each remote host with which it is communicating. A packet switched network, on the other hand, does not necessarily have dedicated connection between two communicating hosts, but rather, breaks up messages between the two hosts into small chunks that are then passed from the source to the destination as resources become available. The idea of packet switching was revolutionary at the time and developed on the research of Paul Baran of the Rand Institute, Leonard Kleinrock at MIT, and Donald Davies and Roger Scantlebury of NPL. (Kurose and Ross 58) Curiously enough, all of the aforementioned researchers conducted their research in parallel without knowledge of the research of the others.
With a general plan for the organization of the ARPANET, a contract was awarded to Bolt Beranek and Newman (BBN) to build the small computers that would act as interfaces for the larger computers on the ARPANET. These small computers were called Interface Message Processors (IMPs). The first IMP was installed at UCLA on Labor Day 1969 under the supervision of Kleinrock. (Kurose and Ross 59) Installations at Stanford Research Institute (SRI), UC Santa Barbara (UCSB), and the University of Utah followed shortly. Thus, in 1969, the ARPANET consisted of four interconnected nodes. The first host-to-host protocol for the ARPANET, Network-Control Protocol (NCP), was finished in 1970. (Kurose and Ross 59) By 1972 the ARPANET had grown to approximately 15 nodes and had seen the development of the first email program.
The success of the ARPANET spawned the development of other networking paradigms. This in turn resulted in the growth of a number of small networks. As the ARPANET continued to grow, the need to connect already established smaller networks to the ARPANET arose (i.e. internetworking). NCP was not up to this task as it was only designed for passing messages from IMP to IMP, thus, Robert Kahn developed the TCP protocol. The TCP protocol differed from NCP in that it provided for end-to-end reliability over (possibly) unreliable networks. (Leiner et al.) This allowed other networks such as packet radio and satellite networks to be internetworked with the ARPANET. Soon the need for an unreliable transport protocol was realized and the forwarding functions were separated from TCP resulting in the IP protocol. The ARAPANET now had a reliable delivery protocol (TCP) and an un-reliable delivery protocol (UDP) which both ran on top of a single forwarding protocol (IP).
The proliferation of small networks had resulted in the creation of several large networks besides the ARPANET including MFENet, HEPNet, SPAN, CSNET, BITNET, JANET etc. (Leiner et al.) In 1986 the National Science Foundation created NSFNET in order to provide access to its sponsored computing centers to the entire academic community. (Kurose and Ross 62; Leiner et al.) To ensure interoperability, NSFNET dictated that all member networks must speak the TCP/IP protocol. Together NSFNET and ARPANET issued RFC 985 - "Requirements for Internet Gateways". The cooperation of NSFNET and ARPANET insured that the two networks would be able to interoperate.
In contrast to the ARPANET, the NSFNET was organized in a single backbone layout. Regional Points of Presence (PoPs) provided connectivity to small localized regions. These PoPs were then linked together over the NSFNET backbone. Since the NSFNET was a government-funded research-oriented network, commercial traffic was prohibited over the backbone. However, the NSF encouraged regional points of presence to provided local data services to commercial entities in their areas. This policy had two results. First, it allowed the points of presence to amortize their infrastructure costs, taking advantage of higher volume to lower connection costs. Secondly, it encouraged the development of private nation-wide data carriers to carry commercial data between PoPs. (Leiner et al.)
Even as the ARPANET was decommissioned in 1989 (Leiner et al.), the NSFNET continued to grow in size and speed. The NSFNET spanned the nation, and even interconnected with national networks on other continents. The backbone grew from 6 nodes with 56 kbps links to 21 nodes with multiple 45 Mbps links and there were over 50,000 interconnected networks. (Leiner et al.) Just as the ARPANET had seen the emergence of email, the NSFNET saw the emergence of IRC and the World Wide Web.
In 1991 the NSFNET removed its prohibition on commercial traffic from the network backbone, spawning the age of the commercial Internet. As the private national networks had matured, the NSFNET backbone was finally decommissioned in 1995, with private networks serving entirely as the Internet's backbones. (Kurose and Ross 62) With the commercialization and privatization of the Internet, there was an sharp rise in commercial use. The network itself moved to a multiple backbone model with private networks exchanging traffic at Network Access Points (NAPs). Private ISPs began marketing dial-up connections for home use. The World Wide Web gave rise to online banking and shopping as well as online communities targeted at home users. The Internet continues to grow both in size and speed with broadband access being made available to more and more end users. As if that weren't enough, research still continues to try to find ways to offer faster, better, more diverse, and cheaper service offerings to Internet users.
As envisioned by Licklider, the Internet has changed the way in which people communicate. With IRC, web chat, and instant messaging, the Internet provides a communication medium that had never existed before in all of history: synchronous written communication. In addition to providing a rich playground for social theorists, the synchronous written communication of the Internet allows geographically isolated individuals, who in all probability have never, nor will ever, meet one another face to face, to share ideas and emotions in real time. In effect, the Internet has shrunk the world, advancing globalization by leaps and bounds. Never before has an individual had the ability to interact with people of differing cultural backgrounds in distant geographical locations with such ease.
Coupled with services like email, the communication channels afforded by the Internet now allows teams to do distributed work on projects in ways never before possible. The combination of synchronous written communication, asynchronous written communication, and the ability to share resources and data allows team members that are geographically separated to coordinate their efforts and work together in a much more effective manner. One outgrowth of this ability to coordinate large decentralized groups is the open source movement.
More than being enabled by the technologies hosted by the Internet, the open source movement finds its root in a concept that the Internet has fostered throughout its lifetime: free exchange of information. From its beginnings as a research and academic network, much information has been exchanged and published on the Internet for free. A shining example is that of the very standards that define how the Internet is to work, RFCs. Since RFC 001, the standard documents for all of the major protocols have been available for free. This is one reason for the rapid advancement of the Internet. Not only was every member provided with a new communication medium, but they also had access to all the documentation in order to understand the Internet as it was and to take it in new directions.
Another freedom that the Internet has encouraged is freedom of the press. The global nature of the Internet and the cost-effectiveness of electronic publication have given rise to independent journalism that would not be able to compete against mainstream journalism in a traditional setting. Email lists, Newsgroups, Blogs, Websites, and IRC are enabling primary sources to communicate their stories directly. These technologies also provide the Internet friendly individual with an editorial outlet without censorship. There are also projects under development to protect the freedom of the press online.
An astute man once observed, "Man's mind, once stretched by a new idea, never regains its original dimensions." (Holmes) I submit that the same is true with regards to the world and the Internet. Not only has the Internet has changed the way computers are used, provided a medium for more effective research on scientific and technical topics, but it has changed the way people communicate. It has changed the way information is disseminated and consumed. Individuals with a personal computer and an Internet connection have the potential ability to access more information than most individuals who have lived throughout history before them.