Chapter 1: A History of Information Highways and Byways

Modern Networks
pp. 7-13

Of what relevance are these ancient networks to our modern computer networks? From the fire signals to the Incan chasqui network, these ancient systems foreshadowed the way computer networks and the Internet would work. To illustrate this, here is a brief layperson's description of the creation of the computer network at New York University (NYU). (Although each campus faced different networking challenges, this example should illustrate the kind of work and thought that went into the network at your school.)

NYU is located in the heart of Greenwich Village in an urban campus that loosely surrounds Washington Square Park and includes buildings scattered across New York City. In the early 1980s, the challenge was essentially to link these buildings together. Among the different solutions that were proposed was using Ethernet. However, while Ethernet worked well over short distances--such as within a building's local area network--it had strict limitations over distances greater than three hundred feet. As Tim O'Connor of the Academic Computing Center recalls:

Ethernet . . . can support a cable of 300 feet long with no more than 30 devices attached; to do more of either, you need to add a repeater, which is like an amplifier; it would again allow for another 300 feet and another 30 devices. Then you hit another limit: the specification stated your limits are only two repeaters with three pieces of [Ethernet] on any segment of cable; for anything beyond that, you need a bridge, which then starts the count all over again. . . . In other words, Ethernet was perfect in a lab . . . or you could run a spine of Ethernet up the length of a building, and attach each floor to it. . . . [However] it was not realistic to cover Washington Square and beyond!22

Like Cyrus of Persia's careful experiments with the exact placement of rest stops in the courier relay system, the laying of the network infrastructure at NYU faced the physical limitations of just how much distance a signal could travel through a wire before it degraded. The only feasible solution lay in using coaxial cable (broadband), the kind that carries cable television signals. Broadband offered a number of advantages over Ethernet, including the fact that it could carry signals over a greater distance. It was also inexpensive. A simple network system, broadband operated in a linear fashion. The network originated at the "head-end" and was expandable in sections, resembling a segmented snake in which "everything emanated from the 'head-end'--the hardware that kept the signal on the network alive."23 Ethernet is simply cable that sits there whether in use or not. Broadband, on the other hand, had the added advantage of being able to be expanded as the need arose. In fact, by the end of the 1980s, as the electronic card catalog was being installed in NYU's Bobst Library, the network had become multilayered, with newer network technology sharing resources with older Ethernet and broadband network lines.

Without being aware of the underlying mechanics, people in Bobst were travelling across our Thinnet [Ethernet] which traversed each floor horizontally, along the traditional thick Ethernet, which ran up the spine of the building, through a bridge to NYU's broadband network, and into a series of devices in [NYU's main computing building], which allowed access to NYU's academic systems and to systems beyond NYU.24

However, broadband was limited; if a section of it went down, all information traveling through the network would stop at that point. Network services for the remainder of the broadband network would be suspended as well. With the introduction of fiber-optic cable, NYU's system was upgraded. In a description that echoes the topography of the Roman and Incan road networks, NYU's new network was "logically (and almost literally) a double ring encircling the campus. The idea [is] that even if one piece of fiber-optic was damaged, the network could work around it."25 The double-ring loosely circles Washington Square Park, while hubs on the ring connect the outlying NYU buildings around Manhattan via fiber-optic cable leased from the telephone company.

While the ancient signal and courier relays laid out the topography for modern networks, it was the invention of the telegraph in 1792 that revolutionized signal networks. Crude telegraph schemes existed before this period, a testament to the tenacity of our ancestors to work with the technology at hand. However, while signal fires and couriers were effective, Herodotus noted that they were crude and insecure means of communication. He observed that the subtleties of "citizens having changed sides or having been guilty of treachery . . . cannot all be foreseen--and it is chiefly unexpected occurrences which require instant consideration and help."26 Anyone who struggled with early clunky word processing packages like WordStar or braved UNIX-based e-mail programs will appreciate the efforts of Polybius in the second century B.C. to create a more efficient system using the existing fire beacon "technology." Polybius devised a code using the twenty-four characters of the Greek alphabet and a telegraph that consisted of two large screens that hid five torches each. Depending on the letter, the signaler raised different numbers of torches behind the two screens. However, this line-of-sight communication was limited until the invention of the telescope in the seventeenth century renewed interest in building communications systems.27

It should be noted that there are line-of-sight network solutions operating today. Wireless network technology has made it possible to cross otherwise impossible distances. Presently, a variation of line-of-sight using wireless networking technology is being used at the Villa La Pietra, a fifteenth-century estate in northern Florence that was given as a gift to NYU in the mid-1990s. Due to the building's status as a historical landmark, tunneling and laying cable as well as stringing wires were out of the question. The alternative solution was wireless technology. Wireless technology, similar to that used in 900 mHz cordless telephones, can carry a signal that can travel throughout a building and is not dependent on direct line-of-sight. Infrared, another common wireless technology, can be employed outside over short distances. However, it is heavily dependent on line-of-sight and can be thrown off by inclement weather, including fog. At the Villa La Pietra, establishing the wireless connection between other buildings on the estate relied on line-of-sight for obtaining the best possible signal from the network inside La Pietra. Connecting the villa with a smaller building used to house NYU students called for the installation of wireless antennas on the roofs of both buildings. These had to be manually adjusted to obtain the best possible line-of-sight over the five hundred yards separating the two buildings.

Claude Chappe (1763-1805) invented the "optical telegraph," a visual telegraph with signals based on the different positions produced by a system of cross-arms and pulleys. The telegraph resembled a windmill, with two lateral arms that rotated freely around a center. The signaling method was similar to that later used on the railroads. The different positions of the arms could transmit nearly 8,500 words from a general vocabulary of ninety-two pages, each containing ninety-two words. Only two signals were required for a single word--the page of the vocabulary and the number of the word. Telegraph towers were set at ten kilometer intervals--within range of telescopes and field glasses.

In March 1792, Chappe showed his invention to the French Legislative Assembly, which adopted it officially. The first connections were made between Lille and Paris, and a line of fifteen stations were operational by August 1794. The first telegram sent with this telegraph announced the victory of the French over the Austrians at Condé-sur-Escaut on November 30, 1794. At this time, the political situation in France was unstable and security was weak. These conditions required a communications system that could operate rapidly, efficiently and secretly. Messengers or couriers on horseback could not meet these criteria. Thus, when Chappe unveiled the Optical Telegraph, the French Assembly was in a particularly receptive frame of mind.28 The network rapidly grew over the next decades into a system that covered approximately three thousand miles, included 556 stations and connected twenty-nine French cities. Many other European states also installed the Chappe telegraph system in their territories.29

Though supplanted in public memory by the electric telegraph, the Chappe telegraph was used extensively in both Europe and the United States. In the United States optical telegraph lines ran from Boston to Martha's Vineyard and connected Staten Island to Manhattan. One line ran from Philadelphia to New York so that the stockbroker who ran it could get word immediately of fluctuations in stock prices. Another optical telegraph operated in San Francisco from 1849 to 1853. Telegraph Hill in San Francisco was one of three optical telegraph stations in that city.30

However, the optical telegraph was slow (a short message could take up to fifteen minutes) and could not be used at night or during periods of poor visibility. Sending messages over long distances required a series of towers, each of which had to receive and then retransmit the message, increasing the chance of errors being introduced that could distort the original message. Also, though cryptic, the telegraph messages were viewable by anyone on the ground and thus were not secure--in some cases people living in the vicinity of a telegraph tower were actually able to learn the signals over time.

However, other developments in the nineteenth century would render the optical telegraph obsolete and have an impact on the development of computer networks. One of the earliest experiments with electrical telegraphs took place in the early nineteenth century. In 1795, Francisco Salva decided to build an electrical telegraph between Barcelona and Mataro. He planned to send his messages by illuminating letters of tinfoil with an electrical spark.31 Alessandro Volta's invention of the battery in 1800 furnished a new source of electricity, better adapted for the telegraph, and Salva was apparently the first to recognize this. In 1804, he successfully developed a battery-powered electric telegraph system by which he could transmit messages over one kilometer, with each letter being carried on a separate wire.32 In 1816, Francis Ronalds, also fascinated by electricity, developed an electric telegraph (using static high-voltage electricity) and succeeded in sending messages through eight miles of iron wire suspended above his garden in London. However, it was Samuel Morse who refined the electric telegraph.

In 1831, Morse, a painter, was forced to cut a trip to Italy short when a revolution broke out in the papal states. On the return trip to the United States he conceived of the idea of a "telegraph" based on electromagnetism. He spent the next six years in his studio at NYU working on his telegraph. In 1837, Morse demonstrated the electric telegraph by sending a signal over 1,700 feet of wire in a room at NYU. This was followed by the development of Morse Code in 1838. On May 24, 1844, Morse sent his first telegraph--containing the words "What hath God wrought!"--from the U.S. Capitol Building in Washington, D.C., to the B & O Railroad Depot in Baltimore, Maryland. Of this invention American author Nathaniel Hawthorne would write: "Is it a fact--or have I dreamt it--that, by means of electricity, the world of matter has become a great nerve, vibrating thousands of miles in a breathless point of time? Rather, the round globe is a vast head, a brain, instinct with intelligence!"33

However, this new discovery was also met with a degree of skepticism. When Morse demonstrated the telegraph before members of Congress in 1842, one senator recalled: "I watched his countenance closely, to see if he was not deranged . . . and I was assured by other Senators after we left the room that they had no confidence in it."34 In 1845, Cave Johnson, the Postmaster General, declared in his annual report that the "telegraph business will never be profitable."35 However, others saw the electric telegraph with an entrepreneurial eye. For example, in 1886 Richard Sears, a telegraph operator and railroad station manager, started experimenting with mail order by selling watches via telegraph--an experiment that eventually led to the formation of Sears and Roebuck. The Associated Press got its start as an alliance of Morse telegraph services and operators dedicated to transmitting news dispatches. The American Civil War was one of the first full-scale demonstrations of the efficiency of this system for transmitting troop deployments and military intelligence.36 By 1851, an international telegraph network was launched with the laying of a cable across the English Channel.37 In 1866, just twenty-two years after the completion of the first telegraph line, the laying of a transatlantic cable connecting the United States with Europe marked the beginning of a new era in telecommunications. In the early years of this international telegraph network, eight words took a full minute to transmit at the cost of one hundred dollars.38

The nineteenth-century revolution in communications technology reached full fruition with the invention of the telephone in 1876 by Alexander Graham Bell. By the next year, the Bell Telephone Company had grown to a network consisting of a thousand phones. By 1880, there were fifty thousand telephone lines in the United States. By 1930, the telephone network had outgrown the telegraph network. (Ninety-three years after Bell's invention, the ARPANET sprang to life in the space occupied by telephone traffic traveling along the wires that first made telegraph possible.)

The development of telegraphs and networks is significant for understanding the Internet because it demonstrates the relentless push toward more speed, more capacity, more raw volume, more "consumers." Each of these inventions was designed for a single purpose but mutated over time to fill other needs. For example, Bell first envisioned that the telephone would be used to pipe music to homes. However, the telegraph and, eventually, the telephone developed into point-to-point communications. Even television, a broadcast medium, has followed this trajectory of diverging from its original uses and modes of delivery. While most television is broadcast in specific wavelengths, cable television literally strings together buildings and neighborhoods. Ironically, we are beginning to come full circle as some cable companies are fighting to provide both voice and data service on their physical networks.

------------