All hail the electronic grapevine

Whitfield Diffie looks at how we learnt to move data without moving matter If one thing distinguishes the past century and a half from all previous human history, it is communication. Some of this is physical: the railway, the steamship, the automobile and the airplane have transformed travel. More of it is electronic. The telegraph, the telephone, radio, television, fax machines, cellphones, communication satellites, optical fibres and the internet have made people,information and ideas from around the globe instantly accessible. The physical and the electronic are far from independent. Electronic communication might conceivably exist without physical communication, but not the other way around. In order to move today's passengers around at 500 miles an hour, an airline cannot do without electronic communications at the speed of light.

If something has wrought such great changes in our lives we shall surely want to know its history, and this is what Laszlo Solymar, a professor of engineering science at Oxford University, offers. The difficulty with presenting the history of communications is that it is not one history but several. Most directly, it is a history of some science, a lot of engineering and even more business. Indirectly, it is a history of the impact of communications on society, business, politics, diplomacy and warfare. Finally, it is a history of the personalities who brought about the developments it records. The difficulty is compounded by the fact that Solymar sets no technical prerequisites for readers and attempts to teach them everything from how messages are represented by signals to the principles of the triode and the super-heterodyne receiver. The sophistication required to absorb this material grows dramatically from the beginning of the book to the end, without building on itself in a way that justifies the growth.

Solymar limits his notion of communications to "telecommunications", that is, ways of moving information without moving matter. This excludes such important communication mechanisms as postal services, magazines and couriers. It is a decision that may have helped keep the task of exposition to manageable proportions, but it may also have cost the author the opportunity to discover many social, and a few technical, parallels between the physical and the electronic. Some history of transport does creep in. The railways were big backers of the electric telegraph because, like the airlines to follow, they needed fast communications to manage their operations. Whatever techniques served to manage shipping when it was the fastest - because the only - means of communication across the oceans were not sufficient for the new electronic medium.

The book's structure is a combination of the topical and the chronological: a sequence of chapters on topics that fall roughly into chronological order. Most of the topics are technologies but a few, such as the deregulation of telecommunications since the 1980s, are sociological issues. The balance between technology and the many aspects of its history varies from one chapter to the next.

Solymar begins with "prehistory", tracing communication back to signal fires and ancient Greek schemes for encoding letters with patterns of torches but overlooking the famous "One if by land; two if by sea" lamp signal that opened the American war of independence. This chapter taught me much, particularly about the mechanical telegraph of the 18th and early 19th centuries, a mechanisation of the semaphore flags used in naval communications. It makes no mention, however, of the heliograph, a similar mechanisation of signals sent by flashing mirrors. The approach is throughout determinedly Eurocentric; there is no mention, for instance, of the smoke signals of the American southwest and the "talking drums" of west Africa, which depend on the tonal nature of languages in the region.

Prehistory is followed by the 19th and early 20th centuries. The first chapter here is on the telegraph, with special emphasis on the development of undersea telegraph cables. This, the author asserts, was the earliest non-trivial application of mathematics to engineering. William Thomson (later Lord Kelvin) used a sophisticated mathematical analysis to determine the diameter of the first transatlantic cable, laid in 1858. The problem was that distance tended to spread out the dots and dashes of the Morse code, running them together so that they could no longer be distinguished and slowing the signalling rate to the point of uselessness. The spreading depends on the resistance of the wire, which increases as the diameter falls. Thomson, one of the directors of the Atlantic Telegraph Company, recommended thicker wire but eventually acceded to thinner (and thus cheaper) wire. Solymar does not say who was right, and whether this decision had anything to do with the fact that the first cable worked for a few months only, or whether thicker wire was used the next time.

On this point, I was disappointed not to find any discussion of what it was like to debug this pioneering connection. Imagine sitting at one end listening to signals from the other end. You want to tell the people at the other end to do something different - but the only way you can get a message to them faster than by steamship is through the cable. This must have been a celebrated problem at the time.

The telegraph is succeeded by discussions of telephone and radio, topics exemplifying the problems of organisation that confront the historian of communications. The telephone is distinguished from the telegraph by the thing to be communicated, voice; the radio by the mechanism of communication, electromagnetic waves, which were made practical by such developments in engineering as the vacuum tube.

The final two-fifths of the book are devoted to "The modern age", from the second world war to the present. Once again, the topics are a mixture of media (microwaves, satellites and optical fibres), applications (fax and digital messaging) and fundamental technologies (transistors and lasers). Curiously, there is no chapter on television, surely the most visible communications development of the postwar period, and not even an apology for its omission.

In running together all later 20th-century developments, Solymar seems to have missed the point that although changes in the 1940s dramatically distinguished the postwar period from the prewar, an equally dramatic change occurred in the 1980s. The introduction of microcomputers into electronics has moved much of modern electronic functioning into software, the greatest manufacturing medium to enter human culture since iron; and the resulting decline in costs, coupled with the declining cost of the processors on which the software runs, is responsible for the vast range of consumer electronics products now available. Equally far reaching has been the transition to "computer-mediated communications". The chapter on "Deregulation and privatisation", for example, treats the drift from communications viewed as a natural monopoly to the current acceptance of competing long-distance and even local carriers as a business and regulatory phenomenon, without adequately acknowledging that the drift was made possible by the introduction of computer-controlled switching. It is this "computer-driven flexibility" that makes so tricky the distinguishing of computing, computation and memory on the worldwide web.

I was pleased to find recognition of two contributions from seemingly improbable sources: the writer Arthur C. Clarke and the actress Hedy Lamarr. Clarke invented, but did not patent, the communications satellite in 1945.

Science fiction can legitimately claim credit for numerous concepts that have later become staples of life - the earliest airbag I know of saved someone in a crash on Venus in one of Robert Heinlein's stories - but Clarke did far more than envisage communications satellites; he went a long way towards designing them in several technical papers. Lamarr, along with her husband the composer George Antheil, invented and patented a frequency-hopping radio, the first spread-spectrum system.

Inaccuracies in the book are more curious than serious. Solymar compares the old Morse code with a five-bit binary code (Baudot) and concludes: "On average the binary code is faster than Morse code by about 50 per cent." My own experiments suggest that with Baudot code as actually used, the ratio of speeds is the other way around. Gordon Moore's famous law about progress in computers is that the number of transistors per unit area doubles every 18 months, rather than every year as Solymar repeatedly states. Last, he asserts that the telegraph had a profound impact on war and diplomacy, while the only diplomatic impact of the telephone has been the "Moscow-Washington" hotline installed in the 1960s. But telephone calls between heads of state have played a crucial role in international relations since Roosevelt and Churchill spoke over the first digital secure telephone during the second world war. Worse yet, the example backfires. The hotline is, or at least was when introduced, a teleprinter connection - a latter-day telegraph.

The exciting fields of signals intelligence and cryptography are barely touched upon. Cryptography is mentioned by name only in connection with protecting the internet, and we read that "for obvious reasons not much is known about the development of cryptography after the second world war". The claim is partially retracted by a sentence later with the observation that "since the middle of the 1970s the subject has shifted gradually into the civilian sphere". Arguably, recent developments in cryptography have been as much public as secret. As with the impact of computers on communications, Solymar notes a phenomenon but underestimates its scope.

Getting the Message is a book packed full of interesting facts and charming stories. It often makes the reader wish it had said more, rarely that it had said less. The reader is nonetheless likely to depart with a sense of having learned many facts about both technology and history but not many guiding principles. If the book was intended to record those events in the development of modern communications that the author considers most noteworthy, it has succeeded. If it was to give a dweller in our communications-drenched world the sense of belonging to that world that comes from understanding how communications works and how it came to be, the book has failed.

Whitfield Diffie is Distinguished Engineer, Sun Microsystems, California, United States.