Down and out on the electronic frontier

Worldwide electronic communication is inevitable. It offers much to those who can use it: but who wins and who loses?

LORD YOUNG told the 1992 general meeting of Cable and Wireless shareholders: "There are only a billion and a half people with adequate telecommunications, another billion with poor services and more than three billion people in the world who can today only dream of a telephone." Of course, as Professor Marvin Minsky at the Massachusetts Institute of Technology comments, "You could say the same about food, clothing, and housing."

The much-heralded age of global electronic communication is, nevertheless, arriving fast. To take just one example, Cable & Wireless is laying thousands of kilometres of fibre-optic cable from Hong Kong into the Special Economic Regions of China. Western and Japanese businesses investing in Shenzen and Guangdong will communicate with their Chinese plants over multiple 64 000-bit-per-second channels.

But will these developments divide the world into "the information-rich" and "the information-poor"? Will the people who, for whatever reason, cannot access electronic information -- and cannot converse with colleagues around the world from their own computer on their own desk -- be consigned to a backwater of human culture? Some say computer networks are the only way to keep afloat in a flood of information. But they also throw into sharper relief some existing questions about access to knowledge.

Concerns about creating an excluded majority of the unconnected fall under three headings. First, many may find the systems unusable: for example, most of the world's population cannot practicably use computers in their own languages.

Second, the technology itself may be inaccessible. Very many people live and work in places lacking the necessary communications. It can take three days to place an international phone call from, say, India or Bangladesh; and then the connection is often too poor to use for computer communications. More cannot afford the necessary equipment.

Third, there are political questions about access to effective education and training to use the technology. The vast bulk of communication on the Internet originates in the USA, a country whose education system professes equal opportunity between the sexes; and still the great majority of that is generated by men. Some ask whether this shows that women are systematically excluded.

What is it that people risk being excluded from? In 1945 Vannevar Bush, former science adviser to President Roosevelt, famously predicted that in the future books would be electronic. In their 1991 book Computerisation and Controversy (Academic Press), Charles Dunlop and Rob Kling quote his detailed prognosis: "The lawyer has at his touch the associated opinions and decisions of his whole experience... The physician, puzzled by a patient's reaction, strikes the trail established in studying an earlier similar case, and runs through analogous case histories, with side references to the classics for the pertinent anatomy and histology. The chemist, struggling with the synthesis of an organic compound, has all the chemical literature available in his laboratory, with trails following the analogies of compounds, the side trails to physical and chemical behaviour."

Bush's vision was uncannily accurate. Law firms use the LEXIS legal database, set up in the UK in 1979 and recording over 3 000 judgements a year since 1945, to search out every last precedent. At least one electronic Personal Medical Reference for doctors has been released (Technology, 12 December). Chemical Abstracts, made available in full on-line by the American Chemical Society in 1980 and on March 2 containing 9 768 152 abstracts, was one of the first great public electronic databases.

These ventures simply add value to information which already exists on paper. Scientific journals are starting to appear published only in electronic form. Last year the American Association for the Advancement of Science launched Current Clinical Trials (CCT). Others include Psycoloquy, another refereed journal sponsored experimentally by the American Psychological Association, and EJournal, which discusses electronic communication. Each of these is available only to those with access to Internet, the global academic computer network.

The Internet is by far the largest computer communications network, with estimates of the number of people having access to it ranging over 12 million. Its origins lie in a project by the US Department of Defense in the late 1960s to develop a command communications system which would function after a nuclear attack. The experimental system which resulted, known as ARPANET, was totally decentralised. Each computer connected to it would send "packets" of information off in the general direction of their goal, leaving routing equipment along the path to send the packets on by whatever route was available. This article was originally filed via Internet. It is quite possible that the first half went in a packet from the author to New Scientist's London offices -- a distance of three kilometres -- via New York and San Francisco, and the second half via Hong Kong.

A columnist in the Communications of the Association for Computing Machinery reports that the Iraqi military command uses this technology, and that this is why their communications held up under massive bombardment in 1990. When the Iraqi Air Force building in Bagdhad was destroyed, as shown on TV, messages were automatically and instantly routed around it. This is why the US government has tried to control the export of modern routing equipment.

In the early 1970s, ARPANET technology was adopted by the US National Science Foundation for the early Internet, at first linking universities to super-computers. An initially incompatible system was developed in Britain for the same purpose, under the name of the Joint Academic Network (JANET). The British network was progressively integrated with the Internet between 1979 and 1984. Computer programmers working for the Department of Defense already had access to ARPANET; an Internet account became the mark of a computing insider -- what Ed Krol, one of these pioneers, described self-mockingly as "a DoD-certified nerd". Given the network's decentralisation, estimates of its usage typically claim accuracy within only a factor of two. Growth of 10% per month in the number of individuals with access is widely quoted on the net.

Why is anyone bothering with this technology, when ink on paper has stood us in good stead for so long and can now be transmitted in seconds by fax? Pam Waddell, senior researcher studying electronic publishing at the Science and Engineering Policy Studies Unit, says that "there are things which are very difficult to express on paper which you can do with an electronic journal -- for example 3-dimensional molecular models." In genetics, DNA sequences which are published only on paper must at present be laboriously typed into researchers' computers before they can be used.

And as Derek Law, librarian at King's College London, points out: "In 20 years Physics Abstracts has grown from 80 000 to 120 000 entries a year. That's 100 abstracts an hour every working day, which means it's impossible to keep up with and is why it's known in the trade as 'the green slime'."

In the future, Physics Abstracts will be a service which tells you in the morning which of the papers published overnight match a set of key-words which you have specified. You'll be able to ask for reverse citations -- all the later papers which reference the one you're reading -- and go directly to read them. And you could see the citation count, and the readership count, change day-by-day. Such tools will be the only way to keep up with the flood of information -- as currency traders, for example, already know.

Bill Gates, president of the Microsoft corporation and thus theoretically America's richest man, predicts that in a decade electronic publications will be worth US$1 billion a year in sales. Most of this will be distributed on various forms of compact disk read-only memory -- CD-ROM, CD-I and so forth -- and will be entertainment, education and home reference. It is projects like the electronic journals -- primary sources of research information available only with new technology -- which set alarm bells ringing for those concerned about what the US organisation Computer Professionals for Social Responsibility calls "the new international information order".

CPSR was formed among Californian computer scientists and technicians in 1981 and is described by journalist and author Bruce Sterling as the US's "veteran cyber-political activist group, with more than two thousand members in twenty-one local chapters".

"The body of human knowledge", a 1992 paper from CPSR's Berkeley group proclaims, "is a social treasure collectively assembled through history. It belongs to no one person, company, or country. As a public treasure everyone must be guaranteed access to its riches." But if some of that treasure is now to be accessible only through computer networking technology, how can everyone be guaranteed access?

The first problem of access is language. Until December Jagdish Parikh worked at NGOnet in Uruguay -- a project conceived at the United Nations Conference on Environment and Development, and funded by the Dutch and Canadian overseas development administrations, to improve communications between UN-affiliated Non-Governmental Organisations, particularly those in "the South". He comments: "Most software is developed in English. Even if it's available (usually at extra cost) in some other languages, it's not ideal for languages using non-Roman alphabets. One cannot communicate [by computer], for example, within India using any of its major languages." Even where scripts are available, different programmes may use different codes to represent the same character, and the files may be very large -- both factors making electronic mail impracticable.

If you accept that in future all business and scientific communication will be in English, the solution to this problem is merely a question of improving education. However, almost every script and alphabet known to humanity should soon be available on computers -- within two or three years, according to Michel Suignard of Microsoft Europe in Paris. Microsoft's "WIN-NT" computer operating system, due to be released later in 1993, will implement the joint International Standards Organisation and Unicode standard. The first draft of this, to be published soon after five years' discussion, will specify a consistent and compact way of encoding dozens of scripts in computer files, including Bangla (Bengali) and Devanagari (Sanskrit and Hindi), Tibetan, Lao, Tamil, much of the Chinese, Japanese and Korean scripts, and dozens of others. The draft even mentions projects to include Cree and Cherokee in the next release.

One remaining difficulty is getting people to design the type- faces to display the characters of the various languages. Another is getting actual user programmes to display and edit text. And for technical reasons the "header" on an electronic message -- who sent it to you, when and whence -- will remain in English- based computer gobbledegook.

The second issue that critics raise is access to communications and to equipment. Philip Machanick of the Department of Computer Science, University of the Witwatersrand in South Africa comments: "I find it ironic that at Stanford [University in the USA], where I've spent some time, there are great libraries but I seldom needed them because the [Internet] access was so good. By contrast, it's been a long uphill struggle to get decent net access here where the libraries are badly starved as a result of a collapsing economy -- and it's even worse in our poorer neighbours such as Zimbabwe."

Besides efforts such as Cable & Wireless' programmes to wire parts of the world, help may be at hand soon from satellite communications. Among several projects, Motorola's Iridium scheme would, if launched, involve 66 satellites in low (777 km) Earth orbit, with links between the satellites. The company estimates the cost of hand-held receivers, with built-in data communications adapters, at US$2500 ([UKpound]1670) when the system is due to be available in 1998, and cheaper later. The cost of a 64k-bits-per-second (up to 1000 words per second) connection from any point on earth to any industrialised country's phone system should be US$3 ([UKpound]2) a minute. Currently, a standard rate call from the UK to Zimbabwe costs [UKpound]1.12 a minute, plus tax, with a data capacity of 20 words a second if you're lucky.

In contrast, UK universities pay a flat subscription of around [UKpound]6000 a year for access to the Internet, plus data line rental and data-storage costs. Individuals can get access in London for [UKpound]10 a month. It would then be possible to read CCT at your desk on basic personal computer equipment costing around [UKpound]500 -- but basic equipment allows only basic, slow and unfriendly access. This minimum price is not likely to fall significantly in real terms.

University libraries are installing moderately powerful workstations, capable of running more friendly user interfaces, at over [UKpound]10 000 per user. And as Sheila Edwards, co-ordinating a study of information systems at the Royal Society, points out, "only one person can look at a computer display at one time -- if you have a multi-volume reference book lots can use it... a terminal on every desk is far from achievable yet."

Derek Law is sceptical, however, of the argument about cost: "At present your Bangladeshi scientist has to find huge wodges of foreign currency for journals.... there are numbers of journals which cost more than [UKpound]4000 a year these days. If they're doing research they have some funds, and you're talking about priorities -- deciding between getting more paper journals and getting electronic access -- not absolute funds."

Electronic information has the potential to be much cheaper than this. A subscription, for example, to CCT, giving permission to access it over the Internet, costs just US$95 ([UKpound]63) a year. But many librarians are worried about the possibility that in the future you'll be charged each time you read an article -- in the jargon, information will be provided as "a transaction-based service".

Derek Law is involved in a feasibility study for a UK-based electronic journal, which looks likely to be launched with the Institute of Physics. On charging, he is "passionately convinced of the benefits of subscription-based rather than transaction- based services -- we have all the evidence we need that as soon as you charge by transaction you drive the [academic] user away. In my experience as a librarian, where searches have been charged on a time basis use is trivial, whereas the on-line UK bibliographical information and data system, charged by annual subscription, attracts 3000 enquiries a day."

Technological and commercial developments are, then, promising to solve the problems of language and of physical access to the network. This leaves the third, and in some ways the most political, question raised by critics. Once a network connection is on place, and a way of charging for it has been worked out, who can or will use it?

It is, given the origins of the Internet, perhaps not surprising that its facilities are notoriously "user-hostile". To use it to connect your computer to a remote service such as CCT involves typing a series of supremely unmemorable commands; and, when you make a mistake, the error message, if there is one at all, will be unhelpful.

Of the Internet's visible users -- those who contribute to its public discussions or "news-groups" -- over 90% of those with identifiable names are male. Is this evidence of discrimination in education convincing women that they cannot use computers? Derek Law has "never seen it myself as a gender issue -- but it's true that in my own daughter's primary school there is a perception that the boys are more exposed to computers than the girls, if only because they knock the girls off."

Sarah Plumeridge is research assistant on a project to study women's use of computers at the University of East London, and teaches students to use programmes. She finds that "If the men don't know what they're doing, they're not flustered. They might say 'I don't know what I'm doing but it's just a question of getting down to it and learning.' Women are more likely to feel that there must be something wrong with them and they're stupid."

She says that "the debate about gender and computing is under- theorised. It's like the debate about mathematics and gender in the early 1970s, when all those pronouncements were made about girls having less developed spatial ability. Feminists then re- analysed the tests which were used, and found that the differences were not as significant as had been said, and do not explain why women were so absent from maths, science and technology."

Plumeridge believes that the differences in the way girls and boys currently learn about computers will also turn out to result from "a complex interplay of classroom interaction and teacher expectation." As Law's daughter finds, the classroom interaction is also a matter of competition for scarce resources. The teacher expectation may take a generation or more to work its way out of the system.

Dorothy Denning, who studies computer security at Georgetown University, Washington, is more hopeful: "I'm teaching a computer literacy class at Georgetown... Enrolment is 50-50 male and female. Based on this, I speculate that gender will cease to be an issue as far as computer and network use goes." And Plumeridge sees hope in "the way that women use the telephone in ways that no-one predicted when it was seen as a business tool."

Brian Lang, as Director of the British Library, is committed to a policy of making information available to people in their homes. He is encouraged by the spread of facsimile technology: "I'm astonished by the rate at which we have become fax-cultured -- the machine has incredibly rapidly become a standard means of communication. A fax used to be a precursor to The Real Thing -- a letter delivered by a person with a sack. Now I get letters marked 'in confirmation of fax' -- obviously this is (still) necessary for legal documents."

Even critics of the way the technology is applied agree that the problems are the same social and economic issues which apply to the distribution of knowledge by any means. Derek Law concludes: "Indian villages may never have access to the Internet, but since at the moment they don't have access to Cambridge University Library we're not disadvantaging them any more." Marvin Minsky is quite clear: "When economic inequalities lessen, so will informational inequalities -- without any particular special attention."

And perhaps we should not under-estimate the importance of fashion in getting people to take up the new technologies. A teenager who does computer graphics for night-clubs opined on the youth-oriented BBC show The Word: "In future, if people want to be hip, they won't be talking about [rock group] Jesus Jones, they'll be talking about getting a modem."