Greetings from the twilight zone

"In the high-tech world, if you're not on the net, you're not in the know." Thus the Economist included the Internet in its festive guide to networks -- alongside the Freemasons, the Trilateral Commission, and others which only the best-informed conspiracy theorists can fret about. More seriously, Lynne Brindley, head of the British Library of Political and Economic Science, asks how, as a young researcher, "you break in to a discipline if you haven't source journals to look at". Increasingly, research is being discussed on the Internet rather than on paper: by "those in the know, in these invisible colleges who can safely whizz their way round draft documents and papers," as Brindley puts it.

Research has always involved "invisible colleges", whether they meet at conferences or exchange ideas in the post -- what the electronic community refers to as "snail mail". Does the age of electronic communication herald newer, more invisible and more exclusive colleges?

"Despite the normative description of science as an arena of fully-open communication, the new communication technologies exacerbate the practical problem of some groups of people having more access to information than other people." That's the conclusion of Bruce Lewenstein, of the departments of communication and Science & Technology Studies at Cornell University in Ithaca, New York state.

The first thing about an exclusive network is that many people don't even know about it. So some history is in order. The Internet grew out of a project by the US Department of Defense to build a communication system which would function after a nuclear attack. In the 1970s, programmers working for the DoD got themselves connected, and started sending electronic messages containing working notes, queries and -- crucially -- gossip.

The technology was taken up by the US National Science Foundation to make super-computer resources available to universities across the country. More and more local university networks joined. The British Joint Academic Network (JANET) gained a high-speed connection to the US, shared with NASA.

The Internet is deeply decentralised: an institution "joining" it need only be connected to a few "neighbours", which forward messages on to their neighbours, by whatever route is available, until they reach their destination. So no-one knows quite how large it is. One recent estimate is that about 7 million people -- somewhere between 3.5 million and 14 million -- have full access through their university or employer.

What's the Internet good for? You could, with permission, sit at a kitchen table on the isle of Jura and run a programme on a super-computer in Cambridge -- or, equally easily, in Cambridge, Massachusetts, or both at once. But most researchers deal more in text than number-crunching.

If you want to exchange text with colleagues around the world, you first need an "account" on a computer, or a local network, with an Internet connection. You compose your message in a word- processor and convert it to unadulterated plain text (ASCII in the jargon). You locate the account name for the person you want to write to -- more on that later. Type the command mail jones@history.winnesota.edu and attach the text; a few minutes or hours later jones looks at her computer in the notorious University of Winnesota and discovers your message waiting for her.

Immediately, you can see the possibility of collaborative writing with anyone, anywhere. You can form a group, too. A "mailing list" re-distributes all the messages it receives to all its subscribers. And you can have public discussions: a message sent to one of the more than 2000 "news-groups" is visible to anyone who cares to look, and possibly to reply.

It's not, of course, quite as easy as that.

Assume, for the moment, that you can type, in English. Assume that you have access to the necessary equipment. Assume that you're able and prepared to learn the sometimes baroque commands needed to access the system. Assume that you're tolerant of the fact that when you make a mistake, as you will, the system may fail to notify you at all, or may throw screeds of gobbledegook at you.

For these assumptions to be true, you're quite likely either to be a member of an academic institution in a Western industrialised country, or very well-to-do in world terms. You're also likely to be male. And the public area of the news system bears this out. An high proportion of messages -- over 90% in an unrepresentative sample of discussions of physics -- comes from the USA. An even higher proportion (of those with identifiable senders) comes from men.

"Women in science worry that these 'private' network exchanges of research results serve to reinforce the 'Old Boy Network' in scientific research circles, especially given the overwhelmingly male demographics of e-mail and news-group users," says Ruth Ginzberg, Assistant Professor of Philosophy at Wesleyan University in the US.

Why should there be this preponderance of men? Sarah Plumeridge is research assistant on a project to study women's use of computers at the University of East London. She comments that "A lot of research suggests that women prefer computing when it's for use, as a tool, when it's not taught as an abstract science." It's clear from the tone of messages in the public news-groups that the boys see them as a playground.

Newcomers are often mercilessly attacked for stylistic solecisms. Kerri Lindo, who teaches philosophy at Middlesex University, saw the Internet for the first time when interviewed for this piece. She immediately related it to her work on the French philosopher Bourdieu and remarked: "it's what I'd call a social Freemasonry -- you can't join a club unless know in advance what the rules are. Someone who learns the rules and then plays the game won't play it as successfully as someone who never explicitly learnt them -- just as people who learn middle-class manners or second languages always get caught out, however fluent they become."

And Josh Hayes, a post-doctorate studying community ecology at the University of Washington, may have hit on a sensible social reason for avoiding electronic communication: "For the moment, those of us who use the net a lot are probably considered to be, well, a little bit geeky. Real ecologists would be out in the field, don't you know."

There are more serious issues too. Cheris Kramerae of the Department of Speech Communication at the University of Illinois at Urbana is, working on the issue of sexual harassment on "the net". This happens in very specific ways -- men sending abusive messages to women, often having obtained their electronic addresses from the electronic "personals column". There is also the problem of socially retarded students abusing the system to distribute digitised pornographic images: the direct equivalent of the calendar on the workshop wall. Kramerae concludes, however, that "Obviously it is not the technology but the policies which are presenting particular problems for women."

Arnie Kahn runs a private mailing list for about 45 feminist psychologists from James Madison University in Virginia. "A few years ago I was sending electronic mail to a few friends who, like myself, were feminist psychologists doing research on gender.... I announced to my friends that if they had a question, they could just send the message to me and I would forward it to the rest of the group."

Kahn's list is, then, exactly an invisible college. Given the vast space occupied by anti-feminist men in the open news-groups which are supposed to discuss feminism, it can only operate if it remains private and by invitation.

Are there, though, fields in which access to the Internet is essential, rather than helpful, to making progress? It seems so. Jim Horne, an Associate Research Scientist in high-energy physics at Yale University in the US, states that "a number of people in high energy (only those with tenure though) have even stopped sending their papers to journals. They only send their papers to the preprint bulletin boards." Paper publication is quite simply too slow to bother with.

These collections of preprints are public, if you have net access and if you've been told where to find them. Stephen Selipsky, a physics post-doctorate at Boston University, points out that, since the preprints were made available in this way, "in the circles I move in, 'private' mailing lists play very little role... There is very little point keeping results secret in theoretical work, and large career rewards from disseminating results... in contrast to areas like biochemistry, where people [want] to stay in the lead on a hot topic."

Computer science is naturally another field where work is exchanged exclusively on the net. A researcher at Edinburgh -- who preferred not to be named "from shyness" -- says that "you tend not to chase up the actual publication (which can be months later). I have seen someone appealing for information about where some papers were eventually published, because you can't (yet) put 'archive@ohio-state.edu' in a bibliography entry." Here, too, there is at least one mailing list which is private -- "in order to keep down the traffic and free it from the 'can anyone tell me what a neural network is?' questions."

In some fields, electronic distribution is the only practical method. If you've ever watched someone laboriously typing DNA sequences out of a journal into a computer -- "ACG ACT AAG TAG" and thus for pages -- you'll see why this is the case for molecular biology.

There are some ways in which electronic communications can break down boundaries. "Speaking as someone at a relatively small and remote institution," says Steve Carlip at the physics department of the University of California Davis, "the biggest handicap is not private electronic distribution, but rather the fact that so much happens at seminars and in conversations."

Robert Gutschera finished his PhD last year and is now an Assistant Professor of mathematics at Wellesley College in the USA. "The heaviest users of electronic mail seem to be younger researchers," he says. "Getting into a field is always hard, but I think e-mail makes it better rather than worse."

Some are positively evangelical. Lewensten quotes Tom Droege, who is looking for "anomalous" heat production from palladium electrodes in heavy water -- the notorious cold fusion experiment -- in his basement laboratory. Droege communicates and discusses all his results publicly on the Internet -- finding negative interest from his work colleagues at Fermilab. "...the real experiment I am trying to do is e-mail science. The 'anomalous heat' project is just an excuse. I think this is the media of the future."

You may notice that most of the people quoted here work in the USA. This is, as you might guess, because their comments were obtained on the Internet -- neatly demonstrating the bias it introduces. On the one hand, the research for this article might have been impossibly expensive without it. On the other, people with net connections are tempted to talk only to the connected.

Kerri Lindo, as a total newcomer, was immediately struck by the possibility of finding others working on Bourdieu -- until she saw the content of the one public philosophy news-group: "It's a real shame, isn't it..." She composed and sent a message anyway -- and was able to predict what the programme would do next, which suggests that the computer software for sending messages isn't as awful as it's often made out to be, at least for post- graduate philosophers. She got just one response, from a group with an estimated 23,000 readers, and this could be summarised as "who he?".

Some of those ten or thirty thousand occasional readers of the philosophy news-group could probably be useful collaborators for Lindo. But how to find them? The sheer volume of public tittle- tattle -- known on the net as "the noise-to-signal ratio" -- means that only those with time to kill will pay attention. The Internet has no equivalent to a phone book. If you know that you want to contact a particular person, you know what institution they work at, and you can guess or find out that institution's electronic address, there are tools which may locate them -- but they're cranky and unreliable.

Often the easiest way to find someone's electronic address is a phone call, which may involve explaining exactly what electronic mail is to three or four departmental secretaries. On the other hand, once you've made contact, the computer screen is a great leveller. If you can work out how to de-gender your personal name, then all the information the reader has about you is what you choose to put into your text. (Or maybe not: Lindo recalls an small experiment in which she could tell the gender of pseudonymous essayists with 93% accuracy, though this was from hand-written scripts.) An intelligent and literate amateur could still conceivably enter into collaboration with a professor...

If you work in the humanities, you can probably put off coming to grips with the technology for a few years. You might want, however, to consider the rich seam of research on how this medium affects the nature of the messages. Lindo is not the only person to speculate that "It's possible that [the net] will influence the whole structure and nature of knowledge as much as the printing press did." Consider, too, that if Cyril Burt's twin studies had been published electronically, some awkward person -- very possibly an amateur -- would have run his figures through a statistics programme and spotted something funny, probably within 24 hours.

If you work in some fields -- certainly high-energy physics and molecular biology, and probably mathematics -- you'd better get connected, get retrained, or get a highly computer-literate graduate assistant ("a nerd", in the jargon) to do it for you. Lewenstein concludes that though electronic communication "will not replace traditional face-to-face interaction... researchers with access to these forms of communication [are] making progress while other researchers, still awaiting information through more traditional slower channels, have not yet begun to work." For them, the ability to use computer communication is an essential part of literacy.

Dorothy Denning works on computer security, and teaches computer literacy, at Georgetown University in Washington DC. She "doubt[s] that the electronic research communities will be any harder to break into than non-electronic ones. Based on my own experience, I expect they will be much easier to join (assuming you have the resources)." Her qualification is vital -- funders, take note.