YET another scare story about computers and pornography has hit tabloids and the Observer, after the Metropolitan Police Vice Squad wrote to Internet service providers suggesting that they censor some of their services. Let's try and separate the different issues.
Two years ago there were a lot of articles about the availability of pornographic images on disk, or downloaded from bulletin boards by children. At that time, an MSc student at Queen's University, Belfast, studied all the arguments raised, and showed that they were rehearsing arguments previously raised about every other medium: books and magazines, theatre, radio and TV. People dissatisfied with the previously established compromises were using the new medium to reopen settled arguments, sometimes allied with psychological attitudes to computers (fear of machines versus inevitable progress).
Separate from this are the issues raised that are specific to the particular technology used. "Porn on the net" is about as meaningful as "Porn on paper". The Internet is merely the wires joining networks of computers together, like a telephone system for data. Over the Internet we can communicate with other human beings in a number of quite different ways, such as real time chat (a kind of typed CB radio), browsing the Web and Usenet newsgroups. People such as the Metropolitan Police Vice Squad and Observer journalists get these confused. So what's special about Usenet?
In the 17th century, Samuel Hartlib corresponded with most of the scientists in Europe. He employed copyists to write out copies of the letters he received and send them on to his contacts interested in the same things. They were interested in all human knowledge, in particular the experiments and ideas that led to the founding of the first scientific societies such as the Royal Society of London. His papers are archived at Sheffield University. But once the Royal Society started publishing journals, scientists no longer needed to engage copyists to keep in touch.
They relied on central publication, with editors and printers. Now in the 20th century, we are using computer networks to do what Hartlib's copyists did in the 17th century: sending one person's short messages to exactly those people interested in the subject. We write our messages in a newsreader, then newsservers copy them and send them on to the next server, then the next, until they get around the world. For 20 years, people with special interests, from artificial intelligence to keeping cats, have used this system, called Usenet, to converse about their pet (sorry) subject, and get answers from the world's experts in their online club.
This is not publishing in the normal industry sense of the word. There is no central journal editor or printer. The computers forward everything. They do not understand the contents of the message, any more than the ill iterate copyists in early monasteries who copied manuscripts as drawings without understanding the Latin words within them. Unlike the monks, though, after a few days they delete the messages (to free up file space). The newsserver at Queen's University only keeps messages for two days. So Usenet messages are also ephemeral.
Those who treat the automatic forwarding of individual letters or messages as publishing clearly understand neither the technical nor social nature of Usenet. To talk about "pornography existing on the Net" in this context is like talking about "pornography existing within the post office". It is hard to see how anyone running a newsserver can, with existing software, exercise control and responsibility for the contents of hundreds of thousands or millions of messages passing through the server on the way to the next "stop, any more than a mobile telephone company.
The Philadelphia Court judges who declared the Communications Decency Act unconstitutional described communications over the Internet as a big, continuous, worldwide conversation. This certainly applies to Usenet newsgroups and the way people use them.
We post messages asking questions, giving information or whatever. Then someone comments publicly on the message, someone else takes it up, and an argument (or "thread") develops. Over a few weeks, several different discussions start, grow and decline, like conversations during an evening at the pub.
Socially, most newsgroups seem like more or less sedate versions of the conversations we hold at academic conference coffee breaks, pubs, football matches or London clubs (from the Reform Club to specialist S&M venues). If we need real world analogies to Usenet discussions, I would think of pubs, Cambridge Union debates, and experimental theatre where the audience gets involved, not publishing. You could think of each group as a different Edinburgh Festival venue, where a performance takes place for an hour or so, then gets emptied for the next one. Maybe we should be looking to apply theatrical censorship standards?
What people discuss depends on the nature of the group, as Usenet has developed a far more specialised and sophisticated classification scheme than the cruder ones we use for magazines: General, Sports, Women's, Computing, Top Shelf, Sex Shop. Talk and pictures about feet in alt.sex.fetish.feet does not interest me. But it is available for those people who are turned on by feet, instead of having to find a specialist club in a big city (which is hard if you are disabled and housebound).
Thanks to this, readers of Usenet can choose what kind of material to read, and avoid that which offends them. But since anyone can post to any group, from time to time we find people posting things that others object to (e.g. hundreds of chain letter money making schemes). This is normally controlled like an offensive remark in a pub. Other people complain, shaming the offender into changing his or her behaviour. As the numbers of posters and messages increased we may need to develop better ways of handling such social problems.
One way is by the application of laws to set limits on social behaviour. So there is a law against assault which helps prevent pub arguments degenerating into physical violence (something that cannot happen when the person you are arguing with is behind a computer half way around the world). We usually apply such laws to the individual offender, not the pub landlord.
People posting messages on Usenet are clearly responsible for what they write. They have been sued for libel and copyright infringement, and can be made responsible for posting images that would be illegal to pass around to people in the pub, or for setting up confidence tricks. Police can set up squads to track down those who posted messages which break particular laws, and charge them. Traditional investigation techniques such as infiltrating groups work online as well as they do face to face, and Usenet messages leave trails or paths showing where they were posted (just like letters).
The current brouhaha is not about individual responsibility for messages posted on Usenet, but about a request to get the ISPs to change the way they administer Usenet to prevent particular messages being posted and/or read. There are two questions: what can they do? and what ought they to do?
1. They could help the police track down the local posters of messages illegal under the relevant nation's law. This is something many would do already when searching for hackers. The ethical questions come when dealing with anonymous messages such as tips sent to journalists: after a recent landmark decision in the European Convention of Human Rights (Godwin vs United Kingdom), case law favours the protection of journalists' sources. But apart from this proviso, this should be a promising line of approach, since child pornographers using Usenet are more easily traceable than those using the post, and more easily deceived by undercover detectives posing as paedophiles.
2. The ISPs could stop passing on particular newsgroups. Unfortunately, anyone in the world can post any message to any newsgroup. This leads to two problems. In any newsgroup, no matter how senous, there will be the occasional posting of say, a picture of bestiality. So if a newsgroup is stopped as soon as an illegal message appears, very quickly most of Usenet would have been closed down, cutting off our country from the professional discussions that maintain the competitiveness of our newest information industries (rather like Japan at the time they allowed no westerners to enter).
Or we could merely cut out the newsgroups whose purpose is illegal (such as the few set up for the distribution of erotic images of children).
Unfortunately, the people - posting to that group can easily start posting to another (which is why there are hard core pornographic images turning up in alt.disney).
The one thing that can be done at the group level is to supply specialised newsreaders or other software to parents that prevent some of the newsgroups from being read by their children.
3. Classify and/or filter individual messages. This is not possible with current technology except at the level of searching for keywords, and not showing an article to children if it contains a word on the banned list. I hope that such software will improve on current levels of intelligence that cut out all references to Scunthorpe (idiot programmers on America Online didn't even think to check for where words start and end) or beer (American suppliers of lists of banned Web sites apply their standards on alcohol to Campaign for Real Ale sites in Britain).
However, there is a general need for some way of rating the quality of the floods of messages coming over Usenet each day, so that readers do not waste so much time finding the one useful message in masses of American college student postings. Solutions to this problem could also help with rating pornography on Usenet.
Now since Usenet is distributed and ephemeral, central rating agencies are impractical. There are no central publishers to send in material for rating, and there are too many messages posted daily for any professional group of reviewers to read more than a miniscule proportion. So instead we can try what papers do at the Edinburgh Festival. They have lots of amateur reviewers going around. I could read several reviews of the same play, and make up my mind which to trust.
Similarly, all experimental Usenet rating systems rely on readers voluntarily rating any message they read. By collecting the opinions of many readers, the systems calculate an average score for the message.
Most of these systems rate messages from very bad to very good. When dealing with pornography, there are two types of rating that might be useful. The first is a minimum recommended age (as for films). If each Usenet message had a minimum age: header, then children's news reader software could use it to prevent them seeing messages intended for adults. The second type of useful rating would be one indicating the actual specialised area of the message (e.g. sexual preference, body part, equipment or whatever). Then the reader's newsreader could skip messages containing keywords of things not to his or her taste, thus avoiding having to clean vomit off the keyboard. There already is a keyword header in Usenet that could be used for his purpose.
How would the messages get the ratings? Well there are several ways of doing this, but here is a suggestion. The initial poster fills in keywords and a suggested minimum age. At present readers of Usenet messages can, from their newsreader software, post a reply that appears in the group. So how about allowing a different kind of comment or reply: a preview message? The reader fills in his or her recommended minimum age, any extra keywords needed to better describe the message, and a rating of how important, or relevant or novel or well argued the message is.
This could generate a special kind of Usenet control message, a review message, which passes from server to server around Usenet. The servers could then combine the keywords into a long list, and average all the ratings, putting the results into the headers of the reviewed message.
Such an approach could both save busy people time in reading Usenet, and help parents protect their children from material that needs more maturity to be appreciated. It would, however, take some time to develop and test such a modification of Usenet. Given the small numbers of people actually using Usenet at present, wouldn't it be better to fund research into Usenet rating and wait a bit, rather than to be panicked into closing down part of Usenet now without actually affecting the amount of pornography on it?
. Next week the EU's Culture Commissioner, Marcelino Oreja, will unveil a discussion paper on new media services and how to target the authors of pornography on the Internet.
"We know that national regulation is not enough, that European regulation is not enough," he said last week. "We may need to have a world regulation of these matters, but let's go step by step."