Jim Smith finds that the Net is no place to anchor references in serious academic studies. I came across some material on the Internet while researching United Nations peacekeeping some years ago which I thought would be appropriate to include in the piece I was working on. I wondered just how to cite the information, and my solution was wholly unsatisfactory. I have finally concluded that not only is citing the Net inherently difficult but it is methodologically unsound in almost all cases.

Before being accused of being a neo-Luddite or a technophobe, I should stress that I think the new information technology can be an incredibly useful tool. It can provide information quickly, cheaply, and which is sometimes unavailable elsewhere. For example, I like being able to access UN resolutions without leaving my desk and to recieve virtually instantaneous reports of the current situation in Bosnia. However, as a reference, as something which I can use in books or journal articles, the Net leaves a lot to be desired and, moreover, may have serious implications for the way we view history itself.

Citing sources of information in academic publications serves two basic purposes: first, readers can discover just where the author found the information which should be verifiable and second, they can look up that source to read more about things that interest them. For the most part, the Net fails on both counts, primarily because electronic information is not stable. Because any information on the Net is, to use Negroponte's terms "bits, not atoms", that information can be altered. This is not always true; CD-Roms, for example, provide a stable source of information which cannot be changed. But most data on the Net is primarily in the form of "gopher" or "html" documents, which consist of little more than standard word-processed text files. These files can be changed by the people who own them at any point, and that makes them unstable and useless as academic tools.

For example, suppose the electronic address of some crucial environmental information is . update.html

ÖйúAƬ

/.

Anyone reading this address will know that it is a World-Wide Web document, which is the property of some company called "Enviro", which could just be a short form of their full name. The name of the vital document is "update.html". But that document, placed on the Web by Enviro, can be changed by Enviro at any time .

Unlike paper documents or books, which are reproduced hundreds or thousands of times, there is only one electronic copy of "update.html", which can be altered by the providers (Enviro) or by a dedicated hacker.

ÖйúAƬ

Thus, Jane Academic, citing "update.html" as her source (and using the full address mentioned above), has no idea whether the document will be the same next time she looks at it, whether it will still be the same if ever one of her readers looks at it, or will even exist at some point the future.

Gopher and Web documents, not to mention Usenet, chat, and discussion group documents, although all having an electronic address, do not remain permanently on the system - unless the users wish them to. Negroponte is simply wrong to say: "Digital books never go out of print. They are always there."

Making a printout of your source of information does not solve the problem. Suppose Jane printed out a copy of "update.doc" in the form she used it when she used it, and suppose that since that time, the original document has disappeared or been altered.

Another researcher contacts Jane, informs her of the situation, and Jane is able to produce her printed copy of the original document for the inquiring researcher.

Unfortunately, Net documents are the easiest of all to forge, because they have no original hard copy. No one knows what the original document looked like, because it was in an electronic form which can be printed in just about any way the user wishes.

ÖйúAƬ

Perhaps, as academics, we have to trust that other academics and providers of information will not alter their material to deceive us. The point is that this capacity exists, and even if you trust the providers and fellow academics, this still leaves the problem of hacking.

Why would anyone want to hack into a UN resolution and change the data? For fun, for profit, or for "reasons of national security". When the Mexican Zapatista movement began posting its information on the Net, anyone who saw that as a threat could have hacked the system, resulting in documents which looked nothing like the original, and put across a message the Zapatistas never intended.

Libellous accusations, lies, and damaging propaganda could all be present where they never were before.

ÖйúAƬ

This has more serious implications. If, as some information technology specialists are predicting, there is a dedicated move from atoms to bits, from paper documents to electronic ones, and if, at some later date, all that exists are electronic documents, how will we trust that any event happened, that anyone said anything at any time anywhere?

The simple truth is that we cannot. History itself can be rewritten.

Unless electronic documents can be verified on paper or in unalterable electronic formats, then housed in multiple locations (preferably publicly accessible libraries) - they simply cannot be cited in academic articles. However accurate researchers are in detailing the electronic source, they have no guarantees that the source will exist unchanged. This may be one area in which the technology (or its use) lags behind, and it may be that with appropriate will, documents can be made more secure and be made unalterable. For now, the Net can only be seen as a source of interesting and useful information - essentially useless as a citation.

Jim Smith is a lecturer and computer officer in the department of peace studies, University of Bradford.

ÖйúAƬ