Prof David Presti suspected some cheating was going on in his classes. He had an idea some of the papers he received - even for his advanced-level neurobiology class - were plagiarised, but he had no clue how many, nor how to catch the plagiarists.
But last spring Presti, a University of California at Berkeley professor, turned the tables on the cheats by using a service which could, for the first time, give teachers an edge in that battle.
Students were told in advance their papers would be vetted for plagiarism, but Presti was sketchy on the details. He didn't want them to know how thoroughly their work would be checked. For that particular assignment, every paper was submitted to a website called Plagiarism.org which analysed the work and assigned a virtual fingerprint. The essays could then be compared to almost every document on the Internet - approximately 800 million of them - and the service's own database of another 50,000 papers.
Despite being warned, about 45 of the 300 papers were flagged as having significant sections of text lifted from other sources. In that one class, Presti had caught almost as many plagiarists as the entire university had in all of 1995, when just 48 were detected. Considering the university offers more than 6,000 classes a year, Presti had uncovered a major can of worms.
"We didn't know what the extent of the problem would be," he says. "We were surprised by the numbers."
Academics all over the world have become increasingly concerned about students turning in other people's work. The US Department of Education recently cited cheating and the erosion of ethics as the greatest problem facing the country's schools - placing it ahead of violence and drugs.
Why this is happening is another matter. There is no doubt the Internet has made it easier than ever for students to lift information and call it their own. Not only can almost any topic be tracked down and downloaded, but a huge array of websites, known as paper mills, now cater specifically to students, providing them with either customised or on-file papers, many for free. Anthony Krier, reference librarian at Franklin Pierce College in Rindge, New Hampshire, tracked down and documented 80 such sites before he turned his project over to the Center for Academic Integrity in March, 1998. They come with names like Cheater.com, schoolsucks.com, or cheathouse.com ("industry leader since 1995," it claims, with 9,500 essays in 44 categories - "over 3 million served"), and all offer their wares for free.
Other sites, such as the Paper Warehouse and High Performance Papers charge from $6.50 for an off-the-shelf report, to $20 a page for customised research. "Students increasingly believe that what's on the Internet is public information and doesn't have to be cited," says Prof Donald McCabe, who has polled more than 15,000 students over the past decade about academic integrity.
In a recent survey by Who's Who Among American High School Students, 80 per cent of top-placing high schoolers admitted to having cheated at some point in their careers, half said they did not necessarily believe cheating was wrong. A US News and World Report poll found 90 per cent of college students believe cheats never pay the price for their actions.
John Barrie hopes his baby, plagiarism.org - the project he and eight friends have been working on night and day for the past two years - will go a long way to changing that. Barrie, 31, is an intense man who seems to have a few too many coffees in his system. His unruly thick black hair looks like a snapshot of the sea on a stormy day. He wears a blue, gold and white striped tie, a droopy beige jacket and two-tone shoes. He becomes particularly animated when discussing plagiarism.org and uses his hands constantly to emphasise points such as declining student ethics which he calls a "ginormous problem". But when he started out he was not seeking to solve any large-scale problems, just prevent students from stealing papers posted on the web for a class he'd been assisting. The papers were made available so students could see the kind of work being turned in by their peers, but by the next term many in the class were complaining their essays had been pilfered and handed in to other professors.
After writing a program which could detect plagiarised work from the papers on file, he turned to the on-line paper mills and then the entire Internet. Barrie soon founded iParadigm and recruited friends - all now part owners of the company - to write the complex codes which allow the website to do its work.
And what remarkable work it is. Within seconds plagiarism.org can trace submissions through 20 of the web's largest search engines - virtually the entire web - a vast and internal database, and the free on-line paper-mills. The goal is to have students submit all papers as a matter of course, before their teachers even look at them. The more institutions that sign-up for the service, the larger and more all encompassing Plagiarism.org's own database grows. Eventually, Barrie hopes it will have so many of the paper-mill essays in the database that students will be too nervous to continue using those services.
"My guess is that it will eventually become so common it will be like spell-checking," says Berkeley's Prof Presti. For every essay submitted to the site, plagiarism.org produces a report in which passages found elsewhere on the web are highlighted in colour - a different colour for every source. It notes locations of every highlighted passage so professors can do the final detective work themselves or students can refer to the work to defend themselves. One paper from Presti's course, for example, came back with five pages of material taken from a single source - the whole thing, except for a few sentence fragments, was highlighted in black.
"If we don't screw up at UC Berkeley, the whole California (university) system will adopt the program," says Barrie. "What's going to happen is the other universities' credibility will come into question if they don't check on the originality of their students' work."
Berkeley will be testing the service on a wide scale this spring. Georgetown University in Washington, Torun University in Poland, and a pilot program for selected universities in the UK are on the table for next year.
German researcher Gunther Eysenbach, recently uncovered plagiarised work in the Journal of the Royal College of Surgeons of Edinburgh using Barrie's site, which allows anyone to try the service for free. Eysenbach had written and published a similar piece one year earlier in the British Medical Journal: "I am not a native English speaker, so it takes hard work to write good English sentences," says Eysenbach. "I recognised the anguish of my work."
So did plagiarism.org, which eventually concluded that about half the article had been lifted from other sources, including Eysenbach's.
Even college professors seem to be vague on the idea. A recent Rutgers University survey found only 41 per cent of faculty considered the failure to footnote citations as being "serious cheating", so at least half of teachers have very different ideas of what the term means.
"To me, it's an ethical question," says Krier, of Franklin Pierce College. "Just because the papers are there on the web, shouldn't we be teaching that it's wrong to take them? The point is we have to teach kids to make the right choices."