Distribute Freely The Case of the Two Cybersex Studies, (c) 1995 By Declan McCullagh declanm@netcom.com July 24, 1995 Dr. Michael Mehta's telephone rang. "Hi, this is Marty Rimm calling from Carnegie Mellon University. I'm the principal investigator of a study on pornography, and I'm leading a 15-person interdisciplinary research team. I'd like to take a look at the research you presented earlier this month." Mehta, then a graduate student at York University in Ontario, remembers answering the phone last fall. He never thought then that an unexpected phone call from Marty Rimm would bring him to accuse the former Carnegie Mellon student of stealing his ideas and his research eight months later. When he first talked to Rimm, he was flattered to be contacted by what he thought was a senior professor at a prestigious university. He tried to help the researchers at Carnegie Mellon University. "Rimm asked for a copy of my paper that I presented at a conference in November. He said Carnegie Mellon was publishing a book, and he might include my paper as a chapter if I sent it to him," says Mehta. Mehta and Dwaine Plaza, also of York University, had studied how adult BBS operators were marketing pornography on the information superhighway. After the two researchers sent "everything they had" to Rimm, they waited hopefully but never heard from him again. "I feel like a total fool now," says Mehta. "I was under the impression that Rimm was a tenured faculty member. He never corrected me when I called him 'professor Rimm.'" He didn't know about the controversy Rimm's paper had caused until last week, when an associate at the University of Waterloo told readers of a cyberlaw mailing list that Mehta and Plaza presented a cybersmut paper at a conference last November. Then researchers and reporters started calling Canada. Mehta is "flabbergasted" by the publicity Rimm's paper has generated "down south." His local papers haven't covered Rimm's study, so he wasn't aware of the controversy its publication caused. Now a post-doctoral fellow at Queen's University, he had given up hope of publishing his paper and moved on to researching democratic rights and public accountability for technology. After learning that Rimm was an undergraduate, Mehta felt silly at first. But after he read Brock Meeks' CyberWire Dispatches, he became convinced of the "seriousness of this issue" and decided to go public with his concerns. Rimm reportedly has retained a lawyer and has declined to comment. "It's funny how a breakdown in trust between people who have never met can happen over the wires," Mehta says. "I trusted Rimm, and he stole some of our ideas, lied to me, and distorted facts beyond belief." Looking back, he says Rimm's behavior was peculiar. He remembers asking Rimm for a copy of the book he was writing. "Rimm said there was an export restriction on it -- his publisher said there was an export ban," Mehta says. Studying CyberSmut: Similarities and Differences "I don't think the parallels between the two studies are coincidences. There are a lot of similarities that can't be accounted for by chance," says Mehta, who's working to document the resemblances. Rimm's paper, published in the July issue of the Georgetown Law Journal and cited on the front page of Time magazine, never mentions Mehta's study. Mehta says that Rimm changed the direction of his research without giving the York University researchers credit: "Rimm wasn't looking at the commercial distribution of BBS pornography until he spoke with us in November. We told him that we were looking for adult BBS logos and telephone numbers in Usenet images." Rimm's study claims that he downloaded and reviewed Usenet images in September, but doesn't say when they were analyzed. His study also lingers on the salacious details of erotica, repeating words like "fuck," "cock," "pedophile," and "paraphilia" dozens of times. "We also started off with emotionally laden language. But in the end, we cut down those terms," says Mehta. "That's what the Journal of Sex Research wanted. You think law journals are tough? Try peer review!" The Carnegie Mellon study somewhat resembles the Mehta-Plaza study, yet it also differs in many ways: * Rimm talks about pornography on adult BBSs, the Internet, Usenet, and the World Wide Web. The Canadian study only looks at images, many from adult BBSs, appearing on a small selection of Usenet newsgroups. * Rimm reviews descriptions of hundreds of thousands of images. The other study looked at the actual images, but only at a few hundred. * Rimm sorts images into only one category. The York University study categorizes each image in 22 different ways. * Rimm waxes poetic about how "pornography permeates the digital landscape." Plaza and Mehta discuss "a symmetric measure of association for 2x2 crosstabulations used when comparing non-parametrically distributed variables." Some of the Mehta-Plaza study's findings contradict Rimm's. Since the Canadian researchers actually looked at the images, they found that many "bestiality" images were actually cartoons: "There were only a few actual digitized photographs in our sample showing such acts." Also, where Rimm talks about "the ease of copying and disseminating digitized child pornography," the Canadian researchers say they found "no actual images showing a sexual act with children or adolescents." After presenting their paper at the November 1994 "Symposium on Free Speech and Privacy in the Information Age," Plaza and Mehta tried for months to find a publisher, but were stonewalled by the Journal of Sex Research. "They didn't like our findings," says Mehta. The journal's editors reportedly also disliked the references to anti-porn activist Catherine MacKinnon. Ironically, MacKinnon published a commentary on Rimm's research in the most recent Georgetown Law Journal, hailing it as a "landmark study of pornography in cyberspace." While both papers conclude that adult BBS operators are using the Internet to market their images, some recent evidence suggests the practice may not be widespread. Brian Reid, a network researcher and Usenet guru at Digital Equipment Corporation, says adult BBS operators become upset when their images appear surreptitiously on Usenet newsgroups. When they find out, they demand that the images be removed. Donna Hoffman, an associate professor of management at Vanderbilt University, agrees: "On the supply side, the operators realize that such 'leakage' hurts the market. On the demand side, the potential customer will say, 'Why should I pay over there when I can see it here for free?'" Whatever the reality, after all this controversy, there may be a happy ending in store for the pair of researchers. Mehta says he was ready to give up on publishing their study, but now "our paper will help shed a little light on this issue." In the wake of the Rimm scandal, he thinks they'll have better luck finding a journal to accept it. [Since his paper has not been published, Mehta decided not to put it online. He has agreed to answer questions about it and can be reached at: mm39@post.queensu.ca] A: ABSTRACT "A content analysis of pornographic images on the Internet" By Michael D. Mehta and Dwaine E. Plaza This paper examines the nature and content of 150 randomly selected pornographic images available through newsgroups located on the Internet computer network. Using content analysis, we identify themes which appear most frequently, and explore differences in the type of material posted by commercial and non-commercial users. Results suggest that commercial vendors are more likely to post explicit pornographic material in public access newsgroups in order to attract new customers to their private, pay-per-use bulletin board services. B: PARALLEL CONSTRUCTIONS Following are examples of where the two studies express similar ideas. Page numbers are in parentheses. Since both studies talk about similar topics, many parallel constructions could be coincidences. Rimm never mentions the Mehta-Plaza study in his paper. ------------------------------------------------------------------------------ RIMM STUDY MEHTA/PLAZA STUDY DRAFT ------------------------------------------------------------------------------ (1867) 17 of the 32 alt.binaries (7) 17 Usenet newsgroups were Usenet newsgroups contained identified that contained sexually- pornographic imagery oriented images ------------------------------------------------------------------------------ (1865) Images are also posted to (16) Other newsgroups besides newsgroups outside the alt.binaries those studied may contain hierarchy sexually-explicit images ------------------------------------------------------------------------------ (1914) Pornography is widely (13) Pornography on computer available through computer networks networks is widely available ------------------------------------------------------------------------------ (1852) Pornography is being vigorously (14) Commercial distributors marketed in computer environments of pornography use the Internet to market it ------------------------------------------------------------------------------ (1851,1875) Pornographers are using (14) Commercial distributors of Usenet newsgroups to advertise pornography post images to newsgroups products and attract customers to attract customers ------------------------------------------------------------------------------ (1868) The role of the NSFnet (4) The role of the NSF backbone backbone began to change is now changing ------------------------------------------------------------------------------ (1862) The Internet is increasingly (3) The Internet has the potential being used by pornographers to disseminate pornography ------------------------------------------------------------------------------ (1861) One of the largest uses for (4) The distribution of pornography computer networks is the distribution is a new use for computer networks of sexually explicit imagery ------------------------------------------------------------------------------ (1910) Pornographers are using (6) Commercial vendors may see newsgroups to advertise at no cost newsgroups as a way to advertise freely. ------------------------------------------------------------------------------ (1862) Until recently, the primary (4) Until recently, the primary use use of the Internet was linking of the Internet was for scientific university and government computers research purposes for research purposes ------------------------------------------------------------------------------ (1868) Multimedia graphics are being (16) There are more advanced types developed of image files which simulate movement ------------------------------------------------------------------------------ (1885) The presentation of kappa (10) Kappa coefficients were values was unnecessary because of the calculated and indicated a good high level of reliability * degree of inter-coder reliability ------------------------------------------------------------------------------ (1916) The study lists 4 boldfaced (22) The study lists 22 categories categories and 18 additional ones ------------------------------------------------------------------------------ (1852) Consumers enjoy considerable (5) Users download images in privacy while downloading images from the privacy of their offices computer networks ------------------------------------------------------------------------------ (1875) An adult BBS sysop can find (14) Most images come from magazines images in magazines or videotapes ------------------------------------------------------------------------------ (1912) Raises questions about (17) Says the Internet has the community standards in cyberspace potential to undermine local laws ------------------------------------------------------------------------------ (1908) Lawmakers are grappling with (18) Content of computer pornography digitized pornography and may decide will change with regulation of to regulate the Internet the "global village." ------------------------------------------------------------------------------ (1853) It may be difficult for (17) It is necessary for others researchers to repeat this study to replicate this study ------------------------------------------------------------------------------ (1875) Adult BBS operators are more (14) Commercial operators are likely to post explicit pornography ** more likely to post explicit pornography ------------------------------------------------------------------------------ * = Mehta says: "The kappa coefficient came right from us. He asked how we calculated reliability, and we told him. It's in there, but he didn't use it. I don't think he knew how." ** = Rimm's study says that adult BBS owners post 71% of hardcore pornography but just 59% of the combined total of hardcore and softcore pornography. C: ESTIMATES OF WORD OCCURRENCES (1) Note the highly emotional words Rimm uses in his study. Where Mehta-Plaza use the term "commercial vendor," Rimm uses "pornographer." Mehta-Plaza Study Rimm Study (2) ----------------- ---------- Paraphilia / Paraphilic 0 82 Pornographer 0 70 Child (3) 1 52 Hard-Core 0 52 Bestiality 2 45 Pedophilia / Pedophile / Pedophilic 0 41 Commercial 42 39 Dog / Horse 0 23 Fisting 0 22 Obscene 0 21 Cock 0 21 Fuck 0 21 Incest 2 17 Pussy 0 17 Pain 0 15 Abuse 0 7 Torture 0 7 Penis 7 4 "She holds the dog cock! Inserts it in her daughter's ass!" 0 2 (1) Since the counting technique isn't perfect the data shouldn't be seen as exact. Also, Rimm's paper is considerably longer so one would expect a word to occur more often. (2) To obtain these figures, the Mehta-Plaza paper was saved to a file with the references and Table 1 included. The Rimm paper and footnotes from http://trfn.pgh.pa.us/guest/mrstudy.html were saved in a single file. The following command sequence was used to extract word counts from the two files: tr '\040' '\012' < {study} | fgrep -i {word} | wc -l (3) Usually used in the phrase "child pornography." In this case, data were reviewed manually for accuracy. The one use of the word "child" in the Mehta-Plaza study was: "We never came across an image depicting a sexual act between an adult and a child/adolescent or acts between children." ###