Critique of Rimm Article on Online Pornography

Introduction

Martin Rimm's work demonstrates that pornography exists on the Internet. However, many of his conclusions about that pornography and about its implications remain problematic, because of biases in the research which result from the purposes of his study. As Rimm states, "The emphasis of this study is to determine how many images pornographers market as pedo/hebephile, an important area which has not been addressed by the legal system." There is evidence to support that this study is a deliberate search for child and teen pornography, though this is not a claim he makes in either his introduction or conclusion: there he argues that his study shows the extent and types of online pornography, a vastly different aim, that would not have obtained the same results. By looking for child and teen pornography, Rimm's study finds it; Rimm's study artificially conflates certain types of "deviant" sexual matter, in order to support his conclusion that the Internet contains significant lawlessness, and merits closer regulation.

Body

1. The categories he uses for "linguistic parsing" are subjective, not objective as he states. In Appendix A, he states that "extremely large breasts" are counted as hard-core. Breasts are intrinsically part of female biology, whether large or not, and need not be soft, hard, or any core at all. Likewise, "amazing" gets counted as fetishistic rather than as a fairly useless adjective. That means that pictures that are described as amazing get counted along with bestiality. Amazing! In Appendix B,"PREGNANT" was counted as "hardcore" perhaps because the woman was described as "spread cunt, milking tits. Lactation and pregnancy happen to be fairly normal female conditions. Since the image portrays no "action" other than milking (which women do) it should not be labeled hardcore.

2. Points 1 and 2 become even more problematic because of the way he sorts material. The flowchart on page 1886 illustrates the problem. All the materials were sorted first by child/teen descriptions. If there are any indications that the material might be child pornography the sorting stops there. If not, the description was tested for paraphilia. And on through hardcore to softcore. The flow of information favors whatever gets tested first. For instance, if an image of an adult woman with shaved pubic hair was described by "pretty, young, hairless, blond," the image would stop and be counted as hebe/pedophilia because of "young and hairless", rather than being counted as softcore because of "pretty and blonde." The flow of information favors ordering the information along the most deviant construction. If Rimm had run his tests looking first for softcore characteristics, his numbers would change.

3. The source for Rimm's Internet usage statistics is Carnegie Mellon University, the world's third-busiest Internet site (according to 1995 NSFNET statistics), and the most heavily networked university on the planet. For several reasons CMU's data (though easily available to Rimm, who conducts his research there) cannot be considered a source of representative USENET traffic: Carnegie Mellon's e-mail, for instance, has built-in facilities for downloading encoded images which other universities do not. CMU is directly connected to the Internet backbone (which runs through Pittsburgh), and images copy much more quickly at CMU than other sites (the standard dorm-room connection there is over seventy times faster than a 14,400-baud modem). CMU also has a bias in the Internet newsgroups it carries: at the time of Rimm's study, for instance, alt.binaries.pictures.erotica was available, while alt.culture.theory was not; the fact that some newsgroups are not available at CMU tends to bias Rimm's conclusions about what percentage of newsgroups post pornography.

4. Another significant factor is CMU's gender ratio: only 29% of students at CMU are women, which is relevant to any examination of sexual relations and practices there. Young men with little contact with women might access pictures of them instead, driving up the use of pornographic materials. This will skew Rimm's findings as well.

5. In his discussion of commercial BBSes, Rimm discards from his study sites which offer only 1200- or 2400-baud connection rates. Though in 1994 (the time of his research) Carnegie Mellon offered its faculty and staff faster connection speeds, many sites (including many universities) did not, and this exclusion biases his study, suggesting that images constitute a larger percentage of digital communications traffic than would a more inclusive survey. It also weights his study away from rural areas with lower tech telephone exchanges.

6. Any description with terms that suggest hebephilic, pedophilic, or paraphilic content are counted as hebephilic, pedophilic, and paraphilic. Elsewhere Rimm states that not everything labeled as incestuous is incestuous, which demonstrates that he understands the difference between labeling and content of images. Nevertheless among the images he classifies as pedo/hebephilic are those that are, for example: "flat-chested, their genitals shaved, clad in pigtails and bobby socks, and sometimes surrounded by stuffed animals" (also in note 71). Making someone look young does not make that person young, but Rimm counts it as a genuine incidence of hebe/pedophilia.

There are other interpretive problems in the study mostly because Rimm is an engineer attempting to look at social issues. He takes an engineer's approach where you identify a problem, find an approach, apply it, and fix it. Social issues like pornography require a great deal of nuance or as he calls them "subjective approaches" because the world, especially the world of fantasy, does not fit into little boxes. Issues like child pornography are hard to address, let alone quantify. What is child pornography? Anyone naked who looks young, even if they're not? Your children in the bathtub? How about photographs of naked children from the nineteenth century which combined sensuality and sentimentalism? Rather than understand pornography on its own terms as a complicated phenomenon, he takes all things that look alike and lumps them together. Moreover, he makes analytical leaps that are mind-boggling; Rimm implies that the desire for privacy on-line equals the desire to molest children (see note 30). He states that the problem of pornography on the Internet is getting worse but the study lacks any data or analysis of change over time. Rimm does not demonstrate that there is anything new in the way of content. He admits that most images are scanned in from other print sources; they are just reused and recycled. So what's new, other than access? Or is access the main problem in which case he should discuss exactly who gained access. Perhaps, he could even talk with them in person or online to discuss how access through an electronic medium has affected them. Using an engineer's approach can be enormously useful for many things, but when that approach gets applied to society through simplifying society to fit approach, more is lost than gained.

At CMU Rimm's study has been taken at face value and (in best engineering fashion) the "problem" has been fixed. Pornographic images have been removed from the university's USENET servers. A thriving underground has developed, allowing students to gain the images they want (however, the content of that underground is not softened by public debate and discussion - pornographic servers at CMU are now private, rather than open, accessible and responsible to the community). Censorship has neither ordered nor controlled pornographic images online at CMU; local sites flourish because CMU students are both canny and computer-literate. Those without the time, skills, or contacts can always turn to commercial servers; money has been able to buy pornography in the past (before the USENET and even before photographs). PLAYBOY and Penthouse are already online, ready to replace the small bulletin boards Rimm discusses in the business of selling sex. Whether we want such forums to be the only sexual discussion available is what's at issue. Rimm calls for a regulation of Internet content, but only sites with experience in that business will likely then be capable of providing sex online: commercial sex or underground sex won't deliver democratic, open, responsive or responsible communication.

Lisa Z. Sigel is a doctoral candidate in Social History at Carnegie Mellon University working under the direction of Dr. Peter Stearns. Her dissertation, entitled "Sexuality and Social Change: British and America Pornography, 1800-1914," looks at the changes in pornography and the pornographic imagination. She has recently published an article on deviancy in the journal Cultronix available only on the world-wide web.

Geoffrey Sauer is a doctoral candidate in Cultural Theory at Carnegie Mellon. He is writing a dissertation about the languages of order and control on the contemporary Internet, and has published "Images of Women at Carnegie Mellon," an ethnography of pornographic servers online. For the past five years he has managed an online publishing site (the English Server at CMU), which today hosts 850,000 readers per month.