Article by Mike Godwin
To anyone with even a slight background in research methodology, Martin Rimm's now-notorious Georgetown Law Journal article is so outrageously flawed and overreaching that you can't miss the flaws even on a cursory first reading. Or at least that's what we've been saying all along.
Yet it's harder to justify that statement in the face of Marty's spectacular success in convincing otherwise intelligent people that the study was a neutral scholarly work of merit. If the study is so obviously worthless, you might ask, how do we explain Marty's having persuaded so many people, from the CMU administration and his own faculty advisers, to the three law professors who endorsed the study in their separate articles in the Georgetown Law Journal, the journal itself, and, finally, the nation's leading newsweekly, that his undergraduate research project was worthy of all its hype?
It took me a while to piece together Marty's winning sales tactics. Some of them have to do with misdirection -- Rimm rarely let anyone actually see what he'd been writing, and even the law professors who were conned into praising his work read only an early draft of Rimm's article -- one different in many respects from the article that finally appeared.
Marty's primary tactic, however, was the most basic of the traditional con man's bag of tricks -- Marty was exceedingly adept at telling people what (he thought) they wanted or needed to hear in order to want to help him.
I first had a clue as to Marty's reliance on this tactic when I started rereading my old e-mail relating to the Rimm study. On rereading, I was struck by the way in which both Rimm and his faculty adviser, Marvin Sirbu, pitched their respective solicitations to me last November, long before the Time Cyberporn scandal, and even a day or two before the free-speech rally I participated in at Carnegie Mellon.
In seeking my collaboration on his article, here's what Marty Rimm wrote me:
"Given that you will be on campus tomorrow, there are two things I would like to discuss with you. First, I would appreciate an independent check of our legal footnotes, which to some extent are based on your postings and articles. Second, our preliminary data indicate that there are no significant differences among individuals in communities across the country in what kinds of erotic materials, including pornography (visual and verbal) and obscenity, they find of interest. In our experiment we began by assuming that there were indeed community standards which differed across communities - that is, that some communities of individuals had no tolerance for or interest in say, pictures of heterosexual anal intercouse. We then began collecting data to allow the evaluation of the ~null hypothesis~ - that is, that there are no differences. Our conclusions are very clear: there are no differences when communities are defined by telephone area codes. This there are no ~community standards~on which communities differ." --Letter from Martin Rimm, Nov. 8, 1994.
In seeking EFF's sponsorship of the Rimm research, Sirbu wrote this:
'As you may have gleaned from reading about the events at CMU, I have been working with Martin Rimm on a study of the availability and consumption of sexually explicit imagery on Adult BBS systems and, to a lesser extent, on Usenet. Andy Blau suggested that EFF might be interested in the work we've been doing. Among other things, we have data which could be analyzed to show the geographic distribution of consumers of adult BBS systems. Such data might be useful in countering or confirming assertions that "community standards" in places like Memphis are different from other regions of the country.' --Letter from Sirbu, Nov. 7, 1994
In other words, both Rimm and Sirbu were telling me that their research provided an empirical basis for an argument that the Supreme Court's "community standards" obscenity doctrine, promulgated in Miller v. California in 1973, is no longer valid, if it ever was.
Now, given that a) both Rimm and Sirbu were reluctant to disclose much detail about the study, and b) the GLJ article, once it appeared, gave little attention to Rimm's "preliminary" finding that there is no appreciable difference among communities in terms of tastes in pornography, you have to wonder why they each chose to share this purported finding with me.
I think it's because they'd read one or both of two articles I'd written in, respectively, summer and fall of 1994. In the summer 1994 article, which appeared in the SF Examiner, I wrote the following with regard to the Amateur Action BBS case:
'[T]he prosecution of the Thomases and their "Amateur Action BBS" calls into question the continuing validity of the Supreme Court's obscenity decision in Miller v. California, now more than 20 years old....
'It's time for the courts to revisit the Miller obscenity standard. In the face of changes in communications media and the evolving nature of "community," the courts should modify the application of the Miller standard to prevent this kind of prosecutorial overreaching. Failing that, the courts should abandon the "community standards" approach altogether.'
In an expanded version of this article -- this for the November 1994 issue of REASON (which was on the stands through most of October) -- I wrote the following:
'The legal framework for the conviction in the Thomases's case is nearly 20 years old, and it reflects both an outdated understanding of 'community' and the ongoing desire of some prosecutors and much of the religious right to turn back the clock regarding sexually explicit materials."
It seems likely that Rimm and Sirbu aimed to give me the impression that their work would support a particular thesis of mine, hoping that this would attract my support and even collaboration. Nothing wrong with that on its face, except that in this instance the tactic communicated a false impression of the focus of the research. (Note: at that point in early November, I had not yet been given a copy of the abstract -- the only direct information I had about the content of the study was that given me in those letters from Rimm and Sirbu. Had it not been for CMU's student-body president, Declan McCullagh, I might never have gotten an early sense of the contents of the study -- Declan sent me a copy of the abstract later in November.)
This and other evidence suggest to me that a standard Rimm approach, when pitching his article as a serious contribution to scholarship, was to tell potential collaborators that Rimm's work complemented theirs. This becomes even more likely when you look into the backgrounds, previous work, and known philosophical positions of the three law professors -- Anne Branscomb, Catharine MacKinnon, and Carlin Meyer -- whose legal- and policy-oriented commentaries follow Rimm's work. For Anne Branscomb, who has written in her book WHO OWNS INFORMATION? and elsewhere about the issues of collection of personal data, Marty's work stands in support of her claim that
A great deal of information we consider to be highly personal, and of interest to ourselves and the town gossip our names, telephone numbers, marital status, educational accomplishments, job and credit histories, even medical, dental, and psychiatric records is now being sold on the open market to anyone who believes he or she might be able to use such information to turn a profit. These transactions usually take place without our knowledge or consent.
Anne Wells Branscomb, Who Owns Information? 3-4 (1994) (quoted in the section IV of Rimm's article).
Take that passage from the introduction to Branscomb's book and compare it to the the first three sentences of the Rimm's introduction to his Georgetown Law Journal article -- Rimm's language is almost an homage to Branscomb:
As Americans become increasingly computer literate, they are discovering an unusual and exploding repertoire of pornographic imagery on computer networks. Every time consumers log on, their transactions assist pornographers in compiling databases of information about their buying habits and sexual tastes. The more sophisticated computer pornographers are using these databases to develop mathematical models to determine which images they should try to market aggressively.
Similarly, Rimm's article, with its discussion of how "power-imbalance" descriptions of imagery (e.g., saying that a woman is "choking" on a penis rather than "sucking" on it) increase the frequency of downloads, seems to provide confirmation of MacKinnon's theories about how porn promotes sexism and the victimization of women. No wonder MacKinnon titled her own GLJ piece "Vindication and Resistance." MacKinnon sees Rimm's study as "vindication" of theories about pornography championed both by her and by her long-time collaborator Andrea Dworkin. This is transcendently apparent in the following passage:
That pornography on computers is part of real life, not apart from it, is made indelibly clear by the Carnegie Mellon study. The content, consumers, and patterns documented in this study are the same as those long observed in the pornography industry and in its tracks through the rest of society.7 The research team documents beyond question the simplest and most obvious, if some of the most contested, facts. Overwhelmingly, it is men who use pornography 98.9% of these consumers, to be exact.8 Even many of the women who use it, Carnegie Mellon found, are paid by pornographers to be there, in order to give men the impression, while online, that women enjoy women being violated.9 Women are disproportionately used in violating ways in pornography. More than ninety-nine percent of all the bestiality pictures studied on the "Amateur Action" bulletin board, for instance, present women having sex with animals,10 in spite of the fact that nearly fifty percent of the pornography studied has men in it.11 The more violating the act, the more women have it committed against them.
Moreover, the more violating the material, the more it is wanted, out of proportion to supply. Sex with children is 6.9% by supply, 15.6% by demand.12 The demand to see vaginal sex (not to be assumed harmless) approaches the vanishing point (6.9% by supply, 4.6% by demand)13, while the demand for women vaginally penetrated by animals on the "Amateur Action" bulletin board approaches fifty percent of all visual bestiality material.14 When a woman is marketed being intensively physically harmed, consumer demand doubles; fellatio gets a lukewarm response, but downloads double for "choking."15
This short passage is buttressed by eight footnotes, citing the writings of three authors: MacKinnon, Dworkin, and Martin Rimm. Indeed, of the 32 footnotes supporting MacKinnon's "Vindication," almost half are citations to Rimm's work.
With censorship opponent Carlin Meyer, Rimm took care to emphasize that the Net can't be effectively censored -- for Meyer, Rimm's article simultaneously proved that a wide range of supposedly nonmainstream images are in fact quite popular and suggested that our culture would have to confront and come to terms with its own appetite for explicit sexual images rather than simply attempt to cover it up through censorship.
For Marty Rimm, enlisting Carlin Meyer as a supporter was very important. There's lots of evidence that Rimm took special pains to persuade Meyer, an outspoken feminist opponent of censorship, that he and Meyer were on the same team. It helps to read Meyer's article with special attention to her footnotes. But before doing so, it's necessary at this point to remember a very different footnote from another article-- Footnote 8 in Rimm's own GLJ article:
'[8] "Study" hereinafter refers to all data collected by the research team, whereas "article" refers only to those findings reported here.'
The nice thing about this footnote is that it allows Marty to cite results of "the Study" without having to explain where those results appeared in "the Article." Hence, except for her first footnote, which simply cites Rimm's article generally, Meyer's footnoted references to "the Study" or to Rimm's work generally seem to be based *wholly on what Marty told her in telephone interviews*.
Here are the relevant quotes from Carlin Meyer's comment on Marty Rimm's article, coupled with the relevant footnotes. ===
'A recent Carnegie Mellon study has confirmed that a lot of people are looking at a lot of explicit sex on computer screens.3 '
3. Marty Rimm, Marketing Pornography on the Information Superhighway, 83 Geo. L.J. 1849 (1995) [hereinafter Carnegie Study or the Study]. As one of a handful of systematic studies of pornography consumption patterns, the Study makes an important and original contribution for all those seriously interested in examining sexuality in American and Western culture. The Study defines pornography to include imagery and text depicting both sexual contact (labeled "hard-core") and nudity or lascivious exhibition (labeled "soft-core"). For purposes of this comment, I adopt the study"s definition. Id. at 1849-50 n.1. ------------
'Pedophilic imagery can easily be described in terms unlikely to be included in a dictionary of off-limits terminology, such as "young girls" or "lovely young girls."86'
86. Although the imagery described in the Study is marketed by the "adult" BBS in quite graphic terms (such marketing takes place even on the Internet and Usenet), it is easy for BBS to change terminology to avoid software screening dictionaries. Amateur Action BBS may have switched its descriptions of some "pedophilic" imagery from "she is spreading" to "girl, naked" to avoid easy identification of illegal child pornography. Telephone Interview with Marty Rimm, Carnegie Study author (Feb. 28, 1995). ------------
'Interestingly, the Carnegie Study never found such descriptors as "snuff," "kill," "murder" and rarely found such others as "pain," "torture," "agony," "hurts," "suffocates," and the like. The term "rape" appeared less than a dozen times in descriptors of over 900,000 images.97'
97. Telephone Interview with Marty Rimm, Carnegie Study author (Feb. 22, 1995). ------------
"Thus, any effort to block access to a particular file, bulletin board, or set of files is nearly impossible to effectuate because such blocks can be easily circumvented. As computer hackers are fond of saying, "for every roadblock, there is a detour."109 "
109.I am grateful to Marty Rimm, Carnegie Study author, for bringing this cyberspace aphorism to my attention. ===
In short, Rimm told Meyer that "the Study" supported two propositions that would necessarily be of interest to an anti-censorship feminist. The first was that the image descriptions didn't seem to be focusing on actual infliction of nonconsensual pain on unwilling women; the second was that censorship, as a technical matter, had already become an impossibility. These were surely revelations for Meyer -- not only did Rimm's claims undercut certain claims made by MacKinnon and Dworkin and the followers, but the sheer impossibility of censorship would mean that society would have to come to terms with its own long-standing yet long-denied interest in sexual images and sexual material. Finally, the Internet, as represented by Marty to Meyer, promised to make it possible for individuals to reclaim the province of erotic imagery from the sadly sterile, unimaginative, and exploitative pornography industry.
Of course, when it comes to the con-man game, Marty is a belt-and-suspenders kind of guy -- he couldn't be sure that the conveniently tailored ways he was presenting his "study" to the law professors would guarantee their endorsement of his work. Which is why he asked at Branscomb and Meyer explicitly to include at least one statement of strong praise for his study. (He may have asked MacKinnon as well, but she is unavailable for comment at this writing.) Branscomb says she had already commended the study in her article even before Rimm casually mentioned to her during a phone conversation how handy it would be she praised it. Meyer, though put off by the effusiveness of Rimm's praise of her work in an April letter, felt at the time that his request in the same letter for a complimentary mention was not too much a favor to ask. (Still, Meyer also recalls being a bit troubled at the time by Rimm's sheer nerve -- he suggested actual wording for her to use in her praise of the study.)
Both Branscomb and Meyer stressed in interviews to me that they didn't intend to be taken as authoritative endorsers of Rimm's methodology -- "I'm not a social scientist, I'm not a statistician," says Branscomb, who simply assumed the methodology was valid since Marty's adviser was Marvin Sirbu, whom she had known when he was at MIT and they had taken a seminar together from communications theorist Ithiel de Sola Pool.
Yet once the law professors could be said, at least arguably, to have endorsed the study, Rimm had much of the ammunition he needed to help Philip Elmer-DeWitt sell the study to Time's editors as a candidate for a cover story. But that doesn't quite explain how Rimm managed to sell Elmer-DeWitt in the first place. After all, the Time senior editor was no law professor, and he had no policy agenda as such. Rimm's problem was straightforward -- what did he need to tell Elmer-DeWitt to win him over?
The clues to his solution to the problem, I think, lie in two interesting quotes from the Hotwired interview of Hotwired interview with Elmer-DeWitt.
Take a look at the first quote, in which Elmer-DeWitt is responding to a question about the sequence of events that led to his decision to push the Rimm study as a cover story. Elmer-DeWitt is relating his first encounter with Rimm, which occurred when the Time writer was working on a piece about the November 1994 censorship flap at CMU (itself triggered by Rimm and his "study."):
Rimm was hard to get a hold of the week that I was doing that story, so I had to interview a number of people, Marvin Sirbu and others, to find out what this study was in order to fully report what had happened. Finally, late in the week he [Rimm] got back to me, and I went over how I had described the study, and he said "geez, I can't believe you got that right!" He ended up being impressed by how I managed, he thought accurately, to characterize his study. A few months later, he emailed me and asked whether Time would be interested in getting first crack at it, when the study was finally ready for publication. So from the beginning, Rimm was talking about giving us an exclusive look at this study.
This is a remarkable quote, in that it tells us just how gifted Rimm was in finding the right buttons to push. For a journalist, there are few compliments more effective than 1) praising one's accuracy, and 2) praising one's insight. What makes the Elmer-DeWitt quotation especially telling, though, is that a review of his Time article concerning the CMU censorship debate reveals it to be *profoundly inaccurate*. Specifically, Elmer-DeWitt's lead paragraph simply misstates what Rimm had claimed to have done:
"The steam began rising for Carnegie Mellon University four weeks ago, when one of its research associates, Martin Rimm, informed the administration that a draft of his study of pornography on the computer networks was about to be released. Rimm had made an elaborate analysis of the sexually oriented material available online. Not only had he put together a picture collection that rivaled Bob Guccione's (917,410 in all), but by tracking how many times each image had been retrieved by computer users (a total of 6.4 million downloads), he had obtained a measure of the consumer demand for different categories of sexual content, some of them, as a faculty adviser put it, ''extremely rough.''
Whatever quibbles one might ultimately have with Marty's research methodology, *it is indisputable that Elmer-DeWitt misreports it here*. For one thing, Rimm had *not* "put together a picture collection" totalling nearly a million images -- the most he claimed to have done was to have collected that many *descriptions* of images. Indeed, the whole thrust (so to speak) of Marty's "study" is analysis of the descriptions, not of the images themselves. For another thing, Marty himself hadn't tracked 6.4 million downloads -- instead, he simply claimed to have totalled the tracking data assembled by the sysops of adult-oriented computer bulletin-board systems.
And finally, of course, the Rimm study ultimately contained little new data about "computer networks" -- instead, the primary source of Rimm's data was, purportedly those same adult BBSs, most of which are unnetworked.
But Marty is never one to let a few inconvenient facts get in the way of a convenient alliance, and it seems likely that his praise of Elmer-DeWitt is what won the Time writer over.
Which brings us to the second, and to me far more interesting, quotation from the Hotwired interview with Elmer-DeWitt.
Quote 2:
HOTWIRED: Is there a link from some conservative group, like the Christian Coalition, to this study?
ELMER-DEWITT: No, there wasn't. Rimm - I didn't even ask him; he volunteered - said "we talked to just about everybody; the one group I didn't want to talk to was the Christian Coalition." He never had any contact with them. There's been a lot of speculation on The Well, that "geez, this must have been funded by the Christian Coalition; they must have given Ralph Reed early access to it." As near as I can tell, that's just the kind of crazy talk that goes on The Well.
Given that there is now strong evidence that the Christian Coalition and other right-wing antiporn acitivists had advance, detailed knowledge of the study and its likely impact (and perhaps even of the Time cover story), one can only wonder why Rimm "volunteered" that tidbit to Elmer-DeWitt. In retrospect, it seems likely that the comment put Elmer-DeWitt off the trail -- his own piece on the Rimm study never mentions the Christian Coalition -- while at the same time convincing the Time writer of Rimm's libertarian bonafides.
If that was what Rimm's intentions were, there's no doubt this tactic worked. But it does give rise to a fascinating question: if Rimm did in fact have prior contacts with the antiporn activists of the religious right -- as he clearly did with the feminist antiporn activists -- one can only wonder how he pitched his study to them.
As yet, none of that segment of the antiporn contingent -- not even Deen Kaplan, the vice president of the National Coalition for the Protection of Families and Children who is also an editor of the Georgetown Law Journal -- are telling.