Would a teenager be harmed by reading Angels in America, the Pulitzer Prize-winning Broadway play? What about viewing erotic Indian sculptures? Or discussing sex in an internet chat room? What harm to minors do various kinds of sexual speech actually cause? These questions have taken on new urgency now that the Supreme Court has set aside the Communications Decency Act (CDA) and the nation is moving toward the widespread use of blocking software to enable schools, libraries, and parents to screen out sexual speech from the Internet. When the Court overturned the CDA last year in Reno v. ACLU, the decision seemed a major turning point in favor of free expression. But much of the new blocking software is based on the unsubstantiated premise that broad barriers to information and discussion about sex serve the interests of young people. Rather than increasing opportunities for kids to learn and talk about sex, America seems poised to close them up.
This trend is especially disappointing because of the shift toward a more nuanced view of sexual speech in the Court's Reno decision. The CDA, passed in 1996, not only criminalized the transmission of "indecent" material to minors over the Internet; it also made it a crime to fail to prevent minors from viewing anything that "in context, depicts or describes, in terms patently offensive as measured by contemporary community standards, sexual or excretory activities or organs." But the Internet is not like a bookshop or a movie theater where one can easily distinguish between minors and adults; the ages and identities of people who visit a Web site or a chat room aren't readily ascertained. So the CDA effectively banned "indecency" from cyberspace altogether. The Court ruled that this violated the First Amendment rights of adults.
The Reno decision has been rightly acclaimed as a landmark, establishing full First Amendment protection for cyberspace, but an important aspect of the ruling has yet to be fully understood. The Court resisted moralizing generalities about harm to minors and recognized that at least some kinds of sexual speech might have value for the young as well as for adults. It noted that the law threatened a wide range of valuable material, including information on safer sex and birth control, human rights reports of rape and torture, artistic images of nude subjects, e-mail between parents and teenagers about birth control, "and arguably the card catalogue at the Carnegie Library." But ironically, while the Court has now rejected the idea that all sexual speech is harmful to minors, our emerging system of internet regulation seems wedded to precisely that view.
SEX IN COURT
By refusing to assume that all sexual speech poses a danger to minors, the Court marked a significant departure from a long history of jurisprudence in which sexual speech could be censored because of the presumed need to protect the young from immoral influences. As recently as 1978, the Supreme Court had ruled in FCC v. Pacifica that the government could ban comedian George Carlin's "Filthy Words" monologue from radio and television—except during late-night hours—in order to protect minors.
The roots of this reasoning stretch much further back into history and across the Atlantic. The 1868 English decision Regina v. Hicklin defined the crime of obscenity as turning on "whether the tendency of the matter charged as obscenity is to deprave and corrupt those whose minds are open to such immoral influences and into whose hands a publication of this sort may fall." This weak-minded category included minors, the lower classes, and women. As Lord Cockburn, the author of Hicklin, explained, the material deemed obscene "would suggest to the minds of the young of either sex, and even to persons of more advanced years, thoughts of a most impure and libidinous character." The Hicklin standard had the effect of suppressing all manner of literary works by authors ranging from Aristophanes and Boccaccio to James Joyce, Theodore Dreiser, and Edmund Wilson.
Hicklin reigned as the American standard for obscenity law until 1957, when two Supreme Court decisions finally interred it. In the more celebrated of the two cases, Roth v. United States, the Court ruled that "obscenity" lacked First Amendment protection because it was "no essential part of any exposition of ideas" and "utterly without redeeming social importance." Sex itself, on the other hand, was "a great and mysterious motive force in human life." The Court reasoned that works with "redeeming social importance" were by definition not obscene even if isolated passages might be thought to "deprave and corrupt" those who are young, feeble, or sensually inclined. Only materials that dominantly appealed to the "prurient interest" and lacked redeeming value could be suppressed as obscene. Butler v. Michigan, the less noted of the two decisions, invalidated a Michigan law that criminalized the sale of "obscene, immoral, lewd or lascivious" materials that "tend[ed] to incite minors to violent or depraved or immoral acts" and that "manifestly tend[ed] to the corruption of the morals of youth." Writing for the Court, Justice Felix Frankfurter explained that the law impermissibly "reduce[d] the adult population of Michigan to reading only what is fit for children."
The obvious response to the Butler decision was for legislatures to pass narrower laws that avoided the constitutional difficulty by specifically banning the sale of sexual material to minors. When such a law came before the Court in 1968, in Ginsberg v. New York, Justice William Brennan crafted a politically shrewd but intellectually dubious justification for upholding it. The New York law in Ginsberg used the then extant definition of obscenity, but adjusted it to ban material for minors that would be permissible for adults. Justice Brennan reasoned that this "variable" obscenity standard meant the banned works were constitutionally unprotected, so that the government did not have to meet its usual burden in First Amendment cases of showing that a "compelling state interest" (that is, protecting minors) justified the law. It was enough that the legislature thought sexual material might corrupt youth.
"Variable obscenity" a la Ginsberg relieved the government of having to prove harm, but it still only allowed the suppression of works found to be "prurient" and lacking in "redeeming value" for kids. The Court's FCC v. Pacifica decision ten years later, permitting censorship of "indecent" on-air language, signaled a seismic constitutional change because it dispensed with the prurience and lack of value requirements of even "variable" obscenity law. Rationalizing the result in Pacifica, Justice John Paul Stevens assumed that indecent speech was harmful to minors and had only peripheral First Amendment protection. Although Pacifica only applied to broadcasting and left the jurisprudence of the print world intact, the decision still marked an important break with the Court's previous rulings. The U.S. government was now empowered to ban constitutionally protected speech, on a theory that subordinated the rights of adults to the nineteenth-century logic of Regina v. Hicklin.
Congress soon began to apply the indecency standard to other media. In the 1989 case Sable Communications v. FCC, the Supreme Court invalidated a law banning "indecent" telephone services while remarking that government does have a compelling interest in barring minors from offensive sexual speech. There was no evidence suggesting what harm would be caused, nor did the Court rely on psychological literature. Instead, it simply cited its Ginsberg decision. But Justice Brennan had crafted Ginsberg on the explicit premise that "variable obscenity" was not constitutionally protected, whereas indecency was. This was no trivial legalism, for the lack of constitutional protection for "variable obscenity" meant that the government could suppress it without meeting the "compelling interest" test. Sliding noiselessly from "variable" obscenity (Ginsberg) to "indecency" (Pacifica and Sable), the Court for the first time suggested that government need not prove any actual need for laws censoring a broad range of "offensive" speech if the justification was the protection of children. While many legal observers thought that a narrow application to broadcasting was inherent in the Pacifica decision, the broad, vague term "indecency" had now been cut loose from its moorings and had become license for censorship whenever the need to protect minors was asserted. Not until Reno did the Court seriously question the broad assumption that minors were harmed by vulgar or sexual speech.
HOW HARMFUL IS SEXUAL SPEECH?
But if, as the Supreme Court suggested in Reno, not all "indecent" speech is dangerous to young people, a number of interesting questions arise. What exactly is the harm to minors that flows from sexual speech? Is it only pornography that causes such harm, or is it any depiction or discussion of human sexuality outside of what socially dominant groups consider an appropriate moral framework? Or is it any explicit information about sex at all—a view that animates much contemporary opposition to sexuality education in public schools? What age groups are affected by what kinds of material, and in what ways?
The answers to these questions are important because they inform not only judicial decisions but a much wider range of policy questions. In the Reno decision these underlying issues remained largely unresolved. Justice Stevens's opinion for the Court assumed that many forms of sexual speech—art, literature, discussions of reproductive politics—would not be harmful to minors. But he nonetheless steered clear of mentioning the testimony of the ACLU's expert witness, Baptist minister and psychologist William Stayton. Stayton had testified that explicit sex information and even pornography do not by themselves cause psychological harm to minors of any age, a view that he said is in the mainstream among sex educators.
In Reno, government lawyers presented no medical expert to contradict Stayton's testimony, but they did provide such expert testimony in a less noticed lower-court case in 1996. Playboy Entertainment Group v. U.S. involved another section of the CDA, which banned (or consigned to late-night hours) audio or visual "signal bleed" of sexually explicit cable television programming to homes that had not ordered them. The government's witness, psychologist Diana Elliott, testified that a study conducted at her child abuse crisis center at the University of California at Los Angeles indicated that children who had seen pornographic television shows were more likely to engage in "sexually reactive" behavior, such as "orally copulating with another child" or "inserting an object" into their own anus or vagina or that of another child.
The three-judge court in Playboy, however, rejected Dr. Elliott's testimony, citing the "anecdotal" and "possibly misleading" character of her methods and conclusions. Nevertheless, this court denied Playboy's request for a preliminary injunction to stop enforcement of the "signal bleed" law. The court satisfied itself on the question of harm by quoting a recently decided Supreme Court cable indecency case and the politically contentious 1986 Report of the Attorney General's Commission on Pornography (the Meese Commission) to support its belief that pornography "is not the best way for children to learn about" sexuality. The judges added: "We believe that the danger of prematurely exposing children to video and audio transmissions of graphic adult sexual behavior is even more troublesome than the exposure to offensive language that was at issue in Pacifica."
These judges' reliance on their own beliefs about child rearing and on the much-criticized Meese Commission is regrettably typical of judicial decisionmaking in this area. At least since the Supreme Court's Ginsberg decision, courts have spoken about harm in terms of "the ethical and moral development of our youth," of the "deep and . . . lasting negative effect" of vulgar words (Justice Lewis Powell's concurrence in Pacifica), and of pornography's not being "the best way for children to learn about sex." Physicians and psychologists also conceive of harm primarily in terms of moral values. They speak of the "bad ideas" that minors might glean from pornography, suggestive language, or even mainstream movies and TV. The implication is that irresponsible sexual messages on television lead to reckless attitudes and behavior.
However, the few relevant social science studies fail to demonstrate a causal relationship even where they find a correlation between TV content and sexual behavior. For example, one 1981 study of 75 adolescent girls, half of them pregnant, found that the pregnant ones watched more TV soap operas and were somewhat less likely to think that their favorite characters would use contraceptives. These findings hardly demonstrate that without racy soap operas the girls would not have gotten pregnant. It may be true, instead, that more sexually adventuresome or irresponsible youngsters are drawn to the racier shows.
The imitative or modeling effects that some psychological theories—and even common sense—suggest might result from young people's watching movies or television have not been proved. The American Academy of Pediatrics acknowledges that "there is no clear documentation that the relationship between television viewing and sexual activity . . . is causal." The evidence is even thinner with specific respect to that subcategory of sexual speech known as pornography. A 1991 field study of juvenile sex offenders reported that although most of them had seen pornography, the primary causes of sexual offenses among these adolescent boys were their own histories of physical and sexual abuse.
Despite this absence of empirical evidence, justifications for shielding minors from sexual speech continued to be based (as in Regina v. Hicklin) on the belief that sexual expression will give youngsters impure thoughts that may lead to risky or immoral behavior. That was the gist of the friend-of-the-court brief filed by the antipornography "Enough is Enough" campaign and other groups in Reno v. ACLU. While mentioning anecdotal evidence about the supposedly unhealthy and addictive effects of pornography on men and adolescent boys, the thrust of their argument was that pornography "is a powerful but deforming tool of sex education" and "encourages neither tenderness nor caring" in sexual relations. Even the 1986 Meese Commission, whose mission was widely understood to be the collection of empirical evidence that pornography causes harm, concluded with regard to minors that the danger "must be seen in moral terms." It explained:
Issues of human dignity and human decency, no less real for their lack of scientific measurability, are for many of us central to thinking about the question of harm. . . . For children to be taught by these materials that sex is public, that sex is commercial, and that sex can be divorced from any degree of affection, love, commitment, or marriage is for us the wrong message at the wrong time.
The Meese Commission's acknowledgment that perceived harm to minors is fundamentally a matter of bad ideas highlights the constitutional problem with indecency laws. Under the First Amendment, the government may not censor ideas simply because it considers them bad or dangerous. In the case of sexual ideas, not only is it difficult to specify what constitutes "harm"; the scientific evidence for such harm remains uncertain at best and inevitably bound up with normative judgments about approved, permissible, and healthy behavior. To regulate a category of expression—even if it could be accurately defined—because of a belief that it will persuade some people to engage in "immoral" and risky behavior is a large step conceptually from the usual constitutional rule that speech can be restricted only in instances where it directly causes discrete, provable harm—for example, slander, invasion of privacy, or the direct incitement of imminent violent acts.
CENSORSHIP GETS PRIVATIZED
After the Court struck down the CDA, a variety of politicians and pressure groups began to suggest new strategies for "protecting minors" from sexuality and other perceived evils they might en counter on line. One strategy has been to draft more narrowly tailored laws that, the authors hope, will withstand constitutional scrutiny. An other approach seeks to replace the constitutionally flawed CDA with classification schemes that would identify and block access to internet venues believed to be offensive, violent, sexually explicit, or otherwise inappropriate for youth. Though distinct from a constitutional point of view, the two strategies share the same underlying assumption that children are always harmed by exposure to "indecent" material.
Commercial filtering products like Cyber Patrol, SurfWatch, and Net Nanny block online sites using one or both of two basic methods. The first focuses on prohibited keywords like "sex" or "breast" while the second relies on tedious and continuously updated site-by-site reviews of hundreds of thousands of Web pages, discussion groups, and chat rooms. The value-laden and sometimes comically overintrusive nature of these programs is by now legendary. Examples range from blocking sites with "sex" in the title (such as "Middlesex") or news articles about AIDS to simply blocking access to the Web sites of the National Organization for Women or the Queer Resources Directory.
But whatever the problems with these commercial filtering products, they at least have the virtue of being voluntary—that is, the private companies do the classifying, and internet speakers are not forced to self-rate. The same cannot be said for "PICS" (the Platform for Internet Content Selection), which is now being widely touted as the nongovernmental way to keep the Internet safe for children. [See Joshua Micah Marshall, "Will Free Speech Get Tangled in the Web?" TAP, January-February 1998.] PICS is not itself a classification system, but a technology that can accommodate multiple self-ratings or third-party ratings of internet sites. Theoretically, parents can choose a labeling system that mirrors their own views. But despite the potential for multiple ratings reflecting different moral values, by the fall of 1997 only three PICS-based ratings systems were in use: Net Shepherd, SafeSurf, and RSACi. The last of these was developed by the Recreational Software Advisory Council several years ago to rate video games. RSACi and SafeSurf call for publishers to rate their own sites under criteria that reflect the same prevailing belief as most of the freestanding software—that is, that the subjects from which minors must be shielded are primarily sexuality and vulgar language. Both systems have calibrated ratings systems that attempt to distinguish among gradations of profanity, innuendo, art, and eroticism.
Net Shepherd, by contrast, uses volunteers from the internet "community" to rate Web sites according to "maturity levels" and "quality" without indicating the content. The company claims to have rated 500,000 sites in a democratic, participatory manner. It fails to note the irony of making access to speech dependent on majority vote in a society that supposedly subscribes to the ideal of protecting expression no matter how unpopular. In any event, the results show the same tendency toward broad censorship as the commercial filtering products. A study done by the Electronic Privacy Information Center (EPIC), released just before a December 1997 Online Summit held in Washington, D.C., and widely promoted by the White House, revealed that Net Shepherd's "Family Search" program blocked between 90 and 99 percent of responsive Web sites when such innocuous search terms as American Red Cross, Museum of Modern Art, National Zoo, or Dr. Seuss were used.
And while theoretically voluntary—that is, internet speakers can choose not to label, and users can choose not to block—PICS-compatible classification systems inevitably create pressures both to self-censor and self-rate. In Britain and the United States, movie producers already negotiate with classification boards in advance about editing their content to achieve a desired rating. Online publishers know that many parents and institutions will choose some kind of filtering scheme. Those who hold out against self-rating and publish material deemed "inappropriate" (among them producers of artistic, literary, sex education, or gay and lesbian sites) will reach a much smaller percentage of their potential audience than they would have in the Internet's earlier, freewheeling days.
Rating and blocking schemes are crude instruments that possess little real value for schools, libraries, or parents who wish to make informed judgments about what internet material their charges should or should not see. Most classifications take little or no account of context or potential educational value. The schemes that do make value judgments are in a sense even more problematic; they tend to block sites that contain unconventional points of view or what is simply controversial information. In either case, rating and blocking will inevitably pressure speakers to dumb down the Internet and create the same drift toward non-controversiality that already characterizes much of the mass media. Net Shepherd's Family Search actually advertises that it "provides families with . . . results that are consistent with the standards of network television programming."
The rush to create a rated and blocked online world has not been without its critics. Since news is by definition likely to include sexual or violent content, journalists have strenuously objected to self-labeling of news sites. (If applied to news, blocking software might well have screened out reports of the Monica Lewinsky affair.) In late October 1997, the Internet Free Expression Alliance was formed by the more civil libertarian of the online advocacy groups, including EPIC, the American Society of Newspaper Editors, the Society of Professional Journalists, PEN, the Electronic Frontiers Foundation, and the ACLU. One of the alliance's goals is to "oppose any government effort to promote, coerce, or mandate the rating or filtering of online content." But this is likely to be an uphill battle against the rating and blocking boosters, given the substantial support the latter are getting from the White House, virtually all of the online industry, many former CDA proponents, and even some of the Reno plaintiffs.
Even setting aside the crudeness of ratings and the inevitable pressures toward self-censorship they create, what will these ratings achieve in teaching young people sexual values? Marc Rotenberg, the director of EPIC, has written that "good parenting is not something found in a software filter." Rather than give the young less information, a better approach might be to give them more, educating them rather than simply slapping labels on forbidden fruit that will only make it more intriguing. Young people will inevitably encounter the "bad ideas" that censors are trying to block. If they are able to see a wide variety of sexual images, from advertising to pornography, they might, with the help of parents and teachers, learn to understand and evaluate their culture's approach to sex, sexism, and gender roles.
At the present political moment, however, things appear to be moving in the opposite direction, and not only in terms of the aggressive marketing of internet rating and blocking schemes. Federal law now mandates abstinence-only, fear-based sexuality education as a condition of government funding. The law forces school districts to suppress potentially lifesaving safer-sex information in order to qualify for federal dollars. Meanwhile, America's teen pregnancy rate continues to be higher than that of any other industrialized country. One prominent sex educator, Michael Carrera, recently lamented that the United States has "slipped 30 years" in its approach to sexuality education. Until that slippage stops and opinion leaders seriously confront the ambiguities underlying the concept of harm to minors, it is doubtful that much dispassionate policymaking will take place, either about speech on the Internet or about sexuality and youth.