5 Va. J.L. & Tech. 7 (2000) <http://www.vjolt.net>
1522-1687 / © 2000 Virginia Journal of Law and Technology Association

VIRGINIA JOURNAL of LAW and TECHNOLOGY

UNIVERSITY OF VIRGINIA

SPRING 2000

5 VA. J.L. & TECH. 7


Erogenous Zoning on the Cyber-Frontier

By Christopher T. Furlow[*]

I. Background of the Problem

II. The Proposal: Restricted Top Level Domain "Cyber-Zoning"

III. Initial Answers and Their Shortcomings

A. Private Solutions: Hardware and Software Content Regulation

B. Failed Government Attempts to Regulate Objectionable Conduct

IV. The Renton Analysis: Reasonable Time, Place, and Manner

V. Keeping It "Behind the Counter": The Ginsberg Rationale

VI. The rTLD Extension and a Possible Labeling Problem

VII. Alternatives and Implementation Problems

A. Competing But Potentially Complementary Philosophy: Lessig and CDA "2.0"

B. Enforcing the rTLD Scheme: The Who and the How Questions

C. The Obscenity/Indecency Problem: Knowing It When You See It

VIII. Concluding Remarks



"The ability of the World Wide Web to penetrate every home and community across the globe has both positive and negative implications--while it can be an invaluable source of information and means of communication, it can also override community values and standards, subjecting them to whatever more may or may not be found online...[T]he Internet is a challenge to the sovereignty of civilized communities, States, and nations to decide what is appropriate and decent behavior."[1]


I. Background of the Problem

  1. Every day, the Internet population grows. One study projects that, globally, the Internet will have in excess of 320 million users by year-end 2000, and a staggering 720 million users by year-end 2005.[2] Of course, the explosion of the Internet is, in large part, due to so-called "e-commerce" business conducted exclusively over the Internet.[3] Nevertheless, the Internet has attracted many who use it as a learning and communications tool, and many of these new "netizens"[4] are children under the age of eighteen.[5] Elementary and high school classrooms are making increasing use of the Internet[6] to supplement inadequate research resources, as are libraries.[7] The result is a new generation of Internet-savvy youngsters, whose ability to navigate the Internet far surpasses their respective parents’ ability to do the same. One study found that, "according to parents, 48 percent of nine- to 12-year-olds are online, compared to 71 percent of 13- to-17-year olds," and that generally nine to 17-year-olds are on line more often than their respective parents.[8]

  2. Most parents would like to be able to regulate the content of the things to which their children may expose themselves, most notably pornography.[9] The Internet is absolutely inundated with pornographic content, which includes either sexual obscenity or sexual indecency, or both. It is certain that the rise in the use of the Internet has provided many benefits to society, most notably with its contributions to education and commerce. Nevertheless, for many parents the Internet is a mixed blessing.[10] While it does give their children access to one of the greatest and most complete learning tools mankind has ever known, it has proved indiscriminate in this regard, being over-inclusive with regard to the things to which it can expose children.[11] In an age where the Internet plays an increasingly vital role to our society and the education of its citizens, it is unlikely that parents will want to forbid their children to use it. Rather, most parents would like for their children to have access to the beneficial materials the Internet has to offer, without worrying that their kids will unwittingly or intentionally expose themselves to sexually objectionable content. Thus, households with both minor children and the capability to access the Internet have been searching for an easy way to filter the content of the web, leaving available to their children the material that is fit for their consumption.

    II: The Proposal: Restricted Top Level Domain "Cyber-Zoning"

  3. In the "real" world, the government may constitutionally require purveyors of sexually objectionable material to verify the age of those wishing to buy such materials or to locate in certain section(s) of town designated for such uses, or both.[12] Such requirements have withstood First Amendment challenges as a result of the government's substantial interest in keeping minors away from such materials, and zoning requirements have been held to further this interest without keeping such materials away from adults who wish to purchase them.[13] This note proposes to evaluate whether a scheme designed to "cyber-zone" the Internet, from the top down, could pass constitutional muster so that it could be implemented and used either to supplant or to augment the effectiveness of current filtering technologies.[14]

  4. Specific Internet sites have discrete "addresses," much like telephone numbers or street addresses with zip codes, consisting of secondary and top-level domain names.[15] The proposal this note seeks to explore is zoning through the use of one or more restricted top-level domains ("rTLDs").[16] Currently, everyone who is familiar with the Internet has heard of "dot com." The termination "dot com" defines a specific generic top level domain ("gTLD") which is itself an area reserved for commercial activities on the Internet. Currently, there are a total of seven gTLDs, including "dot com", reserved for different types of uses.[17] The addition of new top level domain names is not a new idea. Under consideration since at least 1997, there have been numerous suggestions for additions to the root of the domain name server ("DNS"), which administers and recognizes the various "areas" on the Internet.[18] There has been at least one formal proposal to create one or more restricted top level domains ("rTLD"s) into which pornography could be confined, therefore allowing the restriction of the "manner" in which such an area could be accessed.[19] With a rTLD mechanism, parents could more easily and effectively keep their children away from pornographic content.

  5. As recently as 1997, the Supreme Court indicated "that the creation of such [adult] zones can be constitutionally sound."[20] This note seeks to explore the legal and practical obstacles that a rTLD zoning scheme might face. Initially, the note will give a brief outline of previous attempts to prevent accessibility of objectionable material by minors. Part III will begin by exploring the inner workings of private filtering software and standards designed to help parents block speech that is inappropriate for children, and explain why such attempts have failed. Further, the second half of Part III will look briefly at two of the most recent, most important attempts by the legislature to prevent the transmission of certain genres of materials to minors, and why they have been unsuccessful. Part IV will examine how well a rTLD proposal comports with the Renton analysis, the leading zoning case regarding the restriction of "free speech" of this type.[21]

  6. Next, because constitutionality of a rTLD scheme under Renton is not a certainty, Part V examines whether Ginsberg might offer another viable rationale for the government to institute a rTLD scheme constitutionally.[22] Because of the unique nature of the Internet, Part VI examines whether or not a mandatory rTLD zoning scheme could be interpreted accurately as essentially an unconstitutional labeling requirement, and if so, whether a rTLD could still be constitutional.[23] Finally, Part VII scrutinizes the real world value of a rTLD scheme, both with regard to a famous competing solution to the Internet pornography problem, and with regard to certain important unresolved implementation problems that such a move may face. Finally, Part VIII offers some brief concluding remarks on the issues discussed.

    III: Initial Answers and Their Shortcomings

    A. Private Solutions: Hardware and Software Content Regulation

  7. In order to restrict the ability of children to access online pornography and other types of sexually objectionable content, many parents have turned to "filtering" technologies to aid them in their effort. Filtering technologies typically involve software or hardware solutions or some combination of the two.[24] Hardware filtering involves the installation of a physical device into the computer to allow for the filtering of very specific content or to restrict the bandwidth at which information may flow. However, such hardware solutions are relatively uncommon as filtering mechanisms.[25] They are more often employed for security reasons because of their relative permanence, usually requiring a technician to install or remove them, or to simply alter their settings.[26]

  8. Most people who wish to filter Internet content opt instead for the installation of a software program at some point in the chain that connects computers to the Internet.[27] That is to say, software filtering programs will be installed either locally, on home computers (client side) that parents anticipate that their children will or might use to access the Internet, or remotely, at the server level (server side) by Internet Service Providers ("ISPs").[28] Notwithstanding the level at which such technologies are implemented, there are an array of mechanisms used to filter out objectionable content, which can be generally categorized as either "fixed" or more "heuristic" in nature. Still other systems designed to aid in filtering content incorporate aspects of both.

  9. So-called "fixed" solutions are typically the broadest filters, and are set up to filter out content either by blocking entire protocols, entire hosts, or specific pages on a given host.[29] The broadest form of this fixed variety blocks browsers from linking to entire Internet protocols, such as bulletin board services or news groups.[30] However, fixed blocking technologies usually filter Internet content either by blocking entire hosts, or simply select pages on a given host, using a fixed, internally maintained list to prevent children from accessing objectionable material.[31] Host blocking is somewhat narrower than protocol blocking,[32] using a block file list of the uniform resource locators ("URLs") of any and all hosts or sites known to contain objectionable material at the time the list was created.[33] Host blocking filters will prevent the browser from linking to the URLs contained in the block file without proper authorization, such as a password.

  10. These "fixed" technologies, however, have fallen somewhat short of their intended goals. Protocol blocking is far too over-broad, and the problem with filters that use fixed lists is that such lists can never stay current. New objectionable Internet sites are constructed every day, and content from known objectionable sites is easily copied or mirrored to sites with URLs not contained on the list.[34] Users of fixed list filters will need to update their block files constantly as new content appears every day and existing content is moved to different URLs. As this happens, private censors dedicated to characterized content rating must be maintained by companies offering such software to explore (surf) the Internet searching for URLs to incorporate into the block file. Clearly, this process is highly labor-intensive, and updated lists are usually fairly expensive as a result. Because of the continuing growth of the Internet, as well as the increasing speed and ease with which web sites can be constructed and material moved, list-based technologies cannot ever hope to keep pace in any economical way.

  11. Other sorts of software-based filtering solutions typically filter more heuristically, and neither block entire protocols nor rely on a fixed list.[35] These more heuristic technologies tend to stand between a browser and the Internet and are "aware" of "meta-data."[36] Such software searches requested URLs for "meta-data" associated with objectionable content and makes educated guesses about whether or not the requested content is acceptable according to standards preset by the administrator. "Meta-data" is encoded into the web site and may allow search engines and browsers to identify the site with certain subject matter or to attempt to verify assertions that a site may make about itself.[37] Meta-data usually includes keywords for which people may be searching when they wish to view the content of the web site. The meta-data may thus be used to identify the site with keywords. Regrettably, there is currently no state actor or "cybercop" able to police meta-data for its accuracy.[38] Therefore, embedded information cannot always be expected to provide a realistic proxy for the content of the site, and filtering technologies which rely on this embedded information are not wholly reliable as a result.

  12. Finally, there are centralized programs that attempt to standardize a rating system for the Internet that can allow for self-labeling and third-party labeling and generally try to combine the best aspects of fixed and heuristic technologies. The most prominent of such efforts to "rate" sites for content was initiated in 1995 by the World Wide Web Consortium ("W3C") and is called the Platform for Internet Content Selection ("PICS").[39] This project uses a standardized embedded digital label system, created according to a very specific rating scheme, and is used to rate sites much like Hollywood’s voluntary movie and TV rating systems are used to rate programming.[40] Content providers may voluntarily rate their sites using PICS labels, but no authority requires them to do so.[41] Alternatively, using the centralized, cyptographically-secure labeling system PICS provides, third parties may individually tailor a censorship scheme that others (e.g. parents) may use as a model or adopt in full.[42] The PICS program, however, leaves much to be desired. Few content providers who offer objectionable content voluntarily self-label to enable reliable heuristic filtering, and third-party censorship schemes tend to be as under-inclusive as the URL block files used by fixed filtering software.[43]

  13. In the end, the Achilles heel to all such filtering technologies is that they tend to be either over-inclusive, in that they often cut out sites with legitimate value, or they are under-inclusive and fail to filter content from sites that users would prefer to have blocked.[44] As they currently stand, filtering technologies, which have been used to reduce child access to sexually objectionable material, are so inadequate that another, more complete screening mechanism must be sought.[45] To add to the difficulty, such a mechanism must be consistent with the First Amendment and otherwise constitutional.

    B. Failed Government Attempts to Regulate Objectionable Content

  14. Private industry and Internet consortia have not been the only entities seeking a resolution to this problem. Congress too has tried to prevent children from receiving online pornography by enacting legislation that proscribes it. Though there have been other congressional attempts to regulate the transmission of certain types of content on the Internet, two of Congress’ most important efforts to keep children away from pornography have been the Communications Decency Act of 1996 ("CDA") and the Child Online Protection Act of 1998 ("COPA").

  15. The CDA represented Congress’ first attempt to regulate content distribution on the Internet.[46] The CDA sought to criminalize the knowing distribution of "indecent" or "patently offensive" material to minors.[47] The Supreme Court condemned these portions of the act as unconstitutionally over-broad, holding that while the act might succeed in its motives, it would also prevent adults from viewing material that they had a constitutional right to view.[48] The Court found dispositive the fact that identity verification as a criterion for exclusion was crude at best and was unwilling to uphold the act on the government’s promise that technology refinements with regard to reliable age verification would soon render the CDA constitutional.[49]

  16. The next year, Congress returned to the drafting table and wrote COPA, an act whose narrower purpose was to prevent commercial pornographers or pornography providers from knowingly making available to minors such harmful materials for profit.[50] Interestingly, the act grants an affirmative defense to those pornographers who use credit card-based age identification schemes.[51] This is interesting only because the likely result is that COPA will do little to effect its intended goals. First, it does little to affect the current practices of commercial pornography sites, most of which already employ such credit card verification schemes to secure payment for their services.[52] Second, it does not prohibit transmission of pornography by "non-commercial" providers, who represent the bulk of the providers not using credit card systems.[53] COPA has not seen much support from the outset and has already become the object of a preliminary injunction enjoining its enforcement as potentially over-broad, and hence, unconstitutional.[54]

    IV: The Renton Analysis: Reasonable Time, Place, and Manner

  17. The intersection of zoning and First Amendment freedoms was first explored by the Supreme Court in Young v. American Mini Theatres, which involved two 1972 Detroit zoning ordinances that collectively prohibited the location of an adult theater within 1,000 feet of any two other "regulated uses" or within 500 feet of any residential zone.[55] Although a majority of the Court could not agree on a single rationale for its action, the Court nevertheless upheld the zoning ordinances against First and Fourteenth Amendment challenges.[56] Although the adult theaters argued that the zoning ordinances impermissibly encroached on their First Amendment freedoms, the Supreme Court nevertheless held that a state could use the content of adult motion pictures as the basis for treating them differently than other types of motion pictures without violating the First Amendment.[57] The Court seemed persuaded by the "urban blight" argument, which maintained that adult movie theaters tended to denigrate the character of Detroit’s neighborhoods, although this "secondary effects" argument did not win over a majority of the Court.[58]

  18. It was not until City of Renton v. Playtime Theaters that the Supreme Court built just such a majority.[59] In Renton, the challenged zoning ordinance provided that adult theaters could not be located "within 1000 feet of any residential zone, single- or multiple-family dwelling, church, park, or school."[60] Although noting that content-based regulations, enacted for the purpose of restraining speech, presumptively violates the First Amendment,[61] the Court nevertheless found that this was not the case with the Renton ordinance. Rather, retreating from the earlier position in Young that the content of adult motion pictures could validly be used as a basis for restriction, the Renton Court argued that the regulation at issue here was valid as a "content-neutral" time, place and manner regulation. A "content neutral" speech regulation was defined, according to the Court, as a regulation "justified without reference to the content of the regulated speech."[62] The Court noted that such a "content-neutral" regulation was acceptable so long as it was designed to serve a substantial governmental interest and did not unreasonably limit alternative avenues of communication.[63] On the Renton record, the Court concluded that although adult movie theaters were treated differently than other movie theaters, the city’s justification lay not with the content of the speech, but rather with the attempt to prevent the unwanted "secondary effects" that the city determined were incident to such theaters.[64] To bolster this argument, the Court argued that Justice Powell was correct in Young when he observed that if the city had wanted to restrict the message purveyed by adult theaters based on content, it would have tried to close them all or restrict their number rather than narrow the theaters’ choices as to where they could locate.[65]

  19. Under Renton, it would appear at first glance as though an attempt to zone the Internet through the use of rTLDs would pass constitutional muster.[66] Indeed, it seems as though requiring sites containing pornographic content to register solely in one restricted domain would be little more than a content-neutral "place" or "manner" restriction under Renton.[67] However, to assume this outcome is to assume too much, too quickly. To be constitutional under Renton, any proposal to "cyber-zone" pornographic material into a rTLD must further a substantial governmental interest and not unreasonably burden other alternative avenues of communication.[68] Indeed, the Renton Court viewed the city’s desire to eradicate or at least lessen the "secondary effects" of adult theaters, such as prostitution and other forms of street crime, as a substantial governmental interest that was not focused on the content of the adult movie theaters’ speech.

  20. In the Internet context, it is very difficult to understand what colorable, substantial governmental interest could be advanced to justify a cyber-zoning effort as a "content-neutral" time, place and manner restriction. The majority opinion in Reno underscores just how difficult it may be to find a legitimate "secondary effect" that would justify "cyber-zoning." In that case, the government merely advanced that its interest in protecting children from the potentially harmful effects of "indecent" or "patently offensive" speech was enough to justify the CDA’s broad prohibition of the transmission of such speech to minors. The Supreme Court disagreed, saying that an absence of any argument that the CDA was designed to "cyber-zone" to prevent specified "secondary effects" of such speech was what made the CDA an impermissible content-based restriction, not properly analyzed as a "content-neutral" regulation of the time, place and manner of such speech under Renton.[69] In the end, the offensive provisions of the CDA were struck down as impermissibly broad and would not be saved by advancing that the government’s interest in any primary effect of pornography on minors was enough to ensure its constitutionality.

  21. The only "secondary effect" that has been proposed to date is an amorphous, unsupported allegation that the Internet could be expected to lose "legitimacy" if the government does not intervene to cabin pornography into one particular zone.[70] However this bare allegation enjoys no empirical support. Instead it appears that the great economic experiment that is the Internet has continued to flourish some two years after this argument and forecast were first announced, despite both the absence of governmental zoning efforts at any level and the omnipresence of online pornography. As such, a new "secondary effects" argument will be needed to make a regulatory zoning scheme constitutional.

  22. Although the Reno court conceded that the government’s interest in protecting minors from exposure to obscene or indecent speech is indeed legitimate, any attempt to recast the government’s argument for "cyber-zoning" based on a "protection of children rationale" must fail. The reason such a justification must fail lies in the fact that if the government seeks to protect a discreet class of persons from something, the need to afford protection to those persons necessarily implies that the thing against which the government seeks to guard will cause some sort of harm or other adverse effect to the class of persons the government seeks to protect.

  23. Let us assume, for the sake of argument, that exposing children to pornography retards their individual social and emotional growth. Any good story the government could tell regarding the "secondary effects" of exposing children to pornographic material is rendered impotent in light of later Supreme Court jurisprudence. In Boos v. Barry, the Court concluded that "listeners’ reactions to speech are not the type of ‘secondary effects’" contemplated by Renton.[71] Thus, any attempt to tell a secondary effects story, such as the one proposed above, is bound to fail because such alleged secondary effects are inextricably bound up with listeners’ reactions to speech; this, in turn, makes them primary effects. Further, if the argument were recast to allege that the reactions (elicited by the speech we propose to regulate) are themselves adequate to provide a content-neutral basis for regulation, this would fail as well. The Supreme Court would reject this contention based on Forsyth County v. Nationalist Movement, where it held that listeners' reactions to the content of certain speech cannot provide a content-neutral basis for regulation.[72]

  24. This being the case, if a rTLD scheme is to be successful, another way must be found to argue that there are "secondary effects" on the Internet which can anchor a substantial governmental interest, as required by Renton. The best proposal lies in the argument that close proximity to pornography may erode the value, in absolute dollars, of any secondary domain residing in the same gTLD. For an easy, familiar example, take the primary "dot com" gTLD as an example.[73] The value of any secondary domain for a commercial entity is, at one level, whether its "location" is logically correlated to the business it transacts such that it enhances the amount of such business. As such, businesses typically try to use the formula "[business name].com" to denote the address at which they wish potential consumers to seek them out.[74] While a pornographic site will not likely be located at the address of a well known corporation, those businesses in the "dot com" domain might wish to disassociate themselves with sites in the same gTLD purveying pornographic material.

  25. This rationale would have to do with the secondary effects of pornographic material on the volume of business they receive. E-consumers may not wish to inadvertently stumble across pornography while shopping on the Internet. Moreover, just as parents going shopping with their children in the real world would prefer not to drive by or otherwise be near things they do not wish their children to see, so too might Internet surfing parents not wish to "bring the kids" on an Internet surfing session that could include the inadvertent loading of pornographic images or other material.[75] The result might be a diminishment, however large or small, of customers surfing through a particular gTLD. For "dot com" businesses, this would likely mean fewer hits and fewer sales, meaning that such businesses would earn less money.

  26. The logical corollary to this assumed result will be that any "property interest" in certain secondary domains located in the "dot com" gTLD would be worth less to the businesses whose web sites would otherwise occupy them, and thus the property values in the gTLD would be eroded. An argument of this sort, if supportable by some sort of empirical data, would be precisely the type needed to satisfy the "secondary effects" rationale of Renton. In that case, the Court found persuasive the City of Renton’s resolution that the presence of adult movie theaters would have a deleterious effect on neighboring businesses and residences.[76] Admittedly, the Renton zoning ordinance was enacted without the support of any Renton-specific empirical data which suggested that the presence of adult-oriented businesses in close proximity to local residences or businesses would have the "severe impact" alleged by the City Council resolution.[77] However, the lack of written findings of fact by Renton regarding such effects did not trouble the Supreme Court.[78] Asserting that such a standard would be an "unnecessarily rigid burden of proof," the Renton majority felt certain that Renton was entitled to rely on the same sorts of factual findings as conducted by other cities.[79] The Court was content with the city’s right to generalize and adapt the fact-specific findings of other cities to its own situation. This precedent might allow Internet communities to generalize and adapt similar real-world studies to prevent any "secondary effects" argument, such as the one detailed above, from being rejected as simply conclusory and speculative.

  27. Nevertheless, the assumption that Internet "communities" or "cities," as bounded by gTLD restrictions (the city of "dot com"), would be permitted to analogize to real world studies to support "secondary effect" arguments is a tenuous one indeed. Rather, arguments for "secondary effects" on the Internet would likely need to be grounded in data collected from the virtual world, which is a difficult task when identity cannot be readily verified. Still, assuming that Internet traffic patterns can be positively correlated with content, there may still be hope: if network traffic can be parsed to extrapolate what people generally stay away from, then such an argument might succeed.[80] Until then, however, it is difficult to see how any attempt to cyber-zone will pass constitutional muster using this argument. While it is true that a rTLD zoning proposal could potentially meet the constitutional standard under Renton, without written findings to support a secondary effects argument it is unlikely that this will come to pass any time soon.

  28. Another potential "secondary effects" argument would lie with any positive correlation that could be drawn between pornographic content and the likelihood of virus threat. If it could be shown that the presence of pornography is attended in some real way by computer viruses, then there might be a serious argument to cabin pornographic content in one restricted domain. Some viruses have been known to seriously disrupt the networks, servers or individual computers they infect, causing private information to be stolen, and software and/or hardware to be crippled or destroyed. Such destruction of property and theft of information could also be construed as precisely the sort of "secondary effects" the Renton ordinance attempted to eradicate.[81]

  29. Assuming such a "secondary effects" rationale could be adequately documented to reveal something more than just a casual association, the rTLD proposal would likely be constitutional under Renton. In the end, as other authors have noted, the placing of pornographic material into a rTLD does not ban the material altogether.[82] Such material may still be legally published and can still be accessed by those wishing to do so.[83] This being the case, publishers of pornographic content are not restricted from other readily available reasonable avenues of communication. They are not banned from either publishing, performing, or transmitting such content altogether, thus avoiding the problems encountered by the regulations in other important land use decisions.[84] The current proposal simply provides a way to concentrate such material into one "area," much like the Renton ordinance did.

  30. Finally, those who would be affected by the rTLD restriction would have no argument that it foreclosed other reasonable avenues of communication by "placing" them elsewhere. Rather, it would appear that the opposite would be true.[85] The Renton Court was unpersuaded by the argument that commercial viability or "bargain availability" of other real estate on which to locate adult businesses was a meaningful consideration in analyzing whether or not the zoning ordinance had complied with the "reasonable alternative avenues of communication" requirement.[86] Although the Court noted that it has cautioned against zoning regulations that have the practical "effect of suppressing, or greatly restricting access to, lawful speech,"[87] it held that the First Amendment made no guarantees about any speech-related business being able to obtain desirable sites at bargain prices.[88] Any attempt to reargue that point would be fruitless in cyberspace. The area to which pornographic material would be restricted is completely open, would "sell" at the same prices as other secondary domains in the "dot com" domain, and are no "farther" away from browsers than any spaces located in other sister gTLDs.[89] All of these facts work to undercut any allegation that requiring pornographic speakers to remain in one area is any sort of an unreasonable burden on alternative avenues of communication.

  31. In the end, however, precisely ascertaining and documenting "secondary effects" sufficient to support the constitutionality of a rTLD proposal will be a relatively difficult task. The difficulty lies not with the proposal to cyber-zone obscene pornographic content but rather with the attempt to cyber-zone sexually offensive and/or indecent pornographic speech, which may be accorded some First Amendment protection for transmission to adults, but which minors have no constitutionally protected right to view or read. Indeed, although containing the entirety of the inexorable pornography problem may be somewhat difficult under Renton, there remains another viable avenue to pursue for the protection of minors.

    V: Keeping It "Behind the Counter": The Ginsberg Rationale

  32. Even if cyber-zoning were not justifiable on Renton secondary effects grounds, the Supreme Court has nevertheless made clear that government, and more specifically the states, have an interest in protecting minors from pornography. In Ginsberg v. New York, the Supreme Court upheld a New York ordinance that prohibited the sale of pornographic materials to minors.[90] Writing for the Court, and assuming that the legislature had determined that such material would be harmful to them, Justice Brennan articulated two valid reasons why a state could regulate the availability of sex materials to minors that turned on both deferring all child-rearing decisions to parents and the state’s independent desire to safeguard the well-being of its minor citizens.[91]

  33. Beginning in the home context, Brennan opined that, "...constitutional interpretation has consistently recognized that the parents’ claim to authority in their own household to direct the rearing of their children is basic in the structure of our society."[92] In this effort, he concluded that because parents play such a vital role in the upbringing of their minor children for civic duty, they are entitled to the support of laws designed to assist them in this regard (the "Ginsberg entitlement").[93] This support would come in the form of the state making laws designed to prevent children from freely accessing certain types of objectionable material when parents were not around.[94] Against this framework, he had little trouble concluding that the law at issue was clearly valid as supportive of parental control over what material minors could access. He also noted that, in any case, the law did not prevent parents from purchasing such material for their children for consumption in the home.[95]

  34. Next, the Court went further and argued that the state had its own separate interest in providing for the well-being of its children.[96] The state’s interest in this regard is admittedly intertwined with parents’ aforementioned right to the support of the law to aid in parenting, inasmuch as the would help to keep unauthorized material out of the hands of minors when parents are not around. Nevertheless, this rationale was purported to be separate from the first rationale and independently important.[97] Brennan asserted that the state had a further distinct and substantial interest in keeping children "safeguarded from abuses which might prevent their growth into free and independent well-developed men and citizens."[98]

  35. Armed with these two rationales, the Court concluded that the law at issue was constitutional (assuredly as duly grounded in the state’s police power) and did not impermissibly burden any area of "free expression constitutionally secured to minors."[99] The law did not ban the sale of the material entirely but merely restricted the age at which it could be purchased. As such, it avoided the problem of "reducing the adult population to reading only what is fit for children," which was the flaw that had proven fatal to another ordinance that unduly restricted the sale of similar materials.[100] Relying on the New York legislature’s determination that such material was harmful to minors, it was indeed permissible for a state to restrict the sale of pornographic material in an effort both to aid parents and to ensure that children developed into upright adult citizens.[101]

  36. In the Internet context, it seems clear that rTLD zoning would in fact be a mechanism whereby parents could effectively keep pornographic content "behind the virtual counter." New versions of the major competing browsers could be written so as to allow the administrator (read: parent) to password-protect the browser’s ability to enter the rTLD where pornographic material is posted.[102] If a browser were restricted in this way, any attempt to enter an URL in a rTLD, or click a hyperlink into a restricted domain would be met by an "enter password" prompt. Without entry of the correct password, the browser would refuse to link to the site, perhaps even internally logging invalid attempts to do so.[103] In order to avoid any meaningful prior restraint challenges, versions of browsers with this feature should be shipped with the feature disabled, so that no adult end user could later claim that it was "too difficult" to configure the browser to access a rTLD— the browser would be able to link to rTLD URLs upon its initial installation. Ginsberg only says that parents are entitled to laws reasonably calculated to assist parents in their role of parenting, not to assume that role for them. This proposal would have the effect of shifting a portion of the Ginsberg entitlement scheme to parents. That is to say, parents would have to take affirmative steps to activate the browser in this way, but this seems a small price for parents to pay for a more perfect way to filter out pornographic content.

    VI: The rTLD Extension and a Possible Labeling Problem

  37. As is often the case, the solution to a problem is often positively correlated to the way in which the problem is framed. Whatever the outcome of either of the above Renton and Ginsberg analyses, there is the possibility that a rTLD proposal might still raise serious constitutional questions, grounded in a different area of First Amendment jurisprudence. Labeling requirements have long been regarded coolly by the Supreme Court.[104] Requiring a label has been read to violate the negative First Amendment corollary right "...to refrain from speaking at all."[105] Reading a labeling requirement positively, the Court has also said that labeling can have the alternative effect of coercing a speaker to say something that he might not otherwise say, all things being equal.[106]

  38. While, at first blush, it makes perfect sense to analogize the relocation of commercial pornographic content to a specified "place" (a "rTLD") to a type of zoning, this may not in fact be the best way to analyze the situation. Rather, although different domains are very much separate on the DNS, there is no meaningful difference in terms of physical proximity.[107] Any hyperlink can transport a web surfer as easily from one top-level domain to another top-level domain as from one secondary domain to another secondary domain, both of which reside in the same gTLD.

  39. As such, distance is meaningless on the Internet in the traditional sense.[108] One can access a Japanese web site as easily as an American one, for distance does not play a vital role in the ease of Internet travel.[109] Therefore, a rTLD proposal could be interpreted to reduce to a requirement that those who furnish certain types of content "mark" or "label" their sites in a certain way. The rTLD proposal would therefore require the URL to reveal something about the content of the site. By creating a rTLD that is reserved solely for pornographic content, then the URL, which is a visual depiction of the address of the site, would provide the viewer information about the site without the viewer’s browser ever needing to link to that site. That is to say, by looking at a hypertext URL in the browser before connection to the site, the Internet surfer can glean, solely by the TLD extension of the site, that the content there is pornographic. It is in this way that requiring those furnishing pornographic content to register in a specific rTLD will essentially reduce to a labeling requirement.

  40. Assume that a city zoned all of its adult land uses onto a specific street or into a specific zip code. If it were common knowledge in this city that such uses existed exclusively on that street or in that zip code, then perhaps the addresses of the adult establishment could be said to be impermissibly labeled. In other words, if the name of the establishment conveys nothing about its content, then its address certainly would by virtue of the street on which, or zip code in which, it is located. Nevertheless, legal precedent seems to suggest that a city can zone this way, regardless of what the address might "say" about the content or purpose of the establishment.[110]

  41. Yet this analogy fails to account for the fact that in the physical world, addresses and zip codes rarely, and probably never, perfectly correlate with the types of businesses located at or in them. A label, in any real sense, would therefore have to identify perfectly the content of the thing to which it is affixed. Unlike zip codes or street addresses, however, a TLD restricted only to use by purveyors of pornographic content would correlate perfectly to that content, and as such could arguably constitute a labeling requirement for such sites.

  42. Nevertheless, none of this will likely be a concern. No case has ever explicitly held that web site operators have any First Amendment rights in their URL. This stands to reason. Anything that a site might wish to "say" in its URL would be via the registration of a secondary domain named to convey whatever the message might be. Yet, a site operator receives no guarantee that he will be able to register the secondary domain of his choice. For if that secondary domain is already in use, the prospective site operator will have to choose an alternate name. Thus, it is hard to argue that a site operator is free to say whatever he wishes by naming his domain accordingly. He is restricted by the market and, more specifically, by what is still available on the DNS for registration.[111] When site operators are able to register the secondary domain of their choosing, however, the question remains whether a requirement to be located in a rTLD constitutes a labeling requirement which would fail to pass constitutional muster.

  43. Interestingly, due to Internet norms of naming, there may be no problem under this circumstance either. When site operators can in fact get the secondary domain of their choosing, they use their secondary domain as a type of advertising so that Internet surfers might more easily locate their content. The crux of all this is that when site operators can register a domain of their own naming, they voluntarily self-label to take advantage of the URL as free advertising. This fact alone would appear to obviate any concerns that a rTLD requirement reduces to an impermissible mandatory labeling, as most providers of pornographic content would approach the situation with unclean hands.

  44. Still, absent any case law suggesting that site operators have any First Amendment rights in URL naming, the practice of using secondary domain names as a form of advertising suggests something else as well. Advertisements are commercial speech, which receives no First Amendment protection under certain circumstances.[112] The Internet is of course a market of sorts using "hits" as currency, and both commercial and non-commercial providers compete for such "currency".[113] This is because hits on their respective sites correlate with generating revenue for the site from advertisers located thereon.

  45. Secondary domain naming as a form of advertising will likely receive no protection from the First Amendment in the face of a rTLD requirement. Under Central Hudson, the test to determine whether commercial speech receives First Amendment protection came in four parts.[114] First, the speech must concern lawful activity and not be misleading. Second, the regulation must serve a substantial government interest. Third, the regulation must directly advance the state’s asserted interest. Fourth, the regulation must not be more extensive than necessary to serve that interest.[115]

  46. The "speech" at issue here is lawful, indecent sexual speech. Nevertheless, a rTLD requirement would serve a substantial state interest in promoting the parental Ginsberg entitlement, and would directly advance that entitlement through facilitation of user level content screening. Finally, since a rTLD scheme burdens only pornographic speakers, it is not more extensive than necessary to serve the valid government objective of promoting the Ginsberg entitlement. The end result is that due to the admittedly commercial nature of URL naming and the legitimate interest the government could serve by implementing a rTLD, the First Amendment would offer no protection to pornographic site operators.

  47. Much like states can require publishers of magazines containing pornographic content to print a sale restriction on the front cover of each and every issue that goes to press, so too could government legitimately require pornographic site operators to locate in a particular TLD, resulting in something that may approximate a "labeling" of the URL. All such requirements are allowable however, as they facilitate the admittedly constitutional goal of the Ginsberg entitlement. In the alternative, and as mentioned earlier, a rTLD requirement may also be interpreted as nothing more than a content-neutral restriction, applied by the government, and designed to control the manner in which the content is acquired. All of the above suggests that no matter how the situation is approached, a rTLD proposal may nevertheless be constitutional on balance.[116] The burden is minimal, does not by itself stifle or otherwise gag the speaker, and does not ban such material, much less foreclose alternative avenues of communication.[117]

    VII: Alternatives and Implementation Problems

    A. Competing But Potentially Complementary Philosophy: Lessig and CDA "2.0"

  48. The value of implementing a rTLD scheme is immediately apparent: by putting pornographic material into one cyber-zone, parents can more easily administer the access their children have to the content of the Internet. The rTLD proposal creates an easy and inexpensive way to confine pornography such that adults can get into the zone relatively easily, but children cannot do so if parents configure their browsers to restrict access to the rTLD.

  49. Yet the rTLD scheme has not been the only proposal to limit the availability of pornographic material to children. Lawrence Lessig, a noted cyberlaw scholar, has proposed a second generation CDA ("CDA 2.0") designed to restrict access to what he calls "Ginsberg-obscene" material.[118] Under the CDA 2.0, there would be no TLD zoning, but instead a digital certificate[119] administration process run by the Department of Commerce and calculated to screen minors.[120] In addition, Lessig’s CDA would impose civil fines for either knowing distribution of Ginsberg-obscene speech to children or the failure to initiate a screening mechanism based on his proposed digital certificate scheme.[121]

  50. However, Lessig’s proposal suffers from two difficulties which do not trouble the proposed rTLD scheme. The first is a fairly semantic point, which is that the U.S. Department of Commerce’s regulatory authority with regard to the Internet is currently being shifted to the Internet Corporation for Assigned Names and Numbers ("ICANN"),[122] meaning that the former would likely not be in as good a position to function as the certificate authority.[123] The second and more important point, is that Lessig’s scheme appears to suffer from the Coasean problem of assuming no transaction costs.[124] A digital certificate issuance and verification system would undoubtedly be administratively costly, and the effect would be that non-commercial providers could not afford to comply with such a scheme. For even if the cost of setting up and administrating the system could be externalized, such free providers would still have to bear the sunk cost of setting up a verification mechanism and the running cost of maintaining it. This would likely silence some organizational and other free providers and thus validate one of the Supreme Court’s concerns in Reno.[125]

  51. The use of a rTLD might prove to be less expensive both in the short and in the long-term and still manage to achieve the same close tailoring that an age verification scheme would. As noted in a recent article:

  52. Since Lessig’s scheme uses an age verification for narrow tailoring, it eases the current parental Ginsberg entitlement.[127]

  53. By way of contrast, the rTLD proposal seeks to shift the entitlement. By confining pornographic content to one domain, the rTLD idea advances the current entitlement inasmuch as it concentrates such uses. After that, however, the burden shifts to parents to restrict access at the user level by taking affirmative steps to avoid the adult uses concentrated in that area. This shifts the entitlement enough to allow parents, who are better acquainted with their households, to tailor access accordingly and verify age on the user end. The final result is narrow tailoring that is likely much cheaper and simpler to administer.

  54. As a final note, without relocating the pornographers to a rTLD, Lessig’s scheme does little to address their numerous pornographic "banner hyperlinks" or tawdry gateway homepages.[128] Yet when paired with a rTLD scheme, Lessig’s proposal works much better: banner hyperlinks would be legally confined to the rTLD, and gateway homepages could still exist as they do now, only without the risk that children will be able to view them. Implemented simultaneously, both proposals could accomplish together what zoning and age identification schemes do in the real world. Although at one level the rTLD proposal would render Lessig’s digital certificate scheme unnecessary, the digital certificate scheme would in fact provide yet another level of insulation, working to keep pornography out of the hands of children when parents fail to tailor effectively the content to which they want their children to have access. Working in conjunction, the Lessig’s scheme and a rTLD would provide an effective constitutional measure to control the pornography problem on the Internet. To gain access to the rTLD, an Internet user would be required to enter a password at the browser level. Assuming this occurs, the rTLD site visited would request the digital certificate from the browser linking to it. The site would verify the certificate, and allow the authorized user to view the material in one or more directories located there.[129] Finally, paid subscription sites would likely request a credit card number for payment, which could also serve to screen out a large majority of any minors who have circumvented the first two stages.[130] In the event that the CDA 2.0 would not be too costly to administer, use of both schemes could thus potentially provide the maximum level of assistance the law is competent to give parents.

    B. Enforcing the rTLD Scheme: The Who and How Questions

  55. As the administrator of the DNS, ICANN is clearly the best choice inasmuch as implementation of a rTLD on the code side is concerned. This note assumes that governments would not have to compel ICANN to create a rTLD, mainly due to assumptions about the market’s benefit from such an arrangement. Enforcing site operator compliance with the rTLD proposal is another thing entirely though, and it is with this latter notion that government will likely become much more involved.

  56. Both the authority to control the time, place and manner of speech with Renton-style zoning, as well as the "protection of minors" parental entitlement created by Ginsberg, are grounded in the states’ police power.[131] The authority to promulgate and enforce rules on the Internet will have to flow from some deeper font. The authority to see that pornographic content is remanded to a separate Internet rTLD must, of necessity, be much broader than the state police power. Currently, a state’s actions grounded in its police power are only jurisdictionally competent within its boundaries.[132] The Internet, by its widespread use, and to some degree by its architecture, ignores sovereign boundaries the world over. Thus, any state attempt to mandate compliance with a rTLD scheme by domestic ISP’s, or even the Internet at large, would fail. Such an attempt would violate the Commerce Clause, which empowers the federal government to regulate interstate commerce.[133]

  57. This being the case, any effort to force compliance with the rTLD proposal would likely have to come from the federal government, which would be competent to enact laws designed to do this under the auspices of regulating the means and instrumentalities of interstate commerce.[134] Yet even if the federal government were competent to enact such legislation, it could only apply to those located in the United States or one of its territories, or anyone within the "cyber-reach" of an Internet long-arm statute. Still, it is possible that any meaningful effort to enforce the segregation of Internet uses must come at an international level, which could either involve positive comity between nations or a unified international effort.[135] Enforcement mechanisms would include civil penalties for non-compliance, as assessed by whatever government is competent to exercise personal jurisdiction over the site operator in violation.

  58. Enforcement might more logically come from ICANN, however, when such normal avenues fail. As an enforcement mechanism, one proposal has suggested that ICANN could reserve the right to revoke IP rights to secondary domain property for those not complying with the requirement.[136] Such an address revocation could be either temporary or permanent, varying according to the nature and degree of the violation. This note leaves as an open question whether or not the arbitration scheme implemented by ICANN for disgruntled parties to a web controversy is the best option.[137]

    C. The Obscenity/Indecency Problem: Knowing It When You See It

  59. Finally, the prohibition of obscene content is not the problem that a rTLD scheme would seek to address. It is well settled that obscene material receives no First Amendment protection.[138] Material that is merely indecent, however, still receives some First Amendment protection.[139] The problem arises when attempting to distinguish between the two. In order to differentiate the two, the Supreme Court provided a tripartite test in Miller v. California.[140]

  60. Yet in the Internet context, the test is of extremely limited usefulness.[141] Designed before today’s Internet, the Miller test uses "local community standards" as the touchstone of obscenity.[142] The problems in the Internet context engendered by this test are immediately obvious: what is acceptable by local community standards varies greatly the world over, yet all communities are now linked by virtue of the Internet, and all are equally exposed to the content it provides. Any attempt to segregate any type of material, for any reason, requires that there be an immutable point of reference to achieve harmony of the resultant decisions. The current Miller test fails in this regard because the use of a local community standard makes the point of reference highly mutable and useless in a worldwide medium.

  61. It would make a great deal of sense, therefore, to generate some sort of an "Internet community standard", in an effort to be able to allow current obscenity/indecency jurisprudence to be applied in a meaningful way. This suggestion, however, has been summarily rejected by at least one court.[143] Other courts ought to reconsider that position, both to allow United States obscenity jurisprudence to function correctly and to allow worldwide enforcement for non-compliance with the rTLD scheme to be meaningful and give adequate notice to those potentially affected.[144]

    VIII: Concluding Remarks

  62. There never has been, nor will there ever be a good substitute for a live monitor. Even with a rTLD scheme in place, it is impossible to please all of the people all of the time. The architecture of the Internet renders the creation of a perfect filtering mechanism almost sheerly utopian at the present. A rTLD scheme would in fact be an excellent way to assist parents and purveyors of sexually explicit content[145] in preventing children from accessing such media, acting as a cheaper and more easily configured user-side content control measure.[146]

  63. Yet as Lessig and others have noted, cyberspace is a world created and regulated by computer code; as such, perfect segregation of uses in the cyber-world could ultimately be every bit as meaningful as real-world adult use zoning.[147] While it is likely that top-level domain ("TLD") zoning would be a constitutional way to solve— or at least greatly assist in dealing with— the online pornography problem, it is unclear what "secondary effects" rationale the government could proffer to justify a rTLD under Renton. "Secondary effects" would be necessary to support a claim of a content-neutral governmental interest substantial enough to justify "cyber-zoning" pornographic speech into a particular area under Renton.[148]

  64. Notwithstanding a failure to justify rTLD’s under Renton, Ginsberg creates a parental entitlement that would surely suffice to justify the minimal burdens that rTLD requirements would impose on pornographic speakers. These burdens would probably not include an impermissible labeling requirement, as Internet speakers neither have any declared free speech right in secondary domain names, nor any guarantee that such a name will reflect their respective first preferences. Moreover, use of secondary domain names to convey substantive information may be commercial speech not worthy of First Amendment protection in light of the ends a rTLD seeks to achieve, and who it stands to burden.

  65. There are numerous further questions that attend such a scheme, including implementation, effective enforcement, and the call for a new "Internet community standard" for Miller purposes.[149] A rTLD scheme’s virtue lies with its costs and ease of operation and its potential to facilitate the current parental Ginsberg entitlement.[150] While it is true that it would shift some responsibility to parents to achieve its goals, other types of governmental control designed to achieve the entitlement have been less than successful, and in some cases too costly for non-commercial providers.[151] Using a rTLD, all online pornographic content, both commercial and noncommercial, would be kept readily available to adults but could be easily restricted at the user level by parents, who know their respective households best and are best suited to the task.


Footnotes

[*] B.A., 1997, J.D. Candidate 2000, University of Virginia. The author would like to thank professors Jim Ryan, Robert
O'Neil, John Harrison, and Margaret F. Riley, respectively, for their helpful thoughts and commentary on drafts of this
note. Many thanks are also due to Greg Lastowka, Sabra Litherland, Jean-Claude Andre, Andrew Gladin, Christopher
Reed and the editors and editorial staff of the Virginia Journal of Law & Technology for their respective assistance, edits, thoughts and support which were invaluable in the preparation of this note.

[1] 143 CONG. REC. E1633 (daily ed. Sept. 3, 1997) (statement of Rep. Goodlatte). Other authors have also noted the importance and foresight of this quote. See Developments in the Law– The Law of Cyberspace, 112 HARV. L. REV. 1574 (1999) [hereinafter Developments].

[2] See Computer Industry Almanac, Inc., Over 150 Million Internet Users Worldwide at Year-end 1998 (Apr. 30, 1999) <http://www.c-i-a.com/199904iu. htm>.

[3] Id.; see also International Data Corporation, Internet Commerce Will Rocket to More Than $1 Trillion by 2003, According to IDC (June 28, 1999) <http://www.idcresearch.com/Data/Internet/content/NET062899PR.htm> (study projected that e-commerce will top $1 trillion by 2003).

[4] A term which hybridizes the words "net" and "citizen" to create a term which, in Internet parlance, refers to so-called "Internet citizens."

[5] "A joint AOL and Roper Starch study of 500 children between the ages of 9 and 17 found that 63 percent of the youth surveyed would rather surf the Web than watch television, while 55 percent prefer being on the Web to being on the telephone. Children between 9 and 11 years old go online an average of three days per week, while those between 15 and 17 years old go online an average of five days per week." See Henning Schulzrinne, Internet Population Estimates and Demographics, (visited Nov. 20, 1999) <http://www.cs.columbia.edu/~hgs/internet/ population.html>.

[6] See generally President William J. Clinton, State of the Union Address (Jan. 23, 1996) (Clinton pledged that every American classroom will be connected to the Internet by the turn of the millennium).

[7] See U.S. National Commission on Libraries and Information Science, Moving Toward Effective Public Internet Access: The 1998 National Survey of Public Library Outlet Internet Connectivity (last modified Nov. 21, 1999) <http://www.nclis.gov/what/1998plo.pdf> [hereinafter NCLIS Survey](1998 survey of America’s 15,718 public library outlets revealed that 73.3 percent provide general public Internet access).

[8] See Safe & Smart, Key Findings (visited Apr. 2, 2000) <http://www.nsbf.org/safe-smart/key.html>.

[9] See generally Lynne S. Dumas, World Wide Worry, (visited Apr. 2, 2000) <http://familypc.zdnet.com/safety/ security/news/worry/index.html> (Survey indicated that 66% of parents feel that sexually explicit material is a growing problem, and 60% think that it is too easy to access sexually inappropriate content accidentally ).

[10] Another survey conducted by the Annenberg Public Policy Center indicated that while 78% of parents surveyed feared that their children might view sexually explicit material, 59% of parents confessed that, in their respective opinions, children without Internet access are at a disadvantage compared to their peers who do have such access. Moreover, 75% of parents felt that the Internet was a place where children could gain exposure to new things, and 72% felt Internet access helps children with their schoolwork. See Jeffrey Stanger, National Survey Shows Parents Deeply Fearful About the Internet’s Influence in Their Children (May 4, 1999) <http://appcpenn.org/internet>.

[11] That is to say, unrestricted access allows children to come across objectionable material relatively easily. See generally Developments, supra note 1, at 1637 at n.17 (The "reasonable attempt" made by many porn site operators is to warn those visiting the site that it contains objectionable content, and to leave if they do not wish to view the material). See also Dumas, supra note 9 (Higher percentage of teens than parents reported having objectionable experiences online in every articulated category listed, including 69% of teens who confessed to having accidentally come across inappropriate content.)

[12] See City of Renton v. Playtime Theaters, Inc., 475 U.S. 41 (1986); Ginsberg v. New York, 390 U.S. 629 (1968).

[13] The First Amendment provides, in relevant part, that, "Congress shall make no law . . . abridging the freedom of speech, . . ." U.S. CONST. amend. I, § 2.

[14] It is true that at the current time, analogizing principles and jurisprudence of real-world zoning is somewhat difficult, for the same reasons that troubled Justice O’Connor in her concurring opinion in Reno v. ACLU: the twin difficulties of cabining a notion of geography and ascertaining identity. 521 U.S. 844, 886 (Justice O’Connor, concurring in part, dissenting in part). This note is premised on the general agreement among scholars that advances in technology will make zoning the Internet feasible in the future. For arguments both that the Internet does exhibit a type of geography or topography, such that it makes sense to discuss "places" on the Internet, and that zoning is coming to the Internet with advances in identity verification, see Lawrence Lessig, Reading the Constitution in Cyberspace, 45 EMORY L.J. 869, 886-901 (1996). See also Developments, supra note 1, Parts IIA-B, at 1586-96 (generally discussing attributes of "virtual communities" on the Internet).

[15] See infra note 33 and accompanying text.

[16] To be sure, there are other legal authors that have examined this issue in some form or another, with varying justifications and outcomes. See infra note 52 and accompanying text, note 70 and accompanying text, and note 118 and accompanying text.

[17] Currently, the only other gTLDs are: ".edu", ".mil", ".gov", ".org", ".int", and ".net".

[18] In fact, the addition of new gTLDs was one of the primary topics discussed at the March 10, 2000 conference in Cairo, Egypt. ICANN working papers suggest that six to ten new gTLDs should be added. See ICANN, Preliminary Report, Meeting of the ICANN Board in Cairo, (last modified March 11, 2000) <http://www.icann.org/minutes/ prelim-report-10mar00.html> [hereinafter Preliminary Report].

[19] See A. Michael Salim, Proposal to Reserve Restricted TLD’s for Adult-Oriented Domains (January 17, 1997) <http://www.iahc.org/contrib/draft-iahc-salim-restricted-tld.txt>.

[20] See Reno, 521 U.S. at 886. (Justice O’Connor, concurring in part, dissenting in part).

[21] See Renton, 475 U.S. 41, at 54.

[22] See Ginsberg, 390 U.S. at 636 (quoting Bookcase, Inc. v. Leary, 385 U.S. 12).

[23] The labeling referred to would be that of adding a mandatory rTLD termination to the URL. See supra Part II.

[24] See Cynthia Beuselinck, Address at the Odyssey 1998 Conference (Mar. 14, 1998) (transcript available online at <http://www.inet-toolbox.com/toolbox/filters.html>).

[25] Id.

[26] Id.

[27] See Stanger, supra note 10. The Annenberg survey indicated that 32% of parents with online connections are using filtering software, demonstrating what Stanger called " a sign that a substantial portion of parents are taking active strides to shield their children from what parents consider harmful content."

[28] See Beuselinck, supra note 24.

[29] Id.

[30] A "protocol" is a set of formal rules describing how to transmit data, especially across a network. Dennis Howe, Free On-Line Dictionary of Computing, (last modified September 14, 1999) <http://www.nightflight .com/cgi-bin/foldoc.cgi?query=protocol>.

[31] A "host" is nothing more than a computer (or server) that is connected to a network. Id.

[32] Beuselinck correctly notes that protocol blocking is most frequently used to manage limited computing resources. See Beuselinck, supra note 24.

[33] A "URL" is a draft standard for specifying the location of an object on the Internet, such as a file or newsgroup. Howe, supra note 30.

[34] See generally MTI Remote Mirroring Solution Shatters Competitor’s Lock, (Feb. 23, 1998) <http://www.mti.com/press/980223a.html> (press release describing advent of new technology that inexpensively and perfectly mirrors Internet site content).

[35] A "heuristic" is a rule of thumb, simplification or educated guess that reduces or limits the search for solutions in domains that are difficult and/or poorly understood. Howe, supra note 30..

[36] "Meta-data" is "data about data". That is to say, meta-data is definitional data that provides information or documentation of the data to which it refers. Id.

[37] See generally Platform for Privacy Preferences: P3P Project (last modified Oct. 14, 1999) <http://www.w3c.org/P3P> (digital certificates of authenticity may be used by participants to verify privacy of users according to centralized standards).

[38] As a result/example of such duplicity, there are currently numerous unfair competition suits pending in courts regarding misleading meta-data and misappropriation of trademarks by site operators not legally entitled or licensed to use such trademarks. For recent examples of such suits, see Oppedahl & Larson v. Advanced Concepts, No.97-Z-1592, 1997 U.S. Dist. LEXIS 23105, 23108 (D.Col. Dec. 19, 1997); see also Playboy Enters. v. Calvin Designer Label, 985 F.Supp. 1220 (N.D. Cal. 1997).

[39] See Platform for Internet Content Selection: PICS Project, (last modified Oct. 14, 1999) <http://www.w3c.org/PICS>.

[40] See generally Simson Garfinkel, Good Clean PICS (visited Dec. 9, 1999) <http://hotwired.lycos.com/packet/ garfinkel/97/05/index2a.html> (PICS system explained and critiqued).

[41] Ratings systems are intended to be voluntary avoid the prior restraint problem set out by the Supreme Court in Near v. Minnesota, 283 U.S. 697 (1931).

[42] See Garfinkel, supra note 40.

[43] Nevertheless, it should be noted that the Child Online Protection Act of 1998 gave affirmative defenses to commercial online pornographers that voluntarily self-labeled their sites for content. See Child Online Protection Act (COPA), Pub. L. No. 105-277, § 231, 112 Stat. 2681-2736 (1999)(to be codified at 47 U.S.C. § 201).

[44] For the latest round in the filtering wars, see Mainstream Loudoun v. Board of Trustees of the Loudoun County Library, 24 F.Supp. 2d 552 (E.D.Va. 1998) (filtering software used by library was over-inclusive inasmuch as it limited adults’ ability to access legitimate content, and as such was neither the least restrictive means to filter out hard core pornography nor a constitutional prior restraint on such speech).

[45] See generally Maria Seminerio, Report: Filters Too Vigilant, (Dec. 1, 1997) <http://www.zdnet.com/zdnn/content /zdnn/1201/249945.html> (Citing an Electronic Privacy Information Center (EPIC) report that found that a "family-friendly" search engine blocked access to almost 90 percent of materials containing search terms for innocuous, uncontroversial material); EPIC, Faulty Filters: How Content Filters Block Access to Kid-Friendly Information on the Internet (visited Dec. 13, 1999) <http://www2.epic.org/reports/filter-report.html>.

[46] See Communications Decency Act of 1996 (CDA), Pub. L. No. 104-104, tit. V, 110 Stat. 56, 133-35 (1996).

[47] Reno, 521 U.S. at 859-60.

[48] Id. at 885.

[49] Id. at 876-878.

[50] See supra note 43.

[51] Id.

[52] See Timothy Zick, Congress, the Internet, and the Intractable Pornography Problem: the Child Online Protection Act of 1998 32 CREIGHTON L. REV. 1147, 1191 (1999).

[53] Id.

[54] See ACLU v. Reno, 31 F. Supp. 2d. 473, 1999 WL 44852 (E.D. Pa. 1999) [hereinafter ACLU].

[55] 427 U.S. 50 (1976) (plurality opinion of Stevens, J., joined by Burger, C.J., and White and Rehnquist, JJ.).

[56] Id. at 72-73.

[57] Id. at 70-71.

[58] Id. at 71-72.

[59] 475 U.S. 41 (1986).

[60] Id. at 43.

[61] See Cary v. Brown, 447 U.S. 455, 462-463, and n. 7 (1980).

[62] Renton, 475 U.S. at 48 (quoting Virginia State Board of Pharmacy v. Virginia Citizens Community Council, Inc., 425 U.S. 748, 771 (1976).

[63] Id. at 47, (citing Clark v. Community for Creative Non-Violence, 468 U.S. 288, 293 (1984)).

[64] Id.

[65] Id. at 48 (citing Young, 427 U.S. at 82, n. 4).

[66] See supra note 14.

[67] See generally Lessig, supra note 14.

[68] Renton, 475 U.S. at 50.

[69] Reno, 521 U.S. at 867-8.

[70] See April Mara Major, Internet Red Light Districts: A Domain Name Proposal For Regulatory Zoning of Obscene Content, 16 J. MARSHALL J. COMPUTER & INFO. L. 21, 34-35 (1997). The argument seems to be that the so-called "rule of law" confers legitimacy onto that which it is grafted, and that the Internet as a medium will lose legitimacy without the rule of law. She asserts that without governance, Internet content will become shoddy and unreliable.

[71] 485 U.S. 312, 321 (1988).

[72] 505 U.S. 123, 134 (1992).

[73] For purposes of this analysis, I shall assume that "close proximity" can be satisfied merely by location in the same gTLD. While this assumption is clearly open to debate, it seems a logical starting point. For those who are unsatisfied with this assumption, be aware that although defining proximity in terms of similar URLs may be somewhat more valid, that definition presents additional problems that are outside the scope of this note.

[74] Therefore, International Business Machines locates its web site at "www.ibm.com" and Microsoft locates its page at "www.microsoft.com."

[75] See Dumas, supra note 9.

[76] Renton, 475 U.S. at 44. The declining value of surrounding property was indeed one of the purported "deleterious" effects contemplated by the Court in upholding the ordinance.

[77] Id.

[78] Id. at 50-51.

[79] Id.

[80] There are technologies that currently tout the ability to acquire demographic and other personal data about Internet surfers, so that it may be tabulated and sold to corporate and other e-businesses to allow for more targeted marketing. However, there is no reason to suppose that such data could not also be used to support secondary effects arguments of the ilk proposed above. See generally Cyber Dialogue, The Internet Consumer (visited April 2, 2000) <http://www.cyberdialogue.com/products/isg/ic/index.html>.

[81] See Dumas, supra note 9. Of those adults surveyed, 66% said they worried about the abuse of personal information, and 49 percent were concerned about the potential for credit card fraud on the Internet

[82] See Major, supra note 70, at 32.

[83] Note how well that this comports with Justice Powell’s argument in Young. See supra note 65.

[84] See generally Erznozcik v. City of Jacksonville, 422 U.S. 205 (1975) (ordinance prohibiting the display of nudity on drive-in theater movie screen invalidated as facially overbroad); Schad v. Borough of Mount Ephraim, 452 U.S. 61 (1981) (zoning ordinance totally banning coin-operated machines that permitted the viewing of nude dancers struck down as unconstitutionally broad).

[85] See David R. Johnson & David Post, Law and Borders – The Rise of Law in Cyberspace, 48 STAN. L. REV. 1367, 1370 (1996) (arguing that "the cost and speed of message transmission on the Net is almost entirely independent of physical [or virtual] location.").

[86] Renton, 475 U.S. at 53-54.

[87] Id. at 54 (citing Young, 427 U.S. at 71, n.35).

[88] Id.

[89] See Johnson & Post, supra note 85.

[90] 390 U.S. at 636 (1968).

[91] Id. at 639.

[92] Id.

[93] Id.

[94] Id. at 640.

[95] Id. at 639.

[96] Id. at 640.

[97] Id.

[98] Id. at 640-41.

[99] Id. at 637.

[100] Butler v. Michigan, 352 U.S. 380, 383 (1957).

[101] Ginsberg, 390 U.S. at 640-41; see also, M.S. News v. Casado, 721 F.2d 1281 (10th Cir. 1982) (placement of material obscene as to minors in "blinder racks" was constitutional).

[102] See, e.g., Cyber Patrol Partners (visited Dec. 14, 1999) <http:// www.cyberpatrol.com/partners.htm> (indicating that the major competing browsers, Microsoft Internet Explorer, Netscape Navigator, and the America Online browser, currently incorporate features to take advantage of Cyberpatrol’s filtering scheme).

[103] Security or other "access restricting" software typically enables the user to log invalid attempts to bypass the security protocol so that the administrator can later become aware of any and all attempts at breach.

[104] As this discussion does not concern anonymous speech, McIntyre v. Ohio Elections Comm’n, 514 U.S. 334, (1995) (Supreme Court held that such prohibiting the distribution of any anonymous campaign literature violated the First Amendment), and ACLU of Georgia v. Miller, 977 F. Supp. 1228 (N.D. Ga. 1997) (Georgia law prohibiting all electronic communications that did not truthfully identify the sender struck down on First Amendment grounds), are not discussed as clearly distinguishable.

[105] Wooley v. Maynard, 430 U.S. 705, 714 (1977).

[106] See Riley v. Nat. Fed. of the Blind of N. C., 487 U.S. 781, 795 (1988).

[107] See Johnson & Post, supra note 85.

[108] Id.

[109] Id.

[110] See Renton, 475 U.S. at 42. The Renton zoning scheme concentrating adult uses was permissible, and the Court did not allow the First Amendment to be interpreted to require that the city consider the pricing or availability of alternative or preferable real estate on which to locate when it zoned these uses.

[111] Ironically, creation of an entirely new TLD would make it more likely that those wishing to "speak"/advertise through their respective URLs might have a better chance of acquiring their desired secondary domains. The strong, continuing push for more TLDs suggests the truth of this assertion. See Preliminary Report, supra note 18.

[112] See Central Hudson Gas & Electric Corp. Public Service Commission, 447 U.S. 557 (1980) (establishing a four part test to determine whether commercial speech would receive First Amendment protection); Virginia State Board of Pharmacy, 425 U.S. 748 (1976).

[113] A "hit," in Internet parlance, is the record of a browser visiting the site. Most sites tabulate hits to monitor the Internet traffic the site generates. See Dennis Howe, FOLDOC, (last modified September 14, 1999) http://www.nightflight.com/cgi-bin/foldoc.cgi?query=hit&action=search>.

[114] Central Hudson, 447 U.S. at 566.

[115] Id.

[116] See generally, Chris Kelly, The Spectre of a ‘Wired’ Nation: Denver Area Educational Consortium v. FCC and First Amendment Analysis in Cyberspace, 10 HARV. J.L. & TECH. 559 (1997) (finding that any compelled labeling in cyberspace is likely constitutional).

[117] See Sable Communications, Inc. v. FCC, 492 U.S. 115, 126 (1989); accord Reno, 521 U.S. at 844.

[118] Lawrence Lessig, What Things Regulate Speech: CDA 2.0 v. Filtering, 38 JURIMETRICS J. 629, 650 (1998). Lessig coins the term "Ginsberg-obscene" speech to refer to speech which is obscene as to children, but which adults have a constitutional right to view if they so choose. Id. at 666.

[119] The digital certificate would be intended to function as a form of identification. It would either already be embedded in the browser itself, or would be imbedded in the discreet user profile of person using the browser. There are numerous places where such a form of identification could be kept.

[120] See Lessig, supra note 118, at 650.

[121] Id.

[122] Established in the fall of 1998, ICANN was organized to supervise the task of Internet administrative privatization, and will eventually be solely responsible for the addition of new TLD’s and overseeing all operation of the DNS root-server. See John F. Delaney & Robert Murphy, The Law of the Internet: A Summary of U.S. Internet Caselaw and Legal Developments, 570 PLI/Pat 169, 263 (1999).

[123] See generally Generic Top Level Domain Memorandum of Understanding, Latest News, (last modified Aug. 12, 1999) <http://www.gtld-mou.org/>.

[124] See Lessig, supra note 118.

[125] See Reno, 521 U.S. at 856.

[126] See Developments, supra note 1, at 1638.

[127] See Lessig, supra note 118.

[128] "Banner hyperlinks" are nothing more than pictorial hyperlinks. The term "gateway homepages" should be understood to be the page to which a browser travels initially, advertising what is sold "inside" the rest of the site and often including "free samples" of that material. Zick expresses similar concern over the potentially harmful nature of these same sorts of items, which he calls "teasers." See Zick, supra note 52, at 1197-98.

[129] See Lessig, supra note 118.

[130] See Zick, supra note 52.

[131] The police power gives a state the authority to protect or promote the health, safety, morals or general welfare of its citizens. See Brown v. Maryland 25 U.S. (12 Wheat.) 419, 443 (1827); Day-Brite Lighting v. Missouri, 342 U.S. 421, 424 (1952).

[132] See Id.

[133] See American Libraries Ass’n v. Pataki, 969 F.Supp. 160, 167 (S.D.N.Y. 1997) (New York law criminalizing using a computer to communicate with a minor in certain defined ways struck down as a violation of the Commerce Clause.)

[134] See United States v. Darby, 312 U.S. 100 (1941) (commerce power of federal government may be used as the bootstrapping anchor to create a so-called "national police power"). This argument probably still holds after United States v. Lopez, 514 U.S. 549 (1995), where the Supreme Court subjected the Commerce Clause to "outer limits" regarding the scope of congressional legislation it was competent to legitimize. Darby’s concept of a "national police power" derived from commerce power authority may help to quell partially Major’s concern that Internet zoning may not be practicable as the zoning power traditionally resides with the states. See Major, supra note 70, at 32-33.

[135] Other areas of the world have been similarly struggling with controlling online pornographic content, and so the United States would likely find allies to combat this problem. For an example of European Community legislation designed and enacted abroad to deal with these problems, see Council Decision 276/99 of 25 January 1999 Action Plan on Promoting Safer Use of the Internet, 1999 O.J. 276. Further exploration of possible arrangements of this sort is outside the scope of this note.

[136] See Major, supra note 70, at 27. One early IAHC enforcement proposal, which apparently has still not been dismissed, had been to revoke "Internet-based property," namely domain names.

[137] See Internet Corporation for Assigned Names and Numbers, Uniform Domain Name Dispute Resolution Policy, (visited April 13, 2000) <http://www.icann.org/udrp/udrp-policy-24oct99.htm>.

[138] See Roth v. United States, 354 U.S. 476 (1957).

[139] See Sable Communications, 492 U.S. at 126 (holding that sexual speech that is merely indecent, as opposed to obscene, may receive some First Amendment protection).

[140] 413 U.S. 15, 24 (1973). The three-prong test created by the Supreme Court follows: (a) whether the average person, applying contemporary community standards, would find that the work, taken as a whole, appeals to the prurient interest; (b) whether the work depicts or describes, in a patently offensive way, sexual conduct specifically defined by the applicable state law; and (c) whether the work, taken as a whole, lacks serious literary, artistic, political, or scientific value. Id.

[141] It is noteworthy that Zick also takes issue with the current Miller test, but argues instead that the words "whole" and or "value" could be more narrowly or broadly defined to make the test workable. See Zick, supra note 52, at 1198-99. However, while Zick’s concerns have merit, I tend to side with Major here, finding more of a problem with the test’s incorporation of local norms. See Major, supra note 70, at 34.

[142] See Miller, 413 U.S. at 24.

[143] See United States v. Thomas, 74 F.3d 701, 711 (6th Cir. 1996) (concept of Internet community standard for Miller test purposes summarily rejected).

[144] See Patrick T. Egan, Note, Virtual Community Standards: Should Obscenity Law Recognize the Contemporary Community Standard of Cyberspace?, 30 SUFFOLK U. L. REV. 117 (1996) (arguing in favor of an Internet community standard).

[145] In a recent e-mail posting, John Lewis of British Telecommunications indicated support for an rTLD scheme as ICANN moves toward creating new gTLDs. In particular, Lewis pointed to the rapidly approaching possibility that the HDTV (high definition television) digital signal will be distributed on the net, and indicated that this fact alone makes the desire for a rTLD a "significant issue" for him and his firm, as public awareness of the issue is sure to grow. See John Lewis, Re: [BC] Discussion paper on domain names, (Feb. 7, 2000) <http://www.dnso.org/ clubpublic/ga-full/Arc00/msg00090.html>.

[146] Assuming that a rTLD scheme could be put into operation, the author concedes that this would essentially be a simplistic "blocking" technique not that far removed in principle from those described in Part IIIA. Nevertheless, user-defined screening of more select (adult) content would likely pass constitutional muster. This position finds support in other legal literature. See generally Zick, supra note 52, at 1202-03 (arguing that user-based blocking, however it is done, is eminently preferable to any other alternative available currently); see also Major, supra note 70 (broadly arguing that cyber-zoning would be both constitutional and highly effective to accomplish its intended goal).

[147] See Zick, supra note 52, at 1199 ("But make no mistake– zoning is coming to cyberspace. ... The technology of zoning promises to be far more precise than any analog that exists in real space.") See also Lessig, supra note 14, at 888 (Lessig points to a perceived zoning trend that is pervasive, complete and remarkably accurate in its discrimination).

[148] 475 U.S. 41.

[149] Supra note 140.. See also Major, supra note 70, at 33 (Major raises some of the same and still other "issues that remain [to be resolved]" before a rTLD proposal could be implemented).

[150] See generally Developments, supra note 1, at 1638.

[151] Interestingly, at least one survey indicates that only 29 percent of parents (and 24 percent of teens) think that the federal government should even attempt to regulate Internet content. See Dumas, supra note 9.