Features – Thinking About Linking, Part II: Can Law Accommodate the

Bradley J. Hillis is a member of the Washington state bar, and the author of “Internet Experiments in Electronic CourtFiling,” “Considerations When Placing Court Opinions on the Internet,” and “Legal Research on the Internet: A Simple, How To Guide.” He lives in Bellevue, Washington, and is a legal analyst for the Office of the Administrator for the Courts.

Go to Part I

Technology Fixes to Framing and Linking

Some believe that we can better address the need to limit links and framing through technological fixes.9 With framing, one can use a Java script to defeat the ability of another site to frame another. The Java script causes a site to expand to fill entire computer screen of the user, eliminating the navigation frame of the first site. 10 The drawback is that the Java script prevents the original site from using frames because the script works against the host as well as the unwanted framer.

Another fix lies not with the publisher but with the user. The user can defeat frames by typing the uniform resource locator (URL), or Web address, into the browser title bar. The URL is found either in text on the Web page or in the source code. However, it requires some sophistication on the part of the user to employ this technique. In the newer browsers, the user can also simply open the frame in a new window, by left-clicking on the frame and selecting the appropriate command.

Linking is easy to do with standard HTML pages, but is not possible with pages built on the fly with Common Gateway Interface (CGI) scripts or active server pages (ASP). If a page is constructed from discrete components by CGI or ASP technology, the framing site would not have a set URL on which to anchor. Thus, publishers can use this technique to force visitors to go only to the initial gateway page, at the designated domain name.

The drawback is that some companies may want to allow individual visitors the convenience of bookmarking a page of particular interest to encourage customer use of the site. I may visit the L.L. Bean site today, find the flannel shirt I want in the catalog, then bookmark it to return the following day and make my purchase. By defeating the ability of another site to anchor to a specific page, L.L. Bean would unintentionally diminish the chances I would find that shirt again, and might lose a sale.

A newly emerging technological fix lies with digital signatures. New technology allows a publisher to sign an article, and lock the text head-to-toe so that it cannot be altered, or to forgo the signature and simply freeze the content as published. This approach would prevent customization of media. Watermarking of digital images is a way to prevent another entity from taking media, or at least from taking it anonymously. New technologies even promise to interfere with the ability to copy the images at all.

The current use of Adobe Acrobat Portable Document Format (PDF) is sometimes based on the difficulty of a user to alter the text, though cross-platform viewing is the more often cited reason. More precisely, Adobe offers the ability for the publisher to freeze the content of a PDF file. But even if the publisher does not impose this limitation, the practical effect is that the user would have to reprocess the PDF file to a postscript file to alter the text, since “cut and paste” does not work with a PDF file.

Finally, search engines can help copyright owners locate other sites that link to or frame protected content. The copyright owner is thus able to contact the infringer and request that the improper activity cease, under threat of litigation. Some copyright owners will place hidden marks, such as two blank spaces instead of one, between words, to uniquely identify their property, and bolster the ability for search engines to index other publication of the text.

Each of these technological fixes is effective in limiting violation of copyrights but does not replace the need for proper legal rules. Just as we have door locks and still make burglary a crime, in cyberspace there are both technologies to protect intellectual property and rules defining proper use.

Requests at Destination Sites Not to Link or Frame

Of course, one can simply ask not to have other sites link or frame material. One of the best known legal sites, the ‘Lectric Law Library, asks visitors to link only to the front page of the site. The concern is that other pages would deep link directly to files. In this way, an outside site could have a direct link to the Federal Rules of Civil Procedure in a compressed zip file, or a legal form. The ‘Lectric Law Library seeks to control the visitors to its site by requesting the cooperation of other legal Webmasters.

At Reuters News Agency, the copyright notice prohibits the framing of the news releases. The notices provides: “Copyright © 1998 Reuters Limited. All rights reserved. Republication or redistribution of Reuters Limited content, including by framing or similar means, is expressly prohibited without the prior written consent of Reuters Limited….” 11 Reuters specifically prohibits framing in its copyright notice to avoid reliance on a legal doctrine that would limit framing of any copyrighted information. The Reuters statement helps the company both by clarifying the rules of use for purposes of any litigation, and more generally letting users know that the company prefers it not be framed.

While the statement is strong, it ignores the reality that users will readily copy an interesting story to email to friends, or will link to a specific story from a Web site on the same topic as the story, or will even readily copy the story to a Web site archive, since Reuters may post the story for a limited time. Apart from these uses, all arguably as illegal as speeding on the Pasadena Freeway, the Reuters statement ignores the requirement of the user’s computer to cache a copy of the text to view it in a Web browser. Nevertheless, attorneys should advise client companies to include a Reuters-type boilerplate clause against framing in copyright notices.

What Does Not Work at All With Existing Legal Rules: Search Engines, Frames, Links and Basically Everything

Perhaps one has decided that the current model of legal rules for linking and framing work better than is suggested above. According to this view, it is adequate to have a rule that prohibits linking between sites, or the use of frames to repackage the content of another. However, if current rules are adequate, there are some unexamined consequences that prove troubling. First, search engines, such as AltaVista, violate practically every aspect of the current legal model. A search engine provides no content of its own and is arguably as parasitic as TotalNews. Also, a search engine delves deep into a Web site to index words and graphics appearing throughout the site, and then returns links to the precise page where the term appears. A search engine is arguably in violation of the “no deep linking” rule that emerged from the Ticketmaster case. At the Lycos search engine, one can look for images and view a thumbnail before going to the site where the graphic is located. It is practically impossible for an Internet search engine to obtain cross-link agreements with all other sites–that would mean the entire Net.

To prohibit page-specific links at a search engine would undermine the ability of users to locate online information. Until just recently, a link to the gateway page of the U.S. Social Security Administration would do no good if one was searching for Year 2000 information that lied buried four levels down a rabbit hole–it took hours to locate the information through the main gateway page, if it could ever be found at all.

Bookmarks are the real nail in the coffin of a rule against deep linking. Bookmarks are simply links accessed in the menu of a browser. They provide a convenience to the user. One cannot limit how an individual uses bookmarks because they exist only on the desktop. We are left, then, with the uncomfortable choice of saying bookmarks are acceptable because we lack the ability to police them, or even worse that bookmarks to pages deep within a site are illegal.

What is going on here? We tend not to question the copyright status of search engines or bookmarks because the function is essential to the use of the Internet. We all like AltaVista because it provides a convenience. Yet it is commercial and draws visitors for the same reasons that metasite like TotalNews or Sidewalk draws visitors.

Arguably, the problem with framing is not in its abuse but that it exists. Inherent to framing is the phenomenon of repackaging the content of other sites. The problem with the rule against framing is that we have the ability to frame. A good example is the Washburn Law School site, one of the best online law libraries. It uses frames to provide a navigation tool to visitors. However, the effect is that the legal material one views is “framed” by the law school site. Is Washburn commercial since the school charges tuition to attend? If the tuition charged is less than Harvard, does that make Washburn less commercial than Harvard? If I start a non-profit organization, appoint myself director with a six-figure salary, sell advertising at my web and attract visitors by framing another’s content, am I then non-commercial? If I have a for profit business and lose money each quarter am I non-commercial? Of course not.

The same examples apply to linking. Why exactly is it lawful to link to the main page of another site? The reason given is that the Uniform Resource Locator (URL) address is akin to a telephone number or street address, which are not protected under traditional copyright law. But this view fails to explain why deep linking simply takes another, longer URL address at the same site and so deserves equal freedom of use. Taking Washburn again, I use the site precisely because it deep links to the exact legal information I am seeking. I would not use the Washburn site if Washburn linked to the gateway page of the State of Florida, instead of to the exact page where the constitution is located. Perhaps the problem with the rule of deep linking is that we have the ability to link.

A listserv thread about the TotalNews case between Phil Stripling and Dan Burke, Burke asked if a Web browser is framing Web content.12 If frames are impermissible, why is Netscape allowed to have it’s “N” with streaming comets in the top right corner of its browser? We do not deem software to frame content impermissibly because the browser is essential to view content on the Web. The legal rule that browsers framed content would result in undermining the Internet and so we do not take this idea seriously. Perhaps, though, we should in an effort to appreciate the dangers in limiting the ability of HTML to allow users to shape information on their desktop.

If a user opened three browsers as separate windows, and then positioned each window to occupy a portion of the computer screen, it would look an awful lot like frames. If we take the position this is not a copyright violation as to the displayed content, can we allow a software developer include a menu command in a browser to “create three windows (frames)” in which two windows host news content, say from ABC and MSNBC, while the third hosts advertising from the browser company. Doesn’t that look like the TotalNews site?

Taken a step further, we could say the computer monitor frames the content that appears on the screen. Applied to television, we can examine the new double window feature that allows a viewer to open a small window to view a second show. This way, a television viewer can watch a football game and keep track of the golf match on a different channel at the same time. Is this impermissible framing of content? If so, by whom? The television manufacturer, or the user?

The underlying point of these problems is that technology has outpaced the legal rules of copyright that applied with reasonable efficiency in the past. With framing or linking, or any number of issues in cyberspace copyright, we can apply existing rules to solve a particular problem. But if we extend the old legal rules to their logical conclusion we end up with absurd results, that is, we will both hurt freedom of speech and limit the ability of property owners to exploit the commercial potential of cyberspace. It is nobody’s interest to strip property owners of their right to control their information because a “free for all” on the Net would inhibit development of additional online content. But any legal rule that emerges must accomodate the nature of the Internet.


The law of linking between Web sites, discussed in several cases and commentaries over the last year, is based on what at first seems like reasonable commercial rules, but upon further analysis fails to accomodate the future of the Internet. One can draw distinctions between Washburn and Sidewalk, or between Washburn and TotalNews, but we must avoid shallow distinctions that fall apart under impassioned, logical scrutiny. We have a problem in online law that scholars are only beginning to wrestle with. For new media rules to succeed, lawyers must develop a jurisprudence that recognizes the empowerment of the individual by Internet technology.

There is a tension in online law. Intellectual property owners need a stable set of rules to make rational planning decisions. Society needs broad freedoms of speech to ensure the democratic potential of the Internet finds fruition. Whatever happens, do not blame the Internet for its nature of open and free association of ideas, and do not throw out wholesale the ability to link or frame in a desperate effort to prevent illegal infringement of valid commercial expectations.

The cases that arose in 1997 can be read to imply there is no right to link to another Web site, or that the link must lead only to the front page of another site, or that framing is impermissible. In other words, a click on the current law of linking between Web sites leads to an unstable set of expectations that tomorrow will inevitably prove an error 404 dead end. Online lawyers need to recognize that the Internet is a new medium for sharing information that requires creative thinking to establish sound rules of use. The legal issues of derivative works, rights of publicity and dilution of trademark are adequate legal theories if applied correctly to new media technology. But their application to the technological innovation of the Web too often fails to acknowledge that way that electronic information is shared on the Internet.

Law that adopts a simplistic rule that accommodates short term expedience does society no favors, and leaves copyright holders with a false sense of commercial security. Boundaries in cyberspace between commercial and non-commercial use are sometimes difficult to discern. The protection for individual speech and creativity, on the one hand, and protection for copyright holders, on the other, require new legal paradigms. Perhaps it is the destiny of Lawrence Lessig of Harvard, Eugene Volokh of UCLA, and other scholars, to forge a set of rules that explains to society the Internet is naughty by nature, not because it hates you.


Research for this article drew on the case law collection of Robert Cumbow, with the law firm of Perkins Coie, Seattle, Washington, and the articles at the Law Journal Extra, http://www.ljextra.com/.


9. Weiss notes in ““Metasites Linked To IP Violations,” supra note 6, that: “One might argue that technical barricades, rather than reliance on legal principles, should be the primary means of exercising site control. But technical barricades are only viable until some clever developer figures out how to circumvent them. In the long run, Web site operators will need to rely on a range of copyright, trademark and other legal enforcement. This first wave of cases has not yet provided any reliable precedents, but it has certainly offered a glimpse of these emerging issues.” <back to text>

10. Listserv post by Cliff Kurtzman, President, The Tenagra Corporation, to the Online Ads listserv, February 25, 1997, “You can reduce the susceptibility of having this happen to your pages by incorporating the following code into your page’s javascript:

if (top.frames.length 1) {
top.location=”/”; }

<back to text>

11. Reuters copyright found at http://www.internetnews.com/Reuters/china.html.
<back to text>

12. Listserv thread discussing TotalNews framing case, June 6-10, 1997, between Phil Stripling, intellectual property attorney, San Mateo, California, and Dan L. Burke, Seton Hall Law School. <back to text>

Posted in: Cyberlaw, Features, Web Management