Web Site Linking: Right or Privilege

As the World Wide Web continues its exponential growth, disputes about the proper way of linking from one Web site to another will become more heated and conspicuous. Although there have already been several major linking controversies such as the Ticketmaster and Shetland News cases, there is little legal precedent about this matter. Further, the moral propriety of linking in different forms remains a nebulous and largely unexamined topic.

In this concise paper we will seek to explore some of the moral dimensions of this issue. After clarifying the basic mechanics of linking, we will briefly review two major linking controversies which have emerged as paradigm cases, and which can serve as a concrete guide for our discussion. We will then consider some of the applicable legal principles such as copyright infringement and trademark dilution. This will set the stage for an explicit moral analysis of some of the potential problems triggered by linking from one site to another. Although we fully appreciate the dangers inherent in propertizing the Web, we will argue that the "right" to link to another Web site must be tempered and limited by the fact that a Web site is a form of intellectual property. Therefore, its author and "owner" should have the prerogative to control who links to that site and how they do so. We will contend, for example, that if a target Web site is opposed to "deep linking" that wish should be respected. On the other hand, all Web site authors must respect the fact that linking is the heart and soul of the Web. Therefore any arbitrary or excessive restrictions against linking should be eschewed for the sake of the common good of open communications and maximum porosity in the Internet environment. It is obvious that this argument turns to a great extent on the notion that Web sites are property, and thus part of this analysis will seek to develop and defend that position.

Our main concern here will be the acceptablity of deep linking without permission and the importance of respecting copyrighted material in the linking process. There are many other aspects about linking such as one's liability for linking to a site with illegal material, but those complex topics will be left for another time.

A Web site refers to text, graphics, or media content that has been put into an area of the Internet known as the World Wide Web. Each Web site has a unique address or URL (Universal Resource Locator) such as www.bc.edu. A Web site typically consists of many pages which are organized and controlled from a beginning or "home" page.

Quite simply, a link is a connection between two different Web sites. A hyperlink within a Web page contains the URL for another Web site which is activated with the click of the mouse. While most links take the user to the other Web site's home page it is possible to bring the user to subordinate pages within the Web site. This practice has become known as "deep linking." The hyperlink text itself can take many forms. It can be the name of the linked to Web site or a description of what is to be found at that Web site (usually the name or description appears in highlighted text). It can even take the form of a revealing graphic or image such as a company logo.

There are two types of links: an HREF link which instructs the browser to locate another web site and provide its contents for the user, and an IMG or "image" link. An IMG link instructs a browser to enhance the text on the user's Web page with an image contained in a separate file usually located at a completely different web site. For instance, if one is writing a narrative on Monet, the French Impressionist painter, one can include an image of a Monet painting from the Boston Museum of Fine Art's on-line image file to provide an illustration that will accompany that narrative. Unlike an HREF link the user cannot really discern that the image is coming from another source since there is a seamless integration of text and image. It is therefore more customary for users to seek permission for an IMG link. On the other hand, there is a general sentiment on the Web that it is not necessary to seek permission for an HREF link.

The <A HREF tells the browser that this is a link to another Web page to which the user should be taken. The address or URL of that Web page (http://www.bc.edu) follows. And the words "Boston College" represent the text that appears on the screen -- this is the only thing which the user actually sees. When the browser encounters this code it goes out to the Internet and locates the Boston College Web page. It then sends a copy of the Web page to the user's browser.

The value and social benefits of linking are manifold and beyond dispute. Most web pages have multiple links to other web pages and they are also the target of many other links. Links from one Web site to another permit instantaneous access to multiple sources of information. Linking allows users to easily follow a complex and intricate trail of research. Moreover, each user can determine how extensively or deeply to follow that trail. Linking is the essence of the World Wide Web, and there is little doubt that legal or technological constraints on linking would have substantial negative ramifications for the Web.

Although most users concede that there is nothing wrong with linking even without getting the target site's permission, there are problems with the way in which some source sites do their linking. Some of the more prominent and common problems are highlighted in the following two case studies.

In 1997 Ticketmaster Group Inc. filed suit against Microsoft for federal trademark infringement and unfair competition. The facts of this case are clear enough. Microsoft operates a Web site called Seattle Sidewalk which functions as a guide to recreational and cultural activities in the Seattle metropolitan area. Seattle Sidewalk provided many links to related Web sites including a link to Ticketmaster, which operates a popular ticket selling Web site. That link, however, bypassed the Ticketmaster home page and went directly to the respective pages for purchases to events listed in the Seattle Sidewalk page. For instance, a listing on the Seattle Sidewalk page for the Seattle Symphony would include a direct link to a Ticketmaster sub-page that would allow users to purchase their symphony tickets.

This is a prime example of "deep linking," and Ticketmaster raised a number of objections to this practice. According to Ticketmaster, by bypassing their home page Seattle Sidewalk users were not being exposed to the extensive advertising and announcements which were posted there. In Ticketmaster’s view, this diminished the value of that advertising and ultimately the rates that could be charged to future advertisers. Ticketmaster also complained that Microsoft was able to generate advertising revenues on the basis of this link because Microsoft posted a banner advertisement on the same page it displayed the Ticketmaster name and link. And, according to Wagner (1997) Ticketmaster alleged that the links were done in such a way that they “presented information incorrectly and out of context.”

This case certainly raises the fundamental problem with deep linking which circumvents advertising and other identifying or promotional features on the home page. Deep linking not only reduces the value of the target site's advertising but deprives that Web site of its proper exposure and recognition.

On the other hand, it could be argued that Ticketmaster has no real quarrel here. Its Web site information is at least quasi-public. Ticketmaster therefore should not interfere with the Web's free flow of information. In its legal defense Microsoft argued "that Ticketmaster breached an unwritten Internet code in which any Web site operator has the right to link to anyone else's site" (Tedeschi, 1999). Microsoft also argued that it had a First Amendment right to publish this public information.

There was an out of court settlement to this lawsuit in February 1999. Although the terms of the settlement were not disclosed Microsoft did agree to link to Ticketmaster's home page instead of to its sub-pages. The settlement was actually a disappointment for those searching for a firm legal precedent about controversial linking activities. As a result, at least in the United States there are curently no unambiguous legal guidelines on the practice of deep linking.

In a similar case the Sheland Times newspaper in Scotland sued the Shetland News, a news oriented Web site, alleging that the News linked to the Shetland Times Web site and displayed its headlines without any attribution. The Shetland News is an online daily newspaper which posted links to the Shetland Times Online. The News used actual Times headlines as the text of the hypertext link. For instance a link on the Shetland News home page might be "New Prime Minister Elected," which would link directly to a story in the Shetland Times with the exact same headline. The link also bypassed the Shetland Times home page which had copious advertising. In October 1996 Scotland's Court of Session banned the links, "finding it plausible that a headline is a literary work and that the News's practice of incorporating the Times's headlines verbatim in its link lines violated the latter's copyright" (Kaplan, 1997). The Scotish Court did not rule on the matter of deep linking.

After the Court's ruling the Shetland News terminated its links to the Times Web page instead of seeking a legal truce. The feud continued, however, even after the News terminated its links. The Time's posted a story on its Web site that remained for many months describing the New's practice as a "parasitic activity." The News continued to proclaim that it had done nothing wrong. In their view they were merely providing the Times with additional readers.

As one might expect, and as these two cases imply, the law in this area is still rather inchaote and ill-defined. There are, however, legal principles that appear to be applicable to linking disputes. One such principle is copyright infringement. A copyright can be violated in a number of ways. For example, the copyright holder has the only legal right to reproduce the work in question (aside from instances of fair use). This aspect of copyright law probably does not apply in most linking disputes. For example, Microsoft really did not copy anything from the target Web sites. On the other hand the Shetland News did copy the Times' news headlines, and that could be construed as a copyright violation. From a legal standpoint, therefore, copyrighted material such as artwork or headlines should not be used to identify the hypertext link.

Copyright also confers the exclusive rights to display and distribute the copyrighted material. The right to display is probably not infringed by linking. According to Jakab (1997), "the display in this context is probably best understood to be consisting in the copyright holder's site from the moment the content is place on the Web, not in the user's browser; if the act of browsing were to implicate the display right, there would be little to do on the Web." One could argue that Microsoft's Seattle Sidewalk is "distributing" Ticketmaster's Web page so that its users can conveniently purchase their tickets. Legal scholars contend, however, that any appilication of such a broad interpretation would have devastating consequences for the Web, and hence it is unlikey to prevail in any Court of Law.

Another legal principle which may have more applicability is trademark infringement or trademark dilution. Trademark infringement arises when one uses the trademark of another in a way that creates confusion for consumers or in a manner that is deliberately exploitive or deceptive If a new athletic shoe company used the Nike "swoosh" trademark to help sell its shoes there would be a clearcut case of infringement. A link that leads an end user to falsely believe that a Web page is affiliated or approved by the target Web page could be a candidate for trademark infringement. An example could be an IMG link that puts another's trademark on the source Web page to create the impression of an endorsement. Deep linking per se, however, does not necessarily imply such infringement.

Trademark dilution occurs when a trademark is ridiculed or overused by others in such a way that the value of the trademark is diminished. Does the use of a trademark or logo to identify the hypertext link constitute dilution? There are as yet no legal decisions on this matter, but it appears difficult to make the case that this would really cause any serious dilution.

In conclusion, although certain linking activities such as using another's copyrighted material as part of the hyptertext identification or even loading in image files without attribution may violate current copyright laws, other activities such as deep linking do not seem to be covered by such laws. This creates a vacuum which can certainly be taken advantage of by Web site publishers.

There is little doubt that linking is an important activity that yields many social benefits for the World Wide Web and its millions of users. There is also no doubt that there is some legal ambiguity about the proper way to conduct linking activities. But that does not mean that users can link with impunity or without considering the moral implications of their actions. This brings us to the central question of this paper: how can users link with integrity ? Even if deep linking is legally permissible at this point thanks to the lack of a clear precedent, should it be morally acceptable? Are there other aspects of linking that are problematic from a moral point of view? Or is there an intrinsic right to link that supercedes all of these claims?

The notion that users have a right to link to any portion of another web site, i.e., the home page or a subordinate page, is certainly widespread. Lawyers and jurists, for example, who defend this position argue that one's mere presence on the Web is an implied license to link. Further, if there is concern about loss of advertising revenues by linking to sub-pages instead of the home page, the remedy is simple: place ads on the sub pages as well the home page!

To some extent, the moral resolution of this issue depends upon whether or not a Web site is common or private property. When authors create Web sites and put them on the World Wide Web do those sites in effect become part of the Internet commons and does this give others an implied license to link to those sites in any way they choose? Or are they still the intellectual property of their owners despite their public and social nature?

In our view, a compelling case can be put forward that a Web site should be considered as the proprietary and private property of its author and owner. We can cite at least two familiar philosophical arguments to support this claim. The first is Locke's "labor-desert" theory. Recall the essentials of that theory: people have a natural right or entitlement to the fruits of their labor. Thus, if someone takes common, unusable land and through the sweat of the brow transforms it into valuable farm land that person deserves to own this land. Locke's basic argument is that labor is an unpleasant and onerous activity and hence people do it only to reap its benefits; as a result, it would be unjust not to let people have these benefits they take such pains to procure. In short, property rights are required as a return for the laborers' painful and strenuous work.

From a Lockean perspective, therefore, there ought to be property rights in Web sites because their value is based entirely on someone else's labor to construct and set up the site. The production of a Web site is often a labor-intensive activity and this effort should confer a property right for those who made the investment of time and effort to build that site.

Likewise, the utilitarian argument that ownership rights are justified because they maximize social utility and provide an incentive to build future web sites is also apposite. Jeremy Bentham justified the institution of private property by several arguments but the most germane is that knowledge of future ownership is an incentive to work hard to create or improve one’s property. According to Hardy (1996, par. 35), “this rationale directly applies to the creation and improvement of intangible property. . .[and] underlies the notion of copyright and patent laws. . .which exist to provide the incentive for the creation of works of authorship and technology.” It surely seems reasonable to conclude that the prospect of future ownership and all that it entails is an incentive for the creation and embellishment of Web sites, which require a significant investment of time and labor. To some extent, then, a recognition of private property rights in a Web site does provide an incentive to develop new sites, since developers will be motivated by the realization that they will retain firm control over the accessibility to these sites. Part of exercising that control is ensuring that visitors are exposed to the homepage so that advertising revenues will not be compromised.

Right now almost all Web site owners encourage visitors and linnkds from other Web sites, and this is as it should be. The Internet should be a public place, an open forum for the exchange of ideas and information. That does not imply, however, that Web sites are actually common or public property, fair game for manipulation by other Web sites, users, or even hackers.

If a Web site is a form of private intellectual property, there are a number of relevant points to be drawn from this conclusion. It must be understood that a property right is a right of ownership, the right to use or the right to possess, "that is to exclusive physical control of the thing owned; where the thing cannot be possessed physically, due, for example, to its 'non-corporeal' nature, 'possession' may be understood metaphorically or simply as the right to exclude others from the use or benefit of the thing" (Honore, 1961, p. 107). Indeed, the right to exclude is the most fundamental property right of an owner. It follows therefore that Web site owners should have the right to exclude others from use and the right to control how their “property” is utilized by others.

Thus a relevant moral principle here seems to be autonomy or self-determination, that is, the right to control the disposition of what one owns . If the Web site author’s autonomy is duly respected, other Web site publishers will refrain from imposing their activities such as deep linking on unwilling parties.

Our fundamental argument then is that Web site authors as property holders should have full control and autonomy over their sites. This implies that they should be able to determine how source Web sites do their linking. Further, since links do establish a connection between two businesses or institutions, Web site authors should have the right to restrict links from certain sources when such a connection might be a source of embarrasment. If they prefer direct links to the home page to maximize advertising revenues or otherwise increase exposure to announcements there, that preference should be respected by the source Web page.

There are many good reasons why a target Web page would have such a preference. As we have seen in both the Ticketmaster and Shetland cases, the loss of advertising revenues is certainly one valid reason. There may also be reasons based on a concern for aesthetics or the artisitic integrity of a particular Web site's presentation. Let's assume that an artist arranges a Web site with images of her paintings along with some descriptive text; she arranges them in a certain way to show the progression in her work. She wants the works viewed in conjunction with the text and in a certain order. Shouldn't she have the right to restrict individuals from linking to a work in the middle of this presentation or from linking to an image without its accompanying text? Respect for her autonomy and her right to control activities involving her Web site implies a moral imperative to refrain from deep linking in such situations.

This autonomy and control, however, must be balanced with the recognition that linking is a vital activity for the Web which furthers the common good of open communication, that is, the free exchange and free flow of ideas and information. Respect for this common good should set some reasonable limits on the property right enjoyed by Web site authors. One way of achieving this balance is to assume an implied license to link to any target site’s home page. The target site should have the prerogative to demand a more explicit license only under unusual circumstances. Also, in rare cases the target site should have the right to refuse the source Web site’s link because it does not want any association wtih that site for reasons that could tarnish its reputation. Further, we recognize that in some cases there will be private, secure Web sites that cannot be linked to because they contain sensitive information.

In our estimation, however, it would be too burdensome and counterproductive to insist upon an explicit license or explicit permission for every HREF link. The necessity of seeking such a license for every link would diminish the linking that occurs on the Web and thereby impair the common good of the free flow of information. On the other hand, all IMG links and deep linking should only be done with permission of the target site. IMG links are usually seamless, that is, they can be effected without any awareness that an image is coming from another site; it seems only fair that the target site grant its permission and stipulate any acknowledgements it wants for the display of its image at another Web site. Also, as we have seen, deep linking is problematic for many Web site authors since it bypasses promotional announcements on the home page and hence should only be done with consent of the target Web site. Finally, copyrighted material should not be utilized to identify the hypertext link without permission and proper attribution. These prima facie rules represent a prudent way of harmonizing one’s property right in a Web site with the Internet’s common good.