Linking Copyright to Homepages

Matt Jackson

Federal Communications Law Journal, 49 Fed. Com. L.J. 731, April, 1997
[citations omitted]



I. Surf's Up! A Technology Primer

II. Does Linking Infringe on Copyright?

        A. The Reproduction Right
                1. Direct Infringement
                2. Contributory Infringement

        B. The Adaptation Right
                1. Direct Infringement
                2. Contributory Infringement

        C. The Distribution Right
                1. Direct Infringement
                2. Contributory Infringement

        D. The Public Performance and Public Display Rights
                1. Direct Infringement
                2. Contributory Infringement

        E. A Word About New Web Page Design Features

III. Compiling Links for Fun and Profit

        A. Creating Copyrightable Links
        B. How Thin Is Thin?



Mary just started her own business. She created a Web page where users can follow her links to the
coolest sites on the World Wide Web (WWW or Web). It took her weeks of searching to find the best
sites and then a few more days to create all the links - which she plans on updating once a week. She set
up a password system so that anyone who wants to use her Web page has to pay her ten dollars a
month. In her first four months she made almost five thousand dollars. Now the author of one of the
pages she has linked to is suing her for copyright infringement. He wants her to either share her profits or
stop linking to his document. That's not her only headache. She's thinking of filing her own copyright
infringement suit against a former friend who is setting up a business similar to hers. He's created links to
many of the same documents.

How will these cases turn out? Can Mary create links to a Web site without getting the author's
permission? Can she stop her competitor from copying her links? Many law reviews, trade journals, and
daily newspapers have published articles concerning copyright in the digital age, but only a handful have
discussed one of the most common potential forms of copyright infringement - that of "linking" different
documents on the Internet.

The justification for copyright law in the United States is to provide an incentive for the creation of new
works. Article I, section 8 of the Constitution provides: "The Congress shall have Power ... To promote
the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the
exclusive Right to their respective Writings and Discoveries." Any U.S. copyright statute which does not
promote the progress of science and useful arts is presumably unconstitutional.

The economic rationale for copyright is based on the following assumptions: (1) granting property rights
in a work will allow the author to earn a profit from her labor, (2) the ability to earn a profit will provide
the author with the necessary incentive to create, and (3) the more works that are created, the greater
the benefit to the public and the greater the advancement of science. The paradox is that the public can
only benefit if it has access to the work. Access is restricted by granting the author property rights in her
work, for only by restricting access can the author charge users and earn a profit. Copyright law is
designed to resolve this tension in as equitable a manner as possible. Hence, copyright is of limited
duration, only certain categories of expression are protected, and facts are not copyrightable. These are
just a few of the limits placed on the author.

Currently, a heated debate is centered around the proposition that the Internet and the World Wide Web
have upset the sensitive balance between authors and users. Copyright owners, with the backing of the
Clinton administration, claim that unless copyright law is strengthened, content will not be made available
on the Internet and the network will fail. Internet users claim that if their current practices are restricted,
the Internet will fail to live up to its potential as a democratic, interactive medium of communication and
social interaction. This paper examines one of the most unique and important characteristics of the
Internet: the ability to create links between various documents.

One reason links have received such little copyright attention is that linking documents is still a relatively
new phenomenon and there is only one current court case in the United States involving links. However,
these issues will become more important as publishers seek to assert their property rights in information
available on the Internet. A recent issue of a business journal framed the question succinctly: "If I create a
home page and I have my copyrighted material on that page with my trademark, and someone
unilaterally links up to it, this raises the question of whether they're publishing and they're violating my

Linking documents is similar, but not directly analogous, to placing references to other works in a printed
text. For example, a new, printed article might refer to an already published article in Wired magazine.
The reader of the new article would have to find the correct issue of Wired magazine in order to see the
original article. The World Wide Web makes it possible for an electronic version of the new article to be
linked to the on-line version of the Wired article. When the reader reaches the point in the article where
the Wired article is referenced, the reader could select the link and immediately see the Wired article. It
has been widely noted that this ability to link documents is revolutionizing both information retrieval and
the act of reading itself.

This Article will examine two related copyright questions involving links: (1) does linking to a document
constitute copyright infringement, and (2) are links copyrightable? After a brief discussion of the
technology involved, this Article will argue that linking does not infringe on an author's copyright. It will
then discuss to what extent links are protected by copyright.

In our hypothetical scenario, Mary may be liable for copyright infringement in two distinct ways. If the
links she creates violate any of the author's exclusive rights, Mary may be liable for direct infringement. If,
on the other hand, Mary's customers violate any of the author's rights by following the link, then Mary
may be liable for contributory infringement. The basic conclusion of this article is that it is the author of
the document, not Mary, who reproduces the article for Mary's customers. Links are simply addresses
designating the location of a document. Therefore, Mary is not committing either direct or contributory
copyright infringement.

Because links are addresses, Mary is providing her customers with a database. The United States
Supreme Court limited the scope of copyright with regard to databases in Feist Publications, Inc. v.
Rural Telephone Service Co, so Mary's links enjoy very thin copyright protection. Mary may copyright
the selection and arrangement of her links, but not the links themselves.

I. Surf's Up! A Technology Primer

Few Americans over the age of five could have survived 1996 without hearing at least one reference to
the Internet or the Information Highway. Some researchers estimate that in the United States alone, as
many as fifteen million adults already have access to the Internet, and the number of users is expected to
grow exponentially in the next year alone.

The fundamentals of the Internet and the World Wide Web are fairly straightforward. The Internet is
both the hardware which connects thousands of computer networks worldwide, and the protocols which
allow these networks to communicate with each other. The Internet includes e-mail, discussion groups,
chat groups, and information resources.

Each individual network that is connected to the Internet usually consists of a host computer (the server)
and a number of remote computers or terminals (the clients). For example, most universities have
computer networks whereby hundreds of personal computers (clients) are connected to a large
mainframe computer (the server) via fiber optic cable. Users often can connect to the server from a
remote location using a modem and a telephone line.

The Internet is the interconnection of thousands of these servers, each with its own Internet Protocol (IP)
address. Every document has its own "address" on the server, similar to the way files are stored in a
personal computer. A user can access a document by specifying its address, which is known as its
Uniform Resource Locator (URL). A primary purpose of servers is to transmit documents to whomever
requests them.

There are a variety of protocols, such as ftp, telnet, and gopher, that allow a client to search for, and
request documents from a server. The World Wide Web is a newer set of protocols that utilizes
HyperText Transmission Protocol (HTTP) for communication between the server and the client. Client
programs, such as Netscape's Navigator and Microsoft's Explorer, request information from servers.
These programs are known as Web browsers.

One advantage of HTTP is that it can "read" older protocols such as ftp and gopher. Another advantage
is that HTTP lets the author use graphics, video, and audio in her documents. A third advantage (the
topic of this paper) is that the programming language of HTTP allows documents to be linked together -
even if they are stored on different servers. This language, known as HyperText Markup Language
(HTML), is how Web-sites (home pages) are typically created.

Documents which include HTML codes are known as Web documents. An author who creates a Web
document can create links by inserting a special code into the text or graphics. The code contains the
URL of whatever document the author wishes to link to her own document. When a user "clicks on"
(selects) the text or graphic, the browser requests whatever document is specified by the URL. The
server where the document is located then transmits the information to the Web browser.

There are three different types of links: intra-page, intra-system, and inter-system. Intra-page links
connect different parts of the same document. For example, a long document may have a link at the end
which takes the user back to the beginning. Intra-system links connect different documents on the same
server. An intra-system link on a university's server might connect the home pages of two different
departments. An inter-system link connects documents on different servers. Thus, a document
concerning intellectual property law on a university's server might be connected to the home page of the
United States Patent Office. Millions of documents can be linked together through the World Wide
Web. In addition, links can be created in two distinct ways. The most common form of link is a HREF
link. A HREF link is activated when it is selected, usually by clicking on it with the computer mouse. A
second way to create a link is with an IMG command. An IMG link is automatically activated when the
Web page is first loaded. Typically, this is used by the author to "call up" a graphic image stored in a
separate file. When the user looks at the Web page, the graphic is automatically loaded into the page.

Home pages do not have a standard form. They range from a single screen containing only text and no
links, to elaborate multiscreen documents with audio, video, and hundreds of links. Many individuals
have home pages where they include biographical data and links to some of their favorite Web sites. For
example, an individual may include her name, e-mail address, and a photograph of herself on her home
page. If her hobbies included kayaking, she might include a link to a home page created by a regional
kayaking club.

Businesses, universities, and other organizations often have home pages that include extensive links to
other documents maintained by the organization and related organizations. Thus, the local kayaking club's
home page might be linked to its membership list, a calendar of upcoming events, a description of the
club's history, and photos from a recent kayaking trip. It might also be linked to the home pages of other
kayaking clubs around the world.

The World Wide Web is only a few years old, yet its growth has been phenomenal. One survey
estimates that the Web grew from one million users in 1994 to eight million users in 1995. The recent
CommerceNET/Nielsen survey estimates that, as of January 1997, more than 37 million users could
access the World Wide Web in the United States and Canada alone. As of February 1997, there were
more than fifty million home pages on the Web.

Most universities allow their faculty, staff, and students to create Web sites on the university's server for
free. Many businesses have also established their own Web sites, either by purchasing their own server,
or by leasing space on an existing server. Anyone who wants to create his or her own home page can
rent space on servers from one of countless Internet service providers.

Before exploring the copyright issues involved in creating links, a few important technical aspects of links
need to be noted. First, Document A can be linked to Document B without the author of B's knowledge
or consent. However, A cannot link to a specific word or picture in B unless that word has its own URL
address. Thus, links generally go to the beginning of a document or to a link within the document that has
its own URL address.

Second, the link is a one way street - sort of. Someone browsing through A can follow A's link to B.
That user can backtrack from B to A because her Web browser "remembers" the path that was taken.
However, a user who starts at B has no way to connect to A, and furthermore, doesn't even know a link
exists from A to B. So if the author of document A wants to link to a specific section of B or have there
be a two-way link, she must contact the author of document B to arrange the link.

The third important technical note is that for a user to "view" a document, a "copy" of that document must
be loaded into the random access memory (RAM) of the user's computer. Otherwise, no image will
appear on the user's monitor. Whether this temporary copy in RAM should be considered a
reproduction under the Copyright Act is currently the subject of the heated debate.

Finally, the author of document B can use a variety of security measures to prevent anyone from viewing
(or linking to) her document. These measures include encryption of the document, or various levels of
passwords to prevent unauthorized access. In this way, the author of B can charge users each time they
access the document.

Two recent developments in Web page design have further complicated the legal analysis. The first is a
practice sometimes known as "mirroring." The author of A can create an IMG link to a part of B, such
that when a user first looks at A, that portion of B is displayed on A's page. For example, A might
contain a link to a particular graphic image on B. When a user looks at A, the graphic image will be
displayed on A's page, even though the image technically is stored on B.

The second development is a design technique known as "frames." A frame allows the author of A to
create a "window" within her page so that when a user follows a link to B, B appears within the window.
In this way, the border of A "frames" B, and the user always sees the outer portion of A's page.
Typically, A will create a wide border on the left side of the page with links to other portions of A's Web
site. The user could follow a link to B and beyond, and still return to any portion of A's page instantly,
without having to retrace all the links that she has followed. Though frames can be programmed to
display B in various ways, typically, B is not altered to fit the "window." Therefore, portions of B are
obscured by the frame. The user must use scroll bars to view those portions of B.

II. Does Linking Infringe on Copyright?

Copyright as a legal concept was partially a result of the development of the printing press in the fifteenth
century, and it has been adapting to new communication technologies ever since. The rapid expansion of
the Internet has led to many proposals for modifying the current law, including a recent proposal by the
Clinton administration's Information Infrastructure Task Force. The Task Force's proposal would codify
recent controversial court decisions regarding the Internet and computers that are discussed below.
Some scholars feel these modifications to copyright law will favor the copyright industries at the expense
of the general public. They argue that the current law is adequate to protect the copyright owner's

The 1976 Act grants the owner of a copyrighted work certain exclusive rights, which are themselves
subject to limitations contained elsewhere in the statute. The most important limitation is that copyright
protects only original expression, not facts or ideas. The copyright owner has the exclusive right to (1)
reproduce the work, (2) prepare derivative works, (3) distribute copies of the work, (4) perform the
work publicly (excepting pictorial, sculptural, or graphic works, sound recordings, and architectural
works), and (5) display the work publicly (excepting sound recordings and architectural works).
Because these rights may overlap, someone may infringe on more than one right at the same time.

To successfully sue for copyright infringement, the plaintiff must prove: "(1) ownership of a valid
copyright, and (2) copying of constituent elements of the work that are original." For the purposes of this
article, we will assume that Document B consists of copyrightable subject matter and that its author holds
a valid copyright in the work.

The author of A may be liable for infringement in one of three ways: (1) direct infringement, (2) vicarious
infringement, or (3) contributory infringement. Direct infringement occurs if the link itself violates one of
B's five exclusive rights. Vicarious or contributory infringement may result if, by selecting the link, the user
(of A's document) violates any of B's exclusive rights.

Vicarious infringement occurs when the third party (the author of A) has the ability to supervise or
control the direct infringer (the user), and the third party benefits from the infringement. Contributory
infringement occurs when the third party knows the infringement is taking place and "induces, causes, or
materially contributes to the infringing conduct ...." One court summed up the distinction between the two
by saying, "Just as benefit and control are the signposts of vicarious liability, so are knowledge and
participation the touchstones of contributory infringement."

While the author of A may benefit from a user selecting her link to B, she cannot supervise or control the
user. Therefore, a link from A to B does not involve vicarious infringement. However, by providing a
link, the author of A is inducing the user to view B. If viewing B violates any of the copyright owner's
exclusive rights, the author of A may be liable for contributory infringement. But there can be no
contributory infringement without direct infringement.

Contributory infringement is an important concern for the development of the Internet. Internet service
providers, who run the servers that make up the Internet, are justly concerned about contributory
liability. Courts have found bulletin board operators to be liable for infringing actions committed by their
users. This section will analyze the author's exclusive rights in terms of both direct and contributory

A. The Reproduction Right

1. Direct Infringement

The first enumerated right is the right to reproduce the work. The reproduction right is violated when a
copy is made of the original work. According to the 1976 Act, a copy is a material object "in which a
work is fixed by any method now known or later developed, and from which the work can be
perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or
device." The link from A to B only contains the URL address of B. In creating the link, the author of A
has not reproduced any part of B except for B's URL. A URL is a "fact," and as such, it is not protected
by copyright. One could argue that since the URL for B includes whatever name B's author gives to the
document, it contains protected expression. However, short phrases such as titles and names are
generally not copyrightable. Thus, A has not directly infringed B's reproduction right.

2. Contributory Infringement

When a user selects a link from A to B, the information contained in B is downloaded into the
random-access memory (RAM) of the user's computer. When the computer is turned off, all the
information in RAM is lost. Keep in mind that a copy must be fixed in a tangible medium. A series of
controversial cases have suggested that loading a computer program into RAM for viewing creates a
fixed copy and therefore may constitute copyright infringement.

The most important of these cases is MAI Systems, Corp. v. Peak Computer, Inc., in which the
defendant was a service company that repaired computers that were manufactured by MAI. When the
service technicians turned on the MAI computer, the operating software was automatically loaded from
the computer's hard drive to the same computer's RAM. The Court of Appeals for the Ninth Circuit held
that loading the software into RAM created a copy. The court cited the report of the National
Commission on New Technological Uses of Copyrighted Works (CONTU), which stated that, "the
placement of a work into a computer is the preparation of a copy ..." As the MAI court duly noted,
neither the prior cases which it cited for support, nor the CONTU report itself, distinguished between
placement in RAM or read-only memory (ROM).

The context of the CONTU statement was ensuring that the rightful possessor of a copyrighted computer
program would be able to use the program on her computer. In this sense, the Report seemed to be
contemplating the right of the user to load a copy of the program into the computer from a floppy
diskette. There is no indication that the authors of the report believed that once a program was in the
computer, its transfer from ROM to RAM would also be considered a copy.

The MAI court stated that, "Since we find that the copy created in RAM can be "perceived,
reproduced, or otherwise communicated,' we hold that the loading of software into the RAM creates a
copy under the Copyright Act." The court argued that since the computer may be left on indefinitely, the
copy in RAM is "fixed in a tangible medium" as required by the 1976 Act. This interpretation of the 1976
Act has been endorsed by MAI's progeny, and the Information Infrastructure Task Force. By this
reasoning, a slide projector which projects an image on a screen is making a copy. After all, the image
on the screen can be "perceived, reproduced or otherwise communicated" for as long as the slide
projector is left on. Loading a document into RAM for the purpose of displaying on a monitor is directly
analogous to projecting a slide onto a screen.

The MAI decision, and its endorsement by the Information Infrastructure Task Force have been roundly
criticized by leading copyright scholars. The MAI decision appears to be at odds with the legislative
history of the 1976 Act. The House Report accompanying the Act states, "The definition of fixation
would exclude from the concept purely evanescent or transient reproductions such as those projected
briefly on a screen, shown electronically on a television or other cathode ray tube, or captured
momentarily in the "memory' of a computer." The report went on to distinguish between a reproduction
and a display:

"Reproduction" under clause (1) of section 106 is to be distinguished from "display" under clause (5).
For a work to be "reproduced," its fixation in a tangible form must be "sufficiently permanent or stable to
permit it to be perceived, reproduced, or otherwise communicated for a period of more than transitory
duration." Thus, the showing of images on a screen or tube would not be a violation of clause (1) [the
reproduction right], although it might come within the scope of clause (5) [the public display right].

Thus, the legislative history suggests that Congress did not intend for a document temporarily stored in
RAM to be considered a reproduction. If viewing B does not create a copy, then there is no direct
infringement by the user.

Even if their interpretation of the law is wrong, one must accept that the courts have held that a document
in RAM is a copy. Does this mean the viewer has infringed B's reproduction right? To answer this
question one must determine who made the copy that resides in the user's RAM. The author of B placed
the document on a server. When a user who is viewing A clicks on (selects) the link to B, the user's Web
browser requests the document from B's server. It is B's server that actually generates the "copy" which
is sent to the user. Thus it is B, not A, that authorizes the reproduction.

A leading Supreme Court case involving contributory infringement offers insight as well. In Sony
Corporation of America v. Universal City Studios, Inc., the issue was whether Betamax videotape
recorders (VTRs) sold to consumers by Sony were being illegally used to record broadcast television
programs. Universal argued that Sony was knowingly supplying the means by which consumers were
committing copyright infringement, and therefore Sony should be liable for contributory infringement.
Universal relied heavily on Kalem Co. v. Harper Bros., in which the producer of an unauthorized film
dramatization of a copyrighted book was held liable for selling the film to distributors, thereby
contributing to the infringement of the author's public performance right.

In rejecting Universal's argument, the Sony Court distinguished Kalem, stating, "The producer in Kalem
did not merely provide the "means' to accomplish an infringing activity; the producer supplied the work
itself, albeit in a new medium of expression. Sony in the instant case does not supply Betamax consumers
with respondents' works; respondents do." As in Sony, it is B's author who is supplying the user with the
work. A is simply providing the user with an alternative method for viewing B (just as time-shifting in
Sony provided the viewer with an alternative method for viewing Universal's programs).

In Sony, the case turned on whether or not there were "substantial, non-infringing uses" for a Betamax
videocassette recorder. Clearly, there are substantial, noninfringing uses of linking technology in general.
The more important question is whether a link from A to B is capable of substantial, noninfringing uses.
This brings us back to the question of whether reading B (which entails loading B into RAM) is a
noninfringing use of the link.

Recently, the MAI decision was applied to Internet documents for the first time in Religious Technology
Center v. Netcom On-Line Communication Services, Inc. In Netcom, an Internet user posted Religious
Technology Center (RTC) documents on a USENET discussion group. Netcom operated one of the
servers which stored and distributed the discussion group. In a footnote, the court said that under MAI,
"Browsing technically causes an infringing copy of the digital information to be made in the screen
memory ...." However, later in the same footnote, the court said that, "[Browsing] is the functional
equivalent of reading, which does not implicate the copyright laws and may be done by anyone in a
library without the permission of the copyright owner. [Even if one rejects the reading analogy], absent a
commercial or profit-depriving use, digital browsing is probably a fair use." Since viewing a document
does not infringe the reproduction right, providing a link does not constitute contributory infringement.

B. The Adaptation Right

1. Direct Infringement

The copyright owner's second exclusive right is the right to prepare derivative works, the adaptation
right. According to the 1976 Act, a derivative work is "a work based upon one or more preexisting
works, such as a translation, musical arrangement, dramatization, fictionalization, motion picture version,
sound recording, art reproduction, abridgment, condensation, or any other form in which [the
preexisting] work may be recast, transformed, or adapted." The purpose of the adaptation right is to
allow the copyright owner to control more than simply verbatim forms of copying.

Generally, to violate the derivative right, the infringing work must copy part of the underlying work. As
discussed in the previous section regarding the reproduction right, a link from A to B does not
incorporate or copy any portion of B. Thus, a link does not create a derivative work.

2. Contributory Infringement

One commentator has suggested that linking documents may create a derivative work by creating a
"literary "add-on.'" An add-on modifies an existing work and is used in conjunction with that work. If A
contains links to specific sections of B, one could argue that A modifies the way a user views B. In
effect, A is creating an abridged version of B. With printed texts, A would need to copy the desired
sections of B to be an abridgment and hence a derivative work. But with links on the World Wide Web,
A can create an abridged version of B without copying. Thus, the notion is that A is an add-on (i.e., a
supplementary work). The "add-on" concept has appeared in recent court cases involving computer

In Midway Manufacturing Co. v. Artic International, Inc., the Court of Appeals for the Seventh Circuit
ruled that a computer chip manufactured by Artic to speed up a Galaxian video game manufactured by
Midway infringed on Midway's copyright. The court ruled that the speeded up version of the video game
constituted a derivative work. Artic argued that speeding up the video game was like speeding up a
phonograph record and so should not be considered a derivative work. The court rejected this argument
based on the fact that there is a market for speeded up video games while there is no market for
speeded up phonograph records.

Almost a decade later, a similar case was heard in the Ninth Circuit. In Lewis Galoob Toys, Inc. v.
Nintendo of America, Inc., Galoob manufactured a device, a "Game Genie", to be inserted between a
Nintendo home video game cartridge and the Nintendo home video game control unit. The device could
be programmed to change certain characteristics of Nintendo video games. The court ruled this was not
a derivative work and distinguished it from Midway by pointing out that the earlier case involved
substantial copying of a ROM chip while Galoob's device involved no direct copying. The court also
noted that the device manufactured by Artic was used in the commercial setting of a video arcade, while
Galoob's device was used in a noncommercial home setting.

The Galoob court stated in dicta that derivative works should not encompass works whose sole purpose
is to enhance the underlying work. Neither a spellchecking program used in conjunction with a word
processor, nor a kaleidoscope that allows one to view a work in a new way should be considered a
derivative work. The court said, "The Game Genie is useless by itself, it can only enhance, and cannot
duplicate or recast, a Nintendo game's output .... Such innovations rarely will constitute infringing
derivative works under the Copyright Act." The Galoob court went on to state that even if the Game
Genie were a derivative work, its use should be considered a fair use.

The Galoob court ruled that a computer add-on that does not incorporate any part of the underlying
work is not a derivative work. Under the same reasoning, a "literary add-on," such as a series of links,
should not be considered a derivative work either. The links from A to B cannot exist independently of
B. Unlike a printed abridgment or adaptation of a work, the links do not duplicate the original work or
act as a substitute for it.

C. The Distribution Right

1. Direct Infringement

The copyright owner has the exclusive right to "distribute copies or phonorecords of the copyrighted
work to the public by sale or 1111other transfer of ownership, or by rental, lease, or lending." The
distribution right allows the copyright owner to sue a distributor of unauthorized copies even if that
distributor did not make the copies himself. This has been an especially important right with regard to the
Internet, since the person who distributes a document on the World Wide Web does not necessarily
make a copy.

In Netcom, where a user placed an RTC document on the Netcom computer, the court rejected RTC's
argument that Netcom should be liable for direct infringement of RTC's distribution right. The court
reasoned that only the person who uploads the document to the server should be liable for direct
infringement. Similarly, in Sega Enterprises, Ltd. v. MAPHIA, a bulletin board operator was found liable
for contributory infringement rather than direct infringement for allowing users to upload and download
copyrighted Sega video games. These cases suggest it is the person who places the document on the
server who is liable for direct infringement of the distribution right. Since A merely provides a link to the
server where B is located, the author of A should not be liable for direct infringement.

A useful analogy is a telephone answering system. One can program a number into speed dial and then
call the number to reach a business's answering machine and listen to their outgoing message. B's server
is like an answering machine. When B's author places B on the server, it is akin to placing an outgoing
message on the answering machine. The URL that designates B's location is the "phone number" used to
reach the answering machine. When the author of A creates a link to B, she has essentially put B's phone
number (the URL) into a speed dial memory. When the user selects the link, the user's Web browser
"calls" B's server. B's answering machine (the server) then transmits the outgoing message (B) to the
user's Web browser for the user to view. The crucial point is that A does not control the distribution of
B. If B's author no longer wants to distribute B, she can take the document off the server or restrict
access with encryption or passwords. So even if a copy of B has been distributed, the distribution is
being made by the author of B, not A.

2. Contributory Infringment

As long as the copyright owner of B has placed it on the server, its distribution is authorized and A
cannot be held liable for contributory infringement. But suppose the author of A creates a link to a
document that has been placed on a server without the copyright owner's authorization. If the author of
A has knowledge of the direct infringement, she may be liable for contributory infringement, since her link
encourages the further distribution of the document.

In MAPHIA and in Playboy Enterprises, Inc. v. Frena, bulletin board operators were found liable for
allowing users to upload and download copyrighted materials on their systems. In both cases, the
defendants knew that the material was being uploaded without the copyright owners' permission.
Contributory infringement requires knowledge of the infringing activity. In Netcom, the defendant argued
that it cannot "know" of an infringement when it cannot determine whether a subscriber is making fair use
of copyrighted material. The court agreed, noting:

Where a BBS operator cannot reasonably verify a claim of infringement, either because of a possible fair
use defense, the lack of copyright notices on the copies, or the copyright holder's failure to provide the
necessary documentation to show that there is a likely infringement, the operator's lack of knowledge will
be found reasonable and there will be no liability for contributory infringement for allowing the continued
distribution of the works on its system.

This same standard should apply to individuals who create links as well as online service providers. If a
court followed the Netcom reasoning, the author of A would not be held liable unless the copyright
owner had contacted her with proof that B contained infringing material.

Of course, documents on the Web are constantly being updated. The author of A might link to B, and
later find that B has added unauthorized material. A court would have to decide if it is reasonable to hold
the author of A liable in this situation.

D. The Public Performance and Public Display Rights

1. Direct Infringement

The copyright owner has the exclusive right to display or perform her work publicly. According to the
1976 Act:

To perform or display a work "publicly" means- (1) to perform or display it at a place open to the public
...; or (2) to transmit or otherwise communicate a performance or display of the work ... to the public, by
means of any device or process, whether the members of the public capable of receiving the
performance or display receive it in the same place or in separate places and at the same time or at
different times.

A display or performance can occur without a copy of the work being made. Like the distribution right,
the performance and display rights are heavily implicated by the transmission of documents on the World
Wide Web. On the Web, a work can be displayed or performed or both. For example, a Web site may
include text and pictures which are displayed on a monitor, and moving images and audio which are
performed. The differences between a display and a performance are inconsequential for the purposes of
this discussion.

When viewing A, B is not being displayed or performed. Therefore, the author of A is not directly
infringing under clause (1) of the definition. However, the author of A may be liable under clause (2),
which includes the transmission of a work.

Courts have viewed public displays over the Internet much like a distribution. In Frena, the court held
that the public display right was implicated as well as the distribution right. The court stated that "The
display right precludes unauthorized transmission of the display from one place to another, for example,
by a computer system." When a user views B, a transmission is clearly taking place, but it is the author of
B who has displayed (or performed) the document by placing it on the server.

Listening to the transmission of an answering machine's outgoing message over a telephone line would
also be considered a public performance. If someone lets you use their phone and dials the answering
machine for you, they have not violated the performance right. By the same token, A is not directly
infringing on the display or performance rights.

2. Contributory Infringement

Under the current law, establishing links from A to B should not be considered copyright infringement. In
fact, if the author of A wanted to charge users for using her links to B, she could. This would be true
even if she did not share any of her profits with the author of B, and even if the user could access B
directly for free. But can the author of A copyright her links to protect her profits?

E. A Word About New Web Page Design Features

To this point, we have been discussing standard HREF links between A and B. As mentioned at the
beginning of this paper, in the last year some new design features such as "mirroring" and "frames" have
become popular on many Web pages. To what extent do these new features change the legal analysis for
copyright infringement?

"Mirroring," whereby portions of document B are automatically displayed on A's page by using an IMG
link, clearly violates B's display and/or performance rights. Some participants on discussion list argued
recently that since A does not actually contain a copy of the displayed portions, but rather simply follows
its own link to B, that no infringement is involved. While it is true that there is no infringement of the
reproduction right, the "mirror image" clearly violate the public display right by displaying the image "at a
place open to the public."

A more difficult analysis is required to determine if frames violate the copyright act. With frames, when a
user follows a link from A to B, B appears inside a window "framed" by A. As with mirrors, the
reproduction right is not implicated since no part of B is reproduced by A. The question becomes
whether or not a derivative work is created in violation of the adaptation right.

In Mirage Editions, Inc. v. Albuquerque A. R. T. Co., the defendant was found to have violated the
plaintiff's adaptation rights when he purchased the plaintiff's book of photographs, mounted the individual
photos on tiles, and resold them. One could argue that frames have the same effect by "remounting" B in
a frame created by A. As stated earlier, a derivative work is one in which the "[original] work may be
recast, transformed or adapted." Whether framing results in the creation of a derivative work depends on
the court's definition of "recast." The legislative history of the 1976 Act does not offer any insight into
what Congress intended by "recast." If one assumes that to recast means to alter in some fashion, then
framing does not appear to qualify. When placed within a frame, B is not altered in any manner. Rather,
a portion of B is simply hidden from view (depending on how the frame is configured). While more work
is required of the user to view B in its entirety, the complete document remains intact. Thus, frames
should not be considered an infringement of B's derivative rights.

As with contributory infringement of the distribution right, A may be liable if it is linked to an unauthorized
display or performance. But as long as B does not infringe on someone's display or performance right,
viewing B (and therefore, linking to B) does not constitute infringement.

III. Compiling Links for Fun and Profit

A. Creating Copyrightable Links

An author who creates links from A to B or other Web sites may wish to be compensated for her effort
in searching for appropriate documents and establishing the links. For example, there are millions of Web
sites on the Internet and a particular user interested in movies may find only a few of these sites to be of
interest. Searching through all the sites, or even using a search program, can be tedious. If there is
enough demand, the author of A may want to establish links to all the Web sites relevant to movies and
then charge users who want to use A as a starting point. Can the author of A prevent someone else for
setting up a competing Web site with its own links to the same movie Web sites?

Because links are facts, they are not copyrightable. However, a compilation of facts (i.e., a database)
can be copyrighted. "A "compilation' is a work formed by the collection and assembling of preexisting
materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a
whole constitutes an original work of authorship."

The copyright in a compilation extends only to the material contributed by the author, and does not
extend to any preexisting material. For example, an anthology of poems is a compilation. The author of
the anthology can copyright the arrangement and selection of the poems as well as any original
expression that the author adds. But the author cannot copyright the poems themselves. A database is a
compilation consisting of noncopyrightable facts. If A has links to B, C, D, et cetera, then A has
compiled a database consisting of the URLs for the documents to which it is linked.

The leading copyright case involving compilations of facts is Feist Publications, Inc. v. Rural Telephone
Service Co. In Feist, a local telephone company which published its own telephone directory sued a
publisher for copying some of its listings. The Supreme Court ruled that factual compilations must entail
some originality as to the selection or arrangement of the facts they contain. Indeed, the Court repeated
this test throughout its opinion: "If the selection and arrangement are original, these elements of the work
are eligible for copyright protection.... A factual compilation is eligible for copyright if it features an
original selection or arrangement of facts...." The Feist Court rejected lower court cases which had held
that factual compilations deserved protection because of the effort that went into collecting and compiling
the data.

Any expression which the author adds to the facts is, of course, copyrightable: "Thus, if the compilation
author clothes facts with an original collocation of words, he or she may be able to claim a copyright in
this written expression. Others may copy the underlying facts from the publication, but not the precise
words used to present them." So if A includes original descriptions of the links, those descriptions are
copyrightable. However, that protection would not extend to the links themselves.

The difficult question is what is the requisite level of originality required in the selection and arrangement
of the facts. Feist states that:

The compilation author typically chooses which facts to include, in what order to place them, and how to
arrange the collected data so that they may be used effectively by readers. These choices as to selection
and arrangement, so long as they are made independently by the compiler and entail a minimum degree
of creativity, are sufficiently original that Congress may protect such compilations through the copyright
laws. Thus, even a directory that contains absolutely no protectable written expression, only facts, meets
the constitutional minimum for copyright protection if it features an original selection or arrangement ...
Originality is not a stringent standard; it does not require that facts be presented in an innovative or
surprising way. It is equally true, however, that the selection and arrangement of facts cannot be so
mechanical or routine as to require no creativity whatsoever. The standard of originality is low, but it

Feist held that a typical telephone directory white pages, with its selection of basic subscriber information
arranged alphabetically, does not possess enough creativity to qualify for copyright protection. The Feist
ruling has been extended by lower courts to business directories as well.

However, in Key Publications, Inc. v. Chinatown Today Publishing Enterprises, Inc., the Second Circuit
Court of Appeals found that a telephone directory for businesses located in Chinatown, New York was
copyrightable. The court defined selection as, "the exercise of judgment in choosing which facts from a
given body of data to include in a compilation." Because the publisher chose which businesses to include
in its listings and created the categories the businesses would be listed under, the court found that the
directory was copyrightable. The Key case is important for our discussion because, like most Web
pages, it was not a comprehensive listing of all the phone numbers that could have been included in a

A should be copyrightable as a compilation unless it contains a link to every Web site relevant to a topic
and lists them in alphabetical order. Any expression A contains (including descriptions of the Web site
each link is connected to) is also protected by copyright. The more difficult question is whether someone
else can set up a similar series of links.

B. How Thin Is Thin?

As the Feist Court noted, "Copyright in a factual compilation is thin. Notwithstanding a valid copyright, a
subsequent compiler remains free to use the facts contained in another's publication to aid in preparing a
competing work, so long as the competing work does not feature the same selection and arrangement."
Thus, another author is free to use some of the same links as A.

In Key, the Court of Appeals found that a competing telephone directory did not infringe on Key's
copyright - even though they shared many of the same listings - because the competing directory
grouped its listings into different categories and not all of the listings were identical. In explaining its ruling,
the court wrote:

There are a finite number of businesses that are of special interest to a sizable segment of the New York
Chinese-American community, and some substantial overlap among classified business directories
compiled for that community is inevitable. The key issue is not whether there is overlap or copying but
whether the organizing principle guiding the selection of businesses for the two publications is in fact
substantially similar ....

Under the Key analysis, two documents could both be linked to many of the same Web sites as long as
the two documents do not share the same selection and arrangement.

While the court acknowledged that within a particular category some listings will overlap, the listings
cannot be identical: "If the Galore Directory had exactly duplicated a substantial designated portion of the
1989-90 Key Directory - for example, all its listings of professionals such as medical doctors, lawyers,
accountants, engineers and architects, an infringement action would succeed."

The Key court cautions, however, that simply adding or subtracting a single fact (or link) will not prevent
a finding of infringement. Similarly, if A contains links arranged as the "Top 100 Web sites," B cannot
avoid infringement by simply using A's selection to create the "Top 50 Web sites." This is exemplified by
the Key court's reflection on its earlier decision in Eckes v. Card Prices Update:

In that case, we held that a guide to baseball cards infringed a previously published guide, even though
the copyrighted guide listed over 18,000 cards and the infringing guide listed only 5000 cards. Essential
to our finding of infringement was the fact that the 5000 listings duplicated in the infringing guide were the
same 5000 designated as "premium" cards by the copyrighted guide.... The copyrighted guide selected
within the 18,000 a designated group of 5000 that it described as "premium" cards. The infringing guide
then copied that portion wholesale based upon the same principle of selection.

Finally, two documents may be able to share the same links because sometimes there are so few ways of
expressing an idea that the idea and its expression merge. To grant copyright to the expression would
eliminate the idea/expression distinction which is the foundation of copyright law.

In Skinder-Strauss Associates v. Massachusetts Continuing Legal Education, Inc., two publishers
printed competing legal directories for the state of Massachusetts. The district court used the merger
doctrine in its analysis, stating: "The merger doctrine applies here because there are so few ways of
compiling listings of attorneys. This is because, by definition, any directory of lawyers for a given locale
will include virtually the same information." The Skinder-Strauss court held that the alphabetical listing of
Massachusetts attorneys was not copyrightable, but that other elements of the individual directories and
their overall structure were copyrightable. This suggests that, depending on the subject matter, two
documents can share identical links but that the second document may infringe the copyright of the first if
it copies other elements as well. For example, if A and B both attempted to create links to all the Web
sites that contained information about movies, they might share many of the same links. But A might be
arranged by movie genres while B is arranged by director. A closer case would be if A and B both
created links to all the Academy Award-winning movies. In that case, the merger doctrine might apply.


Anyone who has used the World Wide Web knows that links between documents are ubiquitous.
Fortunately, copyright lawsuits involving links are not - yet. To understand the legal implications of links,
one must appreciate both the technical processes involved and the current interpretation of copyright

Links are like telephone numbers; when a user selects a link, she is calling a computer as if it were an
answering machine. When the author of Document B puts B on a server, it is like placing an outgoing
message on an answering machine. Anyone who calls can listen to the message. And just as it is the
owner's answering machine that transmits the message to the caller, it is B's server that transmits the
document to the user.

If Document A contains links to Document B, none of B's exclusive rights are being infringed, since A
simply contains B's "phone number." Even if one accepts the court cases that have held that a document
in RAM creates a copy, no rights are being violated. This is because the author of B has authorized the
distribution and/or display/performance of B by placing B on a server.

The author of A can charge the user for access to A's links to B - even if access to B is free. The
trade-off is that A only enjoys a thin copyright in her selection and arrangement of links. Furthermore, the
links themselves are not copyrightable.

All of this is to the public's benefit. That anyone can create or follow a link gives the public the widest
possible access to information. The thin copyright offered to A encourages the development of useful
links since the author of A can be compensated for her effort. At the same time, the limited nature of the
copyright prevents A from creating a monopoly in links and charging exorbitant prices.

So what about the hypothetical situation presented at the beginning of this article? Mary will not have to
pay the author of the document she has linked to, but he can reconfigure his document so as to require a
password. He would then be able to charge Mary every time someone selects that link. And unless
Mary's competitor is using her selection and arrangement to organize his own links, she will not be able
to prevent him from competing with her. Which means Mary will have to lower her prices or offer a
superior service. Either way, her customers win.

The legal analysis used in this article includes the assumption that B does not contain any infringing
material of its own. If B does not contain any infringing material, then linking to B does not constitute
infringement. However, many Web pages, both personal and professional, do contain infringing material.
A great deal of uncertainty remains as to whether linking to these pages constitutes contributory

I argue elsewhere that the framework for contributory infringement laid out in Netcom, if applied in the
light most favorable to online service providers, would serve national policy goals best. The same
framework should also be applied to authors who create links to infringing documents. The author of A
should not be liable for the infringing actions of B unless: (1) the author of A has knowledge of the
infringing action, and (2) there is absolutely no reasonable fair use defense.

As Congress debates altering the current copyright law, it would do well to note that the phenomenal
growth of the Internet is due in large part to the free flow of information through the World Wide Web.
Authors who place their documents on the Web know full well that others may link to the document and
download it for viewing. That is its whole purpose. The information industries now see the Web as a
potential marketplace to be exploited. Changing copyright law to suit these private industries would
significantly alter the development of the Internet as a public forum dedicated to the free exchange of