Digital media and the law. (Legally Speaking)

Journal:   Communications of the ACM  Oct 1991 v34 n10 p23(6)
           * Full Text COPYRIGHT Association for Computing Machinery 1991.
-----------------------------------------------------------------------------
Title:     Digital media and the law. (Legally Speaking)
Author:    Samuelson, Pamela.

Summary:   Six distinctive characteristics of the digital medium may cause
           disruption and adjustment in existing intellectual property system
           doctrines.  The ease with which digital data can be replicated
           makes the copying of works difficult to control, but digital media
           is particularly vulnerable because the technology that makes the
           medium work is the same one used in making multiple copies.
           Digital form is also easily transmitted and used by multiple
           users, which opens it to more widespread abuse.  The plasticity of
           digital media makes the work easy to amend and complements the
           owners' right to control the making of derivative works, but US
           law does not protect authors from misrepresentations or liability
           from physical harms coming from changes in the text.  The
           distinctions between different types of published work may become
           weakened under the law.  The compactness of works in digital form
           and the form's nonlinear characteristic are also discussed.
-----------------------------------------------------------------------------
Descriptors..
Topic:     Digital audio tape
           Digital recording
           Digital video
           Legal Issues
           System Design
           Copyright
           Compact discs.

Record#:   11 470 216.
-----------------------------------------------------------------------------
*Note*    Only Text is presented here; see printed issues for graphics.

Full Text:

Over the past two years, I have written seven "Legally Speaking" columns and one feature article for Communications about legal issues affecting computing professionals. These writings have covered an array of legal topics including: criminal and civil liability for hackers who breach computer security systems; first amendment issues arising in computing or electronic publishing markets; intellectual property issues, such as patent protection for computer program algorithms; copyright protection for look and feel of user interfaces; what the user interface design field thinks about such protection, and various theories by which a firm might claim to own interface specification information for software systems.

This column will suggest that certain distinctive qualities of the digital medium may be the reason why so many challenging legal issues have arisen over the past few years, and why so many more can be expected in the future. Six characteristics of digital medial seem likely to bring about significant changes in the law. Although the column will focus on intellectual property issues--because much of the contents of the digital medium are intellectual property products--it will suggest that the legal implications of having so many works in digital form are much broader. I invite my readers to augment these observations with their own.

Ease of Replication

The ease with which works in digital form can be replicated poses a great many challenges for the law, especially for copyright law. Copyright was clearly a more useful legal right in a world in which the only methods by which written works could be copied were by a printing press or by hand. Printing presses were expensive and bulky machines, requiring considerable skill and expense to operate. Handcopying was too time-consuming an activity to pose a commercial threat to publishers. Although "pirate" presses were an economic threat, copyright law provided a generally adequate legal weapon with which to fight such piracy in the world dominated by the print medium because it was generally possible to locate printing presses.

The technology of reprography has improved dramatically in the 20th century. As this technology has improved, it has become more difficult for copyright owners to exercise control over replication of their works and to obtain compensation for unauthorized replication. Newer reprography equipment has often been cheaper, less bulky, and more widely available than printing presses, and often requires less skill to operate. Copying by means of these new technologies has, consequently, become more difficult to trace, while at the same time becoming of greater economic concern to copyright owners.

Adding to the copyright owner's problem in controlling the copying of his or her works is the apparently widespread public perception that making copies for personal or private use ins lawful [2]. This view was seemingly made respectable by the Supreme Court's decision in the Sony Betamax case. This case did not, however, rule that all private or personal use copying is legal. It ruled only that Sony was not guilty of contributory copyright infringement for selling videotape recording machines, even though Sony knew some would use the machines to infringe copyrights. But because the primary use of the machines was to allow consumers to copy television programs for time-shifting purposes, the Court thought this should be "fair" and noninfringing.

While all of the improved reprography technologies pose similar threats to copyright owners, what makes works in digital form so much more threatening is that the same technology one needs to use the digital work is often the technology that can be used to make multiple copies of the work. Even more frightening, the technology can be used to produce "perfect" copies. Selling computer programs (or for that matter, other works in digital form) has become comparable to selling a customer the Ford automotive plant at the same time as selling him or her a Ford automobile. Each copy of the program has the potential to become its own factory. It is no wonder the recording industry worked so hard to keep digital audio tape equipment off the market. The controversy seems to have ended in a compromise--royalties will be collected on all sales of DAT machines and tapes and distributed to copyright owners--which may have set an important precedent.

Although the copyright system in the print world has generally forcused on sales of copies of copyrighted works, in the digital world the trend may be to reap the financial rewards for creating and disseminating intellectual products by charging for access to and use of digital works, and limiting rights to use and copy these intelletual products. Copyright law has traditionally not provided copyright owners with exclusive rights to control uses made of copies of their works. Copyright law has generally allowed owners of copies of copyrighted works to lend or even rent out their personal copies of works. The first exception to this rule was legislated a few years ago when "record rental" services were undermining record sales. Las year, a second exception was added to the copyright statute which prohibits the rental of computer programs without the copyright owner's permission.

Ease of Transmission and

Multiple Use

A second characteristics of digital media that poses challenges for traditional intellectual property regimes is the ease with which works in digital form can be transmitted and used by multiple users. Distribution channels for getting printed books from the printer to outlets at which they can be sold to the public have been relatively easy for copyright owners to oversee. Pirate presses and distributors would stand a strong chance of getting caught, for there would inevitably be a public quality to their efforts. Any pirate copy, if it reached the hands of a consumer (let's assume one innocent of the fact that the copyright had been infringed), could still only be used by one person at a time. Even if given to a third party, a pirate copy would leave the possession of one consumer, and only be transferred to the next one. Consequently, although copyright owners were never happy about the existence of pirate copies in the hands of consumers, they were generally content to concentrate their copyright enforcement efforts on shutting down the pirate presses and stopping distributors from selling illegal copies.

Concerns about what consumers will do with pirate copies are more serious when a works are in digital form. When a single pirate copy can be not only put in an isolated personal computer at a user's home, but loaded into a computer hooked up to a network of computers or a network of users of a large computer system, each of whom can have ready and virtually simultaneous use of the same copy, copyright owners are understandably more concerned about controlling pirate copies. Early efforts to excercise such control in the computer software market, through shrink wrap licenses and copy-protect systems, have not been very successful. Market forces led to the abandonment of copy-protected software because users thought they interfered with legitimate uses of the software. Shrink wrap licenses (which are of questionable legality, anyway) are widely ignored by consumers who continue to use and share software as if they had acquired the software without restriction. Other efforts to control transmission of works in digital form, such as export control regulations, seem doomed to failure in a world in which one can make digital transmissions of protected works to sources abroad through a number of means.

Ease of replication and ease of transmission and multiple use present copyright enforcement challenges of the first order. Together, there will be powerful incentives for the owners of rights in such works to attempt to restrict access and use, and to derive revenue more from uses than from sales of copies.

Those who operate computer systems have already developed elaborate systems for restricting access, not only to the computer system itself, but within the system, regulating, for example, which groups of users have authorization to have access to which parts of the system. Eventually, questions are likely to arise about whether the government should regulate (as it has in the communications market) who has what rights to control what kinds of access to information resources in digital form.

Plasticity of Digital Media

A wide array of social and cultural changes can be traced to the textual fixity made possible by the printed medium [1]. Before the advent of printing, written works were relatively few in number, hard to get, and tedious to copy. Because copyists of unprinted texts often amended them in various ways (adding new commentary or deleting what was not of interest), it was sometimes difficult to know what was the official version of a text. With printed works, all of the extant copies of a work (at least for a particular edition of it) were the same.

As beneficial as were many of the consequences of the fixed nature of printed works, not all the consequences of fixation were benign. In a sense, printed works are too fixed, at least for some users and some uses. They cannot, for example, easily be modified. If a text needed to be updated or corrected, publishers sometimes issue a new edition. But in time, the second edition will suffer from the same overfixity problem as had the first edition.

How pleasing, then, is the prospect of having works in digital form. We can correct errors, add commentary, alter the wording, delete what we do not want, update information, and so on, thereby overcoming one of the major disadvantages of the print medium. Such changes can occur quickly and without anyone being able to tell that the changes have been made. Digital sampling techniques allow one to "chop" a sound recording into sound bites that can be remixed and combined with sound bites from other recordings to produce a new recording which is not recognizable as derived from the original recordings. Photographs can be digitally manipulated to add what was not there, or to remove what was, or to combine photographic elements from many different works (which seem likely to make it much more difficult to use photographs as legal evidence). Computer programs can, by being processed through sophisticated reengineering software or compilers, be transmuted into unrecognizable forms. And these are only a few examples of what is characteristic of all works in digital form: they are all inherently plastic.

The plasticity of digital media is not an unmixed blessing. Copyright owners now have more reason to be concerned about what an individual user might do with his or her copy of the work. What if the user now customizes it and resells it to someone else? What if the user changes it in a way that misrepresents what the author meant to say? With a computer program, suppose the user modifies it to correct one error, but in the process of making a correction, inadvertently changes the software in a way that endangers life or property (as might happen with avionics software), thereby affecting the developer's potential warranty liability.

U.S. copyright law gives copyright owners the right to control the making of derivative works. The term "derivative work" is defined quite broadly ("any work . . . based upon a preexisting work . . ."). Thus, the copyright statute would seem to provide some authority for exercising control over what users can do to transform the copies they might have of a copyrighted work in digital form. And yet, much of what a copyright owner might want to control through the reach of the derivative work right is really outside the realm of interests that the U.S. Congress has traditionally meant for copyright to protect.

U.S. copyright law, for example, has traditionally not protected authors from misrepresentations or liability for physical harms resulting from changes in the text of a copyrighted work. Other countries protect the "moral" rights of authors. This might provide a remedy for misrepresentations about who is the author of what text, but not even they would reach the physician harm problem. (The first time the U.S. seriously considered recognizing "moral rights" was in response to the controversy over the "colorization" of movies, another phenomenon computing has made possible.) Generally, one who buys a copy of a copyrighted work reformats it in some way, and is free to remarket it. Under the "first sale rule" of copyright law, this rule has traditionally given owners of copies of protected works the right to exercise personal property rights in their copy of the work so long as they have tithed to the copyright owner upon the first sale of the copy to the public by the price they paid for the copy. In general, the copyright caselaw thus far also seems to give a more restrictive interpretation to the derivative work right than the statutory language might suggest--seeming, by and large, to require that a second work must have copied protected expression from the first one to be an infringing derivative. If a recognizable copy of a protected work is required to infringe the copyright, then digital sampling may not be infringement.

While there is nothing to forbid copyright law taking on other functions besides protection against unauthorized copying of protected expression, such a change will alter copyright law considerably. Congress and the courts may be wary of having copyright law interpreted so strictly that consumers of copyrighted works are forced to forego the significant benefits of the plasticity of the digital medium. Some balance needs to be found between the interests of copyright owners in controlling modifications to their works and the interests of consumers (and perhaps even competitors) in being able to take advantage of the plasticity of digital media.

Equivalence of Works in

Digital Form

A fourth important characteristic of digital media is what I will call the "equivalence of works" in that medium. Copyright law has traditionally treated different kinds of copyrighted works differently. Some kinds of works, for example, do not have the same sets of exclusive rights as do other works. Some special privileges to copy or make certain uses of copyrighted works are available to certain classes of copyrighted works, but not to others.

While there are some historical and public policy reasons for making such distinctions among different classes of works, there are additional important reasons for these distinctions: They reflect differences in the media by which different works have traditionally been made available to the public, as well as the technologies by which the different media are created and the distribution channels by which different media are disseminated to their respective publics. Books, the quintessential work of the print media, are made by printing presses, bound, and sold largely in bookstores. Paintings, sculpture, and photography are quite different media from books, and tend to be produced and distributed in quite a different manner as well. Phonograph records and compact discs are mechanically impressed with encoded information which, when played on a machine, can bring musical performances into one's home. While motion pictures also require a machine to be revealed, they differ significantly as a medium from sound recordings. Yet, they are similar in that they are often broadcast by radio waves, yet another medium whose differences from the printed medium were chronicled by Marshall McLuhan, among others. The point here is that copyright has traditionally conceptualized each entity as being only what it is, and not another thing.

This is not to say that copyright has no experience dealing with mixtures of different kinds of works or with multimedia works. To take a common example, although music is one category of work, a poem another, and a sound recording yet another, the law has come to an accommodation in dealing with what happens if the three are combined. The music and the poem (i.e., the lyrics), if written to go together, are classified together as a musical work. The sound recording, however, is treated as a separate work. Although some works can be made available in different media (music can be written on paper, sung on stage, or recorded on tape), the law still treats each medium somewhat differently.

Any work that can be represented in other media can now be represented in digital form. In this form it can be used in a computer or other data processing unit, whether to be displayed or heard, or to perform some other function. Once in digital form, works protected by copyright are going to become less and less differentiated by type and more and more equivalent to one another because they will now all be in the same medium. This equivalence of works in digital form will make it increasingly easy to create a difficult-to-classify work by combining what have previously been thought of as separate categories of works. (What is an interactive annual report for a company? A literary work? A computer program? An audiovisual work?) Consequently, the elaborate distinctions copyright law has made among different kinds of works will lose much of the meaningfulness they had when media were more differentiated.

Of even greater importance are the consequences of the blurring of what was once a firm distinction between data and writings on one hand and machines on the other. Computer programs have made the distinction between writings and machines, and between data and machines, more difficult to draw. Because of the mechanized character of computer programs, patents are now issued for methods of representing, organizing, and manipulating data in computer programs. One of the most profound consequences of this development is the new entry of patents into an arena in which copyright was once the sole form of intellectual property protection [3]. If a patent issues on a method of organizing data in a computerized database, does a copyright in the database also cover this method as a structural abstraction of the database program, or does the copyright provision excluding methods from the scope of copyright protection apply to exclude this aspect of the program from the reach of copyright? However this question is resolved, the resolution will chnage the face of intellectual property law, for the overlap of copyright law and utility patent law is a new phenomenon.

Compactness of Works in

Digital Form

Compactness is a fifth characteristic of digital media with potential to create new kinds of legal problems. By comparison with books and other traditional media, works in digital form do not take up much space. Consequently, works in digital form are inherently easier to steal. Last year IBM announced it had successfully stored a gigabit of data on a one-inch square magnetic disk. Had this same billion bits of data been typed in double-spaced format, the stack of paper would be three stories tall. It is a lot easier to walk off with a one-inch square than three stories full of documents. This is another reason why distributors of intellectual property will have more interest in controlling access to and uses of protected works in digital form.

The compactness of digital data will also allow new assemblages of materials that would be unthinkable in a print world. Not only whole encyclopedias (which in the print world are generally too bulky to be a single transportable work), but also whole specialized libraries (which have not traditionally been conceived of as themselves being separately protectable compilations of material) are already available on a compact disc. It is becoming commonplace for firms to store all their records on one or more central computers. While we are used to thinking of a firm's records as containing confidential information and perhaps also trade secrets, we are not used to thinking of the firm's records as a copyrightable compilation of materials. In the world of print, the same records would not be thought of as a protectable compilation, nor would a library. But while the compactness of digital media makes it possible to put company records, whole libraries, and the like in one compact form, some new kinds of intellectual property law problems seem likely to result from these new assemblages.

Another consequence of compactness is that it is increasingly difficult to get anything more than a small glimpse at the contents of works in digital form, for the whole is virtually invisible to the user by reason of its storage in electronic form. While it is difficult to look at more than one page of a book at a time, still the very heft of the tome makes printed works have a visibility that digital works do not have. With printed works, there are no significant constraints on one's ability to move from one page to another deep within the text, and to browse through the book at whatever pace and in whatever manner one chooses. With works in digital form, users of the work will tend to become much more dependent on user interface systems which will serve as a porthole through which to see the contents of the compact, but nonetheless oceanic, text of the digital work. Unless one can design one's own specialized portholes for viewing the encoded contents of the work, one's ability to access its contents will be constrained by how well or poorly the generalized user interfaces are for assessing these systems.

Thus, while compactness is a virtue for users because digital works take up less space than paper, the trade-off for users may be in their greater dependence on user interface systems which will not always have been designed with that particular user's needs in mind. Those who already use computing systems to do everyday tasks will have enough experience with how frustrating it can be to have to deal with an unfriendly user interface to understand how critically important user interfaces will be as access systems in a world dependent on digital libraries and other extensive collections of digital data.

Nonlinearity

A sixth characteristic of digital media is the potential it creates for nonlinear experiences with texts, for developing new methods of searching them and linking them together. In the print world, searches through texts tend to be quite linear in character. Although there are some conventions (such as footnotes) by which texts can be linked together, links tend to be linear experiences as well. Some book conventions give some search flexibility: books have tables of contents, indexes, bibliographies, chapters, page numbers, section and subsections, and other such elements. Such items as headings, subheadings, and bold face type help us make our way through books. We underline parts we think we might want to find again later. We turn down page corners or insert bookmarks when we want to hold our place in a printed text.

When one converts printed text to electronic form, one loses the ability to use some of the conventional search techniques of the printed form. Suddenly page numbers are no longer very useful, unless they are artificially added back in to help the reader by signaling where the same part of the text is to be found in the printed form of the work. It will not help someone, looking through an electronic version of the work for a particular passage once seen in the printed version, to recall that in the book the passage was in the upper righthand corner on the left side of the page about halfway through. One cannot browse through a book in digital form the way one can browse through a printed book. And yet, there are things one can do with digital versions of text that are, as a practical matter, very difficult or impossible in printed form (such as all instances of the word "glurp" in a long text).

Depending on how well one designs the conversion of an existing printed text to hypertext, much more than simple keyword searches can be created. Texts can be linked to other parts of the text, so that one can with the "click" of a mouse bring on the screen a related entry or even text from a related but separate document. One can create a kind of bookmark trail through the digital text by examining different aspects of it, one by one, and can navigate back and forth at the bookmark stops. Depending on the design, one may even be able to save one's bookmark trail, so that one can come back and retraverse the same trail at a later time, and perhaps extend it.

Multiple user hypertext systems can be designed in which one may not only be able to blaze one's own trail through the text, but may also be able to follow trails blazed by others. A multiple text hypertext system, permitting users to move from one document to another, provides a tremendous advantage that is unavailable in printed works. One group involved in digital library design envisions the creation of "knowbots" to navigate digital spaces to gather what individual users want, maybe even providing users with a synopsis of what it finds.

There are a host of new intellectual property law questions raised by the new capacity for searches and linking of works in digital form. Some of them are copyright questions; some are patent questions. Because works in digital form are processed by machine, it may be possible to patent search and linking techniques that, applied to works in printed form, would unquestionably not be patentable. Patent law has long had a rule that printed matter, although technically a manufactured product, is not the sort of thing that parent law (a law chiefly aimed at protecting technological innovations) was intended to protect. Under this rule, methods for organizing indexes in printed works would not be protectable. Yet the very same method, when put in digital (and therefore into machine-executable) form may suddenly be patentable, along with such things as methods of footnoting, methods of linking parts of text together, methods of representing symbolically how one can search the text, and methods of displaying information on a computer screen.

Among the many copyright questions raised by hypertext forms of works is whether creating a search trail through the digital text is itself a protectable work of authorship--particularly when one does it within the confines of the copy of the hypertext system copyrighted by somebody else. Another question is whether it is an infringing derivative work to create a program that links a variety of texts and parts of texts together to allow the user of the program to jump from one related part to another but without there being a new, and potentially infringing, copy made of any of these texts. While in the print world a bibliographer might compile a list others might use to search for materials on a particular topic, and this compilation might be copyrighted when printed in some journal, it has not been thought to be a copyright infringement for a reader of the bibliography to traverse the search path set forth by the bibliographer. Especially for works in digital form, creating links within texts and among texts may be a kind of intellectual work for which some recompense is appropriate, and copyright law may need to consider taking such links into the fold of protectable expression. To the extent hypertext allows us to extract the right pieces of information in a timely and economical fashion from a tremendous volume of data, some value is created. The question is who owns it--the person who initiated the search request or the person who made it possible to make the search by designing a good hypertext system?

Conclusion

Any one of the six characteristics of digital media mentioned in this essay would be enough to cause some disruption and adjustment in the doctrines of the existing intellectual property systems. But the six of them in combination to me seem likely to change the face of intellectual property law as we know it. Probably the old legal forms, copyright and patent, will continue to exist and be called by their old names. There will undoubtedly be some significant family resemblances between the old legal forms and the new ones we will create to deal with digital media. But the law of the intellectual property will look different after coming to terms with digital media.

References

[1] Eisenstein, E. The Printing Press as an Agent of Change, I and II, 1982.

[2] Office of Technology Assessment. Intellectual Property Rights in an Age of Electronics and Information (1986).

[3] Samuelson, P. Benson revisited: The case against patent protection for algorithms and other computer program-related inventions. Emory L. J. 39 (1990), 1025.

Pamela Samuelson is a professor of law at the University of Pittsburgh School of Law.