***** Computer Select (Standalone), Apr 1992 : Doc #7317 ***** Journal: Communications of the ACM Feb 1992 v35 n2 p27(6) * Full Text COPYRIGHT Association for Computing Machinery 1992. ----------------------------------------------------------------------------- Title: Copyright law and electronic compilations of data. (Legally Speaking; column) (Cover Story) Author: Samuelson, Pamela. Summary: The US Supreme Court ruled in early 1991 that the telephone white pages are not protected by copyright law because they do not contain any 'original expression'. Creativity in the selection or arrangement of the facts is necessary before a compilation is covered by copyright law, otherwise the facts can be copied by anyone, even somebody preparing a competing compilation. This decision has far-reaching implications for copyright claims in fact compilations, especially for electronic information services. The Feist Publications vs. Rural Telephone Service ruling and its implications are discussed in detail. ----------------------------------------------------------------------------- Descriptors.. Topic: Data Bases Rights Copyright Information services United States. Supreme Court. Record#: 11 936 387. ----------------------------------------------------------------------------- *Note* Only Text is presented here; see printed issues for graphics. Full Text: It has been almost a year since the U.S. Supreme Court issued an important ruling affecting those who compile information and make it available to the public, electronically or otherwise. In the case Feist Publications vs. Rural Telephone Service, the Cout ruled that the telephone white pages could not be protected by copyright law because they contained no "original expression." The Court rejected the "industrious collection" or "sweat-of-the-brow" theory of originality that lower courts had been using for decades to uphold copyright claims in telephone books and similar fact compilations. Only if there is some creativity in the selection or arrangement of the facts, can their compilation be protected by copyright law. The facts themselves, the Court ruled, could be copied at will, even by one who was preparing a competing compilation. This was said to be consistent with the constitutional purposes underlying copyright law. Much of the Court's analysis in the Feist opinion discusses copyright issues raised by fact compilations in general, not just the white pages. Therefore, the decision is likely to have broad implications for copyright claims in fact compilations. Those who maintain electronic fact compilations know that the real value of such compilations is in the data, not in the way it might be arranged. Rearranging digital information is cheap and easy. Some electronic information services will view Feist as undermining their economic incentives to prepare such compilations and make them widely available to the public. Other electronic information providers may see the decision as opeing up new opportunities for making new information products available to consumers. Partly because the Court gave some surprising reasons for its ruling, phe head of the U.S. Copyright Office has referred to the Supreme Court's Feist opinion as having "dropped a bomb" on U.S. copyright law. This column will attempt to explain why some might agree with this characterization. The Facts of Feist Rural Telephone Service is a certified public utility that provides telephone service to several communities in northwest Kansas. It publishes an annual updated telephone directory. The white pages component of the directory contains an alphabetical listing of the names of Rural's subscribers, the towns in which they live, and their telephone numbers. Rural distributes this directory without charging its subscribers, but profits from selling advertisements that appear in the yellow pages component of the directory. Feist Publications was formed by a high school history teacher (named Feist) from northwest Kansas who decided residents would benefit from one area-wide telephone directory containing listings from all 11 telephone service areas in the part of Kansas. Feist went to the 11 companies to obtain a license to use their white pages listings in his directory. All but Rural Telephone granted him the requested license. Faced with a choice between having an incomplete area-wide directory or copying Rural's listings, Feist chose the latter. Aware that telephone companies put in fictitious listings in order to catch those who copy information from their directories, Feist hired people to check the nearly 5,000 listings in the Rural directory that he wanted to put in his area-wide directory. Nevertheless, four fictitious listings from the Rural directory showed up in the Feist directory. More than 1,300 of the Feist directory's nearly 47,000 listings were identical to the listings in the Rural directory, and about 3,600 were the same except for Feist's addition of addresses to the entries. The Issues in the Lawsuit Rural Telephone Services sued Feist for copyright infringement, and won in the trial court. The trial judge relied principally on a number of prior court decisions holding that telephone directories were protectable by copyright law on a "sweat-of-the-brow" theory of originality in a compilation of white pages listings. These cases had also ruled it was an infringement of the copyright in such a directory for someone to copy listings from one directory to put in a competing directory. The appellate court affirmed for substantially the same reasons as the trial court. (In a related antitrust case, the trial judge found that Rural's refusal to license the white pages listings to Feist was motivated by its desire to maintain its monpoly over yellow pages advertisements.) It was somewhat surprising to those who concern themselves with copyright law when the Supreme Court decided to hear the appeal in Feist since there had been a number of prior cases involving copyright claims in directories whose appeals the Court had not taken. The issue on which the Court agreed to hear the Feist case concerned the scope of copyright protection available to telephone directory white pages. (Note that this issue would seem to presume the validity of the copyright, raising the question only as to the extent of protection such a copyright provided.) But the issue on which the Court ruled was the validity of Rural's claim of copyright in white pages listings. Stated in copyright terms, the question was whether there was sufficient originality in white pages listings to make them copyrightable. To this question, the Supreme Court's answer was a very firm "no." The Pros and Cons of "Sweat-Of-the-Brow" Copyrights One argument that might be made in favor of "sweat-of-the-brow" copyrights is that humans generally feel a sense of ownership in the product of their labors, whether the products be highly artistic, highly inventive, or very mundane. The notion that property rights arise from the investment of labor in the creation of something of value has deep roots in Anglo-American jurisprudence. U.S. intellectual property law, however, has always had special rules about what kinds of intellectual products are eligible for protection and under what circumstances. Not everything in which labor is invested can be protected under these rules. Copyright law, for instance, requires the presence of some "originality" of "expression" in a written work in order for copyright protection to be available to its author. The term "originality" has never been defined in the U.S. copyright statute. Congress seems to have been content to let the courts develop its meaning on a case-by-case basis. Over the years, U.S. case law has defined originality in somewhat varying ways. One commonly quoted definition of the term derives from a nineteenth century Supreme Court decision. It speaks of originality as meaning that the work merely "owes its origin" to the person claiming to be its author, and was not copies from another work. Rural's white pages (and many other fact compilations) seem to satisfy this standard. Thus, the "sweat-of-the-brow" standard for originality seems to have some basis in the case law, to which Congress arguably deferred by not defining originality in the statute. (There is general agreement that the originality standard of copyright law under U.S. law is very low, although some disagreement on just how low. Some other countries have higher standards than U.S. law for how much originality is required in order for a work to qualify for copyright protection, and for the person responsible for bringing the work into being to be considered an author. The term "author" is sometimes reserved for those who speak with "authority.") Compilations of information have been protected by U.S. copyright law practically since its inception. Although long recognized as straining the copyright concepts of authorship and originality in expression, fact compilations have often been described as in need of copyright protection in order to provide adequate economic incentives to induce firms to invest in the creation and maintenance of this socially valuable kind of work. Though it is perhaps a misnomer to speak of the work required of one who compiles facts as "sweat-of-the-brow," there is no question but that it is an arduous and expensive task to gather information, compile it, make sure the information is accurate and the gathering complete, print the compilation, prepare updated versions, and so on. If, after a firm has spent considerable sums of money to collect and clean up the data, someone else can print a compilation of the same facts in direct competition with the first compiler without undergoing the effort and expense, incentives to produce compilations may be lacking. Largely in response to this argument, courts developed the "sweat-of-the-brow" originality standard for copyright protection for directories. Many cases embraced the theory before the Feist decision. Some cases, although declining to apply the "sweat-of-the-brow" standard in other kinds of cases, seemed to accept the standard as applied to directories. Still other decisions rejected the theory. One reason the Supreme Court probably took the Feist case was to resolve the conflict in the lower court case law on what originality standard should be used in fact compilation cases. But Rural was surely correct in its claim that there had been a rather large number of decisions upholding copyrights in white pages and other similar compilations which were ample precedent to support a ruling in Rural's favor--at least until the Supreme Court's decision in its dispute with Feist. A Statutory Argument Against "Sweat-of-the-Brow" In 1976, Congress passed a revised copyright statute which made all "original works of authoriship . . . fixed in [a] tangible medium of expression" protectable by copyright law. Two provisions of the new statute referred to compilations. One defined the term compilation as "a work formed by the collection and assembly of preexisting materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a whole is an original work of authorship." The second concerned the scope of copyright protection available to compilations. It states that the "copyright in a compilation . . . extends only to the material contributed by the author of such work, as distinguished from the preexisting material employed in the work, and does not imply any exclusive right in the preexisting material." Justice Sandra Day O'Connor's opinion in Feist points out that in order for the industrious collection (or sweat-of-the-brow) theory of originality to be consistent with the statute, the statutory definition of compilation would have to end after the word "data." Instead, the definition goes on to require selection, coordination, and/or arrangement of the preexisting materials. In addition, this selection, coordination, or arrangement must be substantial enough to make the work as a whole original enough to be copyrightable. The Court construed these provisions of the 1976 Act to mean the originality in a compilation must lie in the selection, coordination or arrangement of data. That, and that alone, is eligible for copyright protection as the "expressive" contribution of the compiler. Because Rural was required by law to publish a directory with listings of its customers and their telephone numbers, Rural's white pages lacked originality in selection. The Court also precieved no originality in Rural's alphabetical arrangement of customers' names and phone numbers in columns on the pages of the directory. Although repeating the oft-cited statement that the standard of originality in copyright law is low, the Court found the white pages of Rural's telephone directory to be "utterly lacking in originality," and "devoid of even the slightest trace of creativity." It was, consequently, not protectable by copyright law. Nor was Feist's copying from it copyright infringement. Unlike some other fact compilers, Rural still had plenty of incentive to publish the white pages listings, for Kansas had made publication of such a directory one of the conditions for Rural's ability to provide telephone service to customers in its allotted region. Also, unlike many fact compilation cases, Rural distributed its directories free. The Supreme Court, however, did not rest its decision against Rural on its being a case in which there was no incentive-to-publish problem. In fact, the Court did not even mention this issue. One reason Feist seems to lawyers like such a broad decision is that this possible ground for distinguishing between telephone directory and other kinds of fact compilation cases did not appear to affect the Court's judgment at all. Had the Supreme Court's Feist decision simply made the points about the lack of originality in Rural's selection and arrangement of data, its decision, even if displeasing to the telephone companies, would not have been controversial. But that is not where the Court's analysis dropped. If the Court's opinion dropped a "bomb" on copyright law, it came from the additional constitutional ground it gave for its ruling in Feist's favor. The Constitutional Argument Against "Sweat-of-the-Brow" Congress was granted power in a special clause of the U.S. Constitution to enac laws that would promote the progress of science and the useful arts by conferring on authors and inventors exclusive rights to their respective writings and discoveries for limited times. It is to this clause that the federal copyright and patent laws owe their origin. Copyright law is said to fulfill the constitutional mandate in promoting the progress of "science" (by which the drafters of the Constitution meant "knowledge") by giving "authors" exclusive rights in their "writings." Patent law promotes the progress of the "useful arts" by giving "inventors" exclusive rights in their "discoveries." The constitutional clause has sometimes been said to impose limitations on Congress' power, as well as being a grant of power. Under this clause, for example, Congress is understood to be unable to grant perpetual patent rights, for that would contravene the "limited times" provision of the constitutional clause. Similarly, some cases suggest that Congress does not have power to give exclusive rights to those whose improvements in the useful arts do not rise to the level of an "invention," for the clause speaks only of giving rights to "inventors." One century-old Supreme Court decision struck down the first federal trademark statute as beyond Congress' power because the constitutional clause spoke only of authorizing Congress to grant rights to "authors" and "inventors." These constitutional limitation decisions have been out of fashion for some years. Few intellectual property lawyers expected the Supreme Court in Feist to reinvigorate these musty old cases, but it has done so. The Supreme Court in Feist states that "originality" of expression is a constitutional requirement in order for copyright protection to be available to a work. That is, in order for someone to be considered an "author" within the meaning of the constitution, and in order for the work to be a "writing" under the constitutional clause, there must be originality in expression in the work. Originality, the Court decided, requires that there be some "creative spark"--not much, but some--present in the work. The Court in Feist seems to say that Congress is constitutionally forbidden to amend the copyright law to extend copyright protection to uncreative works, such as the white pages of a telephone directory, even if it wanted to. Some lawyers think Congress might still be able to pass a sui generis (of its own kind) statute to protect uncreative compilations, just as it did some years ago when it passed the Semiconductor Chip Protection Act which was said to be partly authorized by the patent and copyright clause of the Constitution and partly by the clause that gives Congress the power to regulate interstate commerce. Such a law, however, had better not look too much like a copyright law, or the Court might regard it as an evasion of its ruling in Feist. There are hints in the decision that fact compilations might still be protected against some kinds of unfair competition. However, given some other statements the Feist decision makes about the unprotectability of facts, it would seem that merely copying information to prepare a competing directory would not be the kind of unfair competition the Court thinks might be unlawful. The questions the Feist decision raises about Congress' constitutional powers and what kinds of unfair competitive copying might be illegal help to explain why some regard the Court's decision as a bombshell. It is puzzling that the Supreme Court based its ruling in Feist on this constitutional theory, given there was an adequate statutory basis for its ruling in Feist's favor. In general, the Supreme Court does not rule on a constitutional issue when there is an adequate basis for its ruling on another ground. It did not explain why it made an exception to this general rule in Feist. Law professors will be kept busy for some time speculating on why the Supreme Court made a constitutional ruling in this case. This sport is unlikely to be of interest to Communication's readers, but what the Court had to say about why facts are unprotectable by copyright law should be of interest. Why Facts are Unprotectable by Copyright Law The Court in Feist regarded it as a fundamental axiom of copyright law that although original compilations of facts can be protected by copyright law, facts themselves cannot. One reason the Court gave for its assertion that facts are uncopyrightable was that facts are "discovered" by humans, rather than being "created." (Section 102(b) of the copyright statute states that even when embodied in a copyrighted work, "discovieries" are not protectable by the copyright.) That is, facts are not "original" to a human author who might write about them, although the collocation of words used by an author to describe the facts would be "original" in a copyright sense. There are some interesting epistemological assumptions underlying the Court's assertion that facts are discovered whereas texts are created, and it would be fun to debate this issue. This aspect of the Feist ruling brings to mind a similar issue raised in the Supreme Court decisions concerning the unpatentability of algorithms, which the Court regards a having been "discovered" by humans (like laws of nature or scientific truths which are also said to be unpatentable for this reason), rather than being "invented" by them. Yet, another manifestation of a related assumption underlying U.S. intellectual property laws is the position of the Copyright Office that computer-generated works are unprotectable by copyright law because they lack a human author. The works, therefore, cannot be original in a copyright sense. Rural admitted that individual facts were not copyrightable, but insisted that compilations of them were. The Court, however, did not think the presence of 100 or 1,000 facts in one place could magically change the copyright status of the facts from unprotectable to protectable elements of the work. A compilation's only characteristic making it protectable by copyright law, said the Court, was originality in the selection and arrangement of the facts. The Court recognized that this inevitably meant copyright protection for a fact compilation would be "thin," for the "raw facts" in the compilation, said the Court, could be "copied at will," even by someone preparing a competing work. In order not to infringe, the second compiler must merely make his or her own selection and arrangement of the facts. The Court recognized that some might think it unfair that a second compiler could make uncompensated use of the valuable fruits of the first compiler's labor. But the copyright rule that facts can be copied at will was, the Court said, not some unfortunate or unforeseen byproduct of the copyright scheme. It was, instead, its very essence. The ultimate purpose of copyright law, said the Court, was not to maximize the financial return to individual authors, but to promote the dissemination of knowledge. Allowing second-comers to copy facts in preparation of another work would aid in achieving the constitutional purpose of the copyright. This was yet another reason the "sweat-of-the-brow" theory was rejected by the Court. By its insistence that facts (as least published onces) are not the property of those who compile them and that facts need to be in the public domain in order to promote the constitutional purpose of the copyright, the Feist decision stands in striking contrast to many other court decisions issued in recent years. In the March 1991 "Legally Speaking" column, I discussed some court decisions saying information could be property, theft of which could be criminally prosecutable. To be consistent with the Feist decision, those earlier decisions should now be construed as making only unpublished information the property of its possessor. In Feist the Court revives the traditional legal approach toward published information, namely that no one owns it. Originality in Selection and Arrangement After Feist The selection of preexisting materials (information or anything else) is one source of potential originality in a compilation. Pablo Picasso's collages of newspaper clippings are an example of an artistic type of selection. A more typical case in which selection has been held to be expressive was the Eckes case, in which someone claimed copyright protection for his selection of the 5,000 "premium" baseball cards out of approximately 18,000. When another firm copied this same selection in order to provide updated price information for those 5,000 cards, it was found guilty of copyright infringement. The compiler's subjective judgment in choosing these 5,000 cards seems to have been what made the selection original. Many times, however, the only creativity in selection that a compiler may make is in deciding what categories of information to collect. There is one post-Feist decision ruling that while this kind of selection may satisfy the copyright standard, there may be limits to how far courts will go on this kind of selection theory. Some categories will seem natural to the kind of information being collected (when one collects names, one often collects addresses and telephone numbers). Others may reflect a method or system the compiler may have devised (a method, for example, concerning baseball player performance may involve selection of specific kinds of data about the players). Since methods and systems are not protectable by copyright, categories of information that are constituent elements of the method or system may not be original enough to support a copyright in the compilation. Selection is a limited kind of theory for copyrighting compilations, in part because there are many compilations that are chiefly valuable because they are complete (that is, because the compiler did not select among them, but included all of the information on a certain topic in the compilation). Rural's white pages is one example. Readers can no doubt think of others. After Feist, compilers may get more selective than they used to be, just in order to make their compilations more copyrightable. Coordination and arrangement of information is the other potential source of originality for copyrighting compilations. The arrangement of articles, ads and commentary in a particular issue of a journal is an example of a potentially expressive arrangement of preexisting materials in the compilation that might constitute the journal. So, too, may be the ordering of recipes in a cookbook or algorithms in a book of them. (The individual recipe or algorithm would probably not be protectable by copyright law because it is a "procedure" which the law says is an unprotectable idea.) We know from Feist that alphabetical ordering is not original enough to be protectable by copyright. Chronological and numerical orderings are likewise unlikely to be original enough to make a compilation protectable by copyright. Arrangements that are constituent parts of a particular method or system for arranging information may likewise be uncopyrightable. Originality in the idea (here, a method or system) for a compilation does not mean there will be originality in expression. This is illustrated in the old Baker vs. Selden case, in which the Supreme Court found no original expression in the arrangement of headings and columns in Selden's ledger sheets because the arrangement was in accordance with the bookkeeping system that Selden had devised. Readers may infer from this discussion that after the Feist decision, firms maintaining electronic compilations of data have reason to be somewhat nervous about the copyrights they claim in their compilations. The quirkier and more subjective the compilation, the more original and expressive it is likely to be, which will ensuer copyright protection for it. The more methodical the compilation is, the less original it may be, and the less protectable by copyright law. Yet methodical arrangements of data are more valued than quirky and subjective arrangements by people in the computing field. The implications of Feist for electronic information providers may be obvious, but let's spend a few minutes spinning them out. The Implications of Feist Many firms now derive substantial revenues from selling or licensing use of electronic information on specialized topics. They typically assert copyright protection in their databases of this information, although often also relying on licensing contracts with customers, and not just on copyright alone. For some of them, Feist probably did seem like the dropping of a bomb. The economic incentives concerns that are so important to their business were given no weight in the Feist opinion. One reaction of electronic information service providers to Feist may well be to cancel or modify plans to make their valuable data available on CD-ROMs for general distribution to the public. They may choose to rely more heavily on connect-time licensing arrangements. Another reaction might be to encrypt electronic data made available to the public and strictly control access to decryption devices. As noted in the October 1991 "Legally Speaking" column, one of the challenges digital works poses for the future is that it is so much easier and cheaper to copy digital information than to copy its print counterpart. The ease of digital copying makes it much more likely that firms will restrict access to their valuable digital products. Now that the Feist decision says it is lawful to copy facts as long as one arranges them in a different way (something any darned fool can do), the likelihood of restricting access to digital information may be even greater. Some may argue the Feist decision might have inadvertently contributed to disabling the achievement of the constitutional purpose of promoting the dissemination of knowledge. In reaction to Feist, electronic information service providers may restrict access to their databases and limit plans to make copies available for general distribution. As a result, less information may be disseminated than if copyright protection were available to the valuable compilation. On the other hand, Feist may spur the development of new kinds of electronic information services because now competitors may be able to offer new selections and arrangements of data, adding value to preexisting compilations by offering new services that will benefit consumers. Some of these services may enhance the value of the compilation on top of which they may be built, for in order to use the value-added service, users may also pay money for access to or rights in copies of the original compilation. As a result of Feist, some firms may also seek patents for methods of organizing data for inclusion in a computer database. While database patents are likely to be controversial within the technical community and some parts of the legal community, they may offer protection for certain kinds of compilations that copyright will not provide. It will not be surprising to see more electronic information providers relying on patents in the aftermath of Feist. Feist may also have some important implications for the scope of copyright law in computer program cases under U.S. law. Feist undercuts some of the rationale for strong copyright protection for programs, such as that found in the Lotus vs. Paperback opinion. The Court made clear in Feist that copyright-protectable "expression" should not be found merely because a work (or some aspect of it) is valuable or required considerable effort to produce, or because, without copyright protection, there might be insufficient incentives to invest in creation of certain kinds of works. The Court's strong reliance in Feist on the century-old Baker vs. Selden decision, and its endorsement of the "thin protection" doctrine which has long been associated with Baker vs. Selden, may also be significant for computer program cases. Further, the Court in Feist made clear that competitors have an affirmative right to copy valuable but unprotected elements of copyrighted works. After Feist, it would seem that courts would need to do more in computer program copyright cases than recite the number of ways in which something could be done (e.g., the number of different ways in which commands might be organized in a user interface) to justify a conclusion that "orignal expression" is present which can serve as a predicate to a finding of copyright infringement. Conclusion The white pages of telephone books do not quality for copyright protection under the Supreme Court's ruling in the Feist case (not withstanding the copyright notices on every page of the Bell of Pennsylvania phone book I have in my office). There was ample basis in the statute for the Court's ruling. While there are many statements in the Feist opinion that seem to suggest the ruling will have broad implications for all compilations of facts, how broad or narrow a ruling Feist really is will only be known in the fullness of time, as competitors copy information from one anothers' compilations and new litigations are begun to resolve the issues. The potential effect on electronic information providers is great. I will leave it to you to decide if the effect will be for good or for ill. Pamela Samuelson is a professor of law at the University of Pittsburgh School of Law.