MISUSE OR FAIR USE: THAT IS THE SOFTWARE COPYRIGHT QUESTION

JAMES A.D. WHITE

TABLE OF CONTENTS

I. INTRODUCTION

II. THE UTILITARIAN ROLE OF INTELLECTUAL PROPERTY IN OUR ECONOMIC SYSTEM

A. The Utilitarian Public Policy Rationale For Intellectual Property Protection In The United States

B. Balancing The Rights Accorded Copyright And Patent Holders With Those Retained By The Public

C. Balancing The Rights Granted By Federal Intellectual Property Law With Those Accorded By State Law

III. THE ROLE OF INTELLECTUAL PROPERTY MISUSE IN ENFORCING UTILITARIAN PUBLIC POLICY

A. The History Of The Patent Misuse Doctrine

B. The Recent Emergence Of The Copyright Misuse Doctrine

IV. THE PROPER TEST TO ENHANCE UTILITARIAN PUBLIC POLICY: ANTITRUST "RULE OF REASON" OR MISUSE "SCOPE OF THE GRANT"?

A. The Misuse Analysis Provides Advantages Over An Antitrust Analysis In Balancing The Rights Granted

B. The Unique Status Of Software As A Copyrighted Work

C. A "Scope Of The Grant" Copyright Misuse Doctrine Should Be Applied To Computer Software

V. THE PAST USE OF OTHER JUDICIAL DOCTRINES TO LIMIT SOFTWARE COPYRIGHT IN MISUSE CONTEXTS

A. Recent Growth Of The Software Industry Creates A Need For Copyright Limitations

B. Employing Fair Use To Prevent Misuse

C. Employing Other Copyright Doctrines To Prevent Misuse

VI. APPLYING SOFTWARE COPYRIGHT MISUSE TO TRADITIONAL MISUSE CONTEXTS

A. Software Copyright Misuse And Tying Arrangements

B. Software Copyright Misuse And Restrictions On Competition

VII. APPLYING SOFTWARE COPYRIGHT MISUSE TO CURRENT SOFTWARE PRACTICES

A. The Licensing Rather Than Sale Of Software

B. License Restrictions On Reverse Engineering

VIII. CONCLUSION

I. INTRODUCTION

The emergence of the copyright misuse doctrine is one of the most significant trends in copyright law in recent years. A basic premise of our intellectual property system is that various rights granted to creators as an incentive to produce intellectual property must be carefully balanced with the rights retained by the public. The judicial doctrine of intellectual property misuse limits efforts of intellectual property creators to inappropriately extend their rights and alter this careful balance. While the patent misuse doctrine has for decades prevented patent holders from unduly extending the rights granted to them beyond the scope of the patent, an appellate court did not extend an analogous limitation to a copyright owner until 1990. In the ensuing period, a debate has emerged whether the copyright misuse doctrine should exist, and if so, whether it should exist outside the scope of an antitrust violation.

This paper argues that the copyright misuse defense should exist, and that at least in the context of computer software it should be separate from antitrust analysis. Computer software is distinct from other forms of copyrighted works in various ways, providing more market power to the copyright owner and impacting federal patent law in ways that require restraints on the actions of software copyright owners. In addition, application of traditional antitrust analysis does not extend well to the software industry because actions that evade antitrust law are nevertheless misuse of copyright.

After examining how courts have previously applied doctrines such as "fair use" to restrict over-reaching by software copyright owners, the paper concludes that a software copyright misuse doctrine would provide a more appropriate vehicle for defining the scope of rights accorded to software copyright owners. Finally, the paper briefly employs a misuse analysis to consider various situations that occur in the software industry.

Part II of this paper begins by examining the role that intellectual property plays in our economic system, briefly tracing its history and examining the balance of rights between creators and the general public. Part III provides a brief history of the development of the intellectual property misuse doctrine. Part IV examines the advantages and disadvantages of an antitrust approach to copyright misuse, and after exploring the unique aspects of computer software relative to other copyrighted works, concludes that a software copyright misuse doctrine should exist independent of antitrust law. Part V considers the relatively recent arrival of the software copyright misuse doctrine, and concludes that, in the absence of such a doctrine in the past, courts have used other doctrines to limit software copyrights. Parts VI and VII briefly apply a copyright misuse analysis to various practices in the software industry.

II. THE UTILITARIAN Role of Intellectual Property in our economic system

Commentaries on the copyright misuse doctrine often refer exclusively to the patent misuse doctrine and antitrust laws when reaching competing conclusions as to the correct application of copyright misuse. Many proponents of a strong copyright misuse doctrine argue that patent misuse has a long history, and that the similarity of copyright law to patent law thus legitimizes a copyright misuse doctrine in analogous situations. Alternately, proponents of restricting copyright misuse to a purely antitrust analysis focus only on how a copyright misuse defense would affect the application of antitrust principles.

When examining the proper scope of copyright protection for software, it is instructive to explore the historical application of the copyright1 and patent2 laws in a broader sense, and to review other situations which require limitations on the rights of patent and copyright owners. This exploration will illustrate that the patent and copyright laws have indeed evolved together and that they are based on a common public policy of benefiting society through the encouragement of creation, discovery and dissemination of novel ideas and creative expression. It will also illustrate that this utilitarian goal is achieved through the grant of property rights in a limited monopoly to inventors and creators.

Although society wants to provide incentives to create, underlying public policy necessitates the careful balancing of the rights granted to creators with the rights retained by the general public. The rights which are granted to creators are determined uniformly through statutory grants and limitations, but subjective judicial doctrines allow courts to modify these rights in individual situations. Enforcement of an appropriate balance prevents reordering of these rights, whether the reordering occurs through state law or by private action. While the misuse doctrine can be used to prevent such reordering, it is only one of many limitations placed on the rights of intellectual property holders and it should be applied in a manner consistent with these other limitations.

A. The Utilitarian Public Policy Rationale For Intellectual Property Protection In The United States

The current federal copyright and patent laws share underlying public policy rationales, and are the result of a similar evolution over time. As with much of the law in the United States, the impetus for federal patent and copyright laws originated with English law, which strove to benefit society by encouraging the creation of new inventions and new works of authorship.3 In continuing this English tradition, the copyright and patent laws in the United States are designed to benefit society as a whole by providing incentives to creators.4 These federal laws stem from a constitutional grant of power to Congress: "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."5 The copyright/patent clause is notable in that it explicitly mentions the purpose behind the grant of power, an unusual occurrence in the Constitution.6 By placing the copyright and patent grant of power together in this manner, and by explicitly incorporating the purpose of the grant in the Constitution, the framers indicated the importance of these doctrines and continued the English tradition of encouraging the creation of intellectual property.7

American courts have affirmed the utilitarian nature of intellectual property rights8 by repeatedly emphasizing that the federal intellectual property rights granted to creators serve as an incentive to produce. "The economic philosophy behind the clause empowering Congress to grant patents and copyrights is the conviction that encouragement of individual effort by personal gain is the best way to advance public welfare through the talents of authors and inventors in 'Science and useful Arts.'"9 Thus, both patent law and copyright law derive their underlying authority from the common public policy of benefiting society by granting incentives for the production of novel ideas and creative expression.

B. Balancing The Rights Accorded Copyright And Patent Holders With Those Retained By The Public

In order to achieve the public policy objectives underlying copyright and patent law, it is necessary to limit the rights that are granted to the owners of copyrights and patents.10 If the ultimate utilitarian goal is to increase the number of inventions and works of authorship that are available to the public, then legal monopolies granted to creators can have a negative short-term effect on this goal. To the extent that creators can restrict access or use of their creation, whether absolutely or through exorbitant monopoly rents, the public fares less well than if the same creation had been available without legal restrictions. However, the premise of utilitarian theory is that some incentive is necessary to induce the creation in the first place-that is, the public is better off with a new creation at a monopoly price than with no creation at all. Furthermore, the limited duration of the rights granted guarantees that at some point the creation will enter the public domain and become freely available to the public.

Since the legal incentives granted to creators deprive the public of a benefit in the short term, the utilitarian goal of the intellectual property system can only be fully achieved if the incentives granted to creators are the minimum necessary to spur the creation and dissemination of inventions and works of authorship. In an effort to achieve this optimal level of innovation at the lowest public costs, the rights granted to owners of copyrights and patents are carefully balanced with the rights retained by the public. In a world with perfect information and zero transaction costs, the optimal solution could be attained by varying the rights granted to each creator on an individual basis. Thus, each creator could be compensated at the minimum level necessary to compensate for each creation produced, regardless of the underlying value of the creation.

Transaction costs render it impractical to grant rights to each creator on an individual basis. Therefore, the legal system instead imposes systemic constraints on the rights of creators. These constraints take the form of both statutory limitations and judicial doctrines. While some systemic constraints are objective limitations that apply to all creators equally, others are subjective or equitable in nature. To the extent that subjective and equitable constraints are imposed on an individual basis, these efforts can be viewed as an attempt to tune more finely the balance of rights granted to a particular creator. Conversely, courts can enforce subjective doctrines broadly and use them to alter the balance of rights granted to creators either for a broad class of transactions, or for a type of protected subject matter.

1. Statutory Definitions Of Copyright And Patent Rights

The statutory grants of power under chapters 17 and 35 of the United States Code define federal intellectual property rights of creators. In the absence of such explicit statutory grants, a creator would not enjoy any federal intellectual property protection for their creations. Under the protections granted for both copyrights and patents, the federal statutes include broad provisions that grant power to the creator, and narrower provisions that restrict the granted power in a variety of specific ways.

A patent conveys the right to exclude others11 from making, using, selling, offering for sale, and importing the patented invention.12 This right can be enforced through a combination of injunctions,13 damages,14 and attorney's fees.15 While these rights can convey significant economic power for a useful invention, the statutes limit the rights in various ways. As a threshold matter, a patent is only granted to inventions that are "worthy" of these legal protections. That is, to be patentable, an invention must have the appropriate statutory subject matter,16 be novel,17 be useful18 and be non-obvious.19 Further, the patent rights granted are of a limited duration.20 In addition, the rights of patent owners are statutorily limited in other ways.21 Through these statutory grants, patent law provides relatively broad rights that last for a relatively short time.

A copyright, on the other hand, conveys the affirmative exclusive right to reproduce, distribute to the public, perform publicly, display publicly, and prepare derivative works of the copyrighted work.22 Authors of works of visual art have additional rights of attribution and integrity,23 and performers of live musical performances have additional rights to control the use of their performances.24 Copyright protection is available for original works of authorship that are fixed in any tangible medium of expression25 and lasts for a limited duration, but does not prevent independent creation of the same work.26 As with patent law, the copyright statute provides various statutory limitations on copyright owners,27 including special provisions for certain classes of users such as libraries28 and instructors,29 and special provisions for certain classes of copyrighted works such as computer programs.30 In addition, section 109 provides limitations on a copyright owner's ability to control an item embodying a copyright after the item has been sold (the "first-sale" doctrine),31 and section 107 codifies the equitable "fair use" limitation which permits certain types of uses of copyrighted works.32 These statutory copyright grants provide relatively narrow rights that last for a relatively long time.

2. Judicial Modifications Of Copyright And Patent Rights

While the copyright and patent statutes provide the primary definition of the rights granted to creators, the courts have adopted judicial doctrines which alter the scope of these rights. Over time, some of these doctrines were codified and became part of the statutory framework. For example, the "fair use" doctrine of copyright law originally arose judicially to restrict the rights of copyright owners, but is now statutorily codified.33 Although it is possible for judicial doctrines to provide an objective expansion or limitation of a creator's rights, they generally are more subjective and equitable in nature. As such, courts can enforce these doctrines on an individual basis and can alter the rights granted to an individual creator to reflect the amount of protection that is needed or deserved. Alternatively, courts can enforce such doctrines broadly and use them in such a manner so as to alter the balance of rights granted to creators for a class of transactions or for a type of protected subject matter.34

Two notable judicial doctrines used in patent law are the requirement of non-obviousness for patentability35 and the doctrine of equivalents for determining the breadth of claim scope.36 While courts describe these doctrines as objective standards,37 both doctrines involve subjective tests that provide results which are notoriously difficult to predict. For example, consider an invention that involves the combination of two techniques, A and B, and prior art references that describe either A or B but never their combination. An objective standard might hold that the combination of the two techniques is not obvious unless there is an explicit suggestion in the prior art to combine these techniques. Instead, the judicial determination is whether the combination of these techniques was obvious at the time of the invention to one of ordinary skill in the art. This determination is inherently subjective.

In the area of copyright law, the judicial doctrine of "fair use" is an extremely broad, subjective limitation on the rights of copyright holders. The fair use doctrine permits certain uses of copyrighted material (e.g. making copies of copyrighted material for news reporting or criticism), despite a literal violation of prohibited statutory activities. Although now codified in the copyright statute,38 the determination of fair use is inherently subjective and thus dependent on judicial implementation.39 While fair use has traditionally been viewed as a limitation on the rights of the copyright owner (i.e., a balancing of the rights that favors the public),40 some scholars have conceived fair use in certain cases to merely reflect market failure; if the transaction costs to obtain permission to use a right are higher than the value of the right, it would not warrant granting that right.41 Regardless of the underlying conception, such a broad-based doctrine gives significant power to the courts to modify the rights granted by copyright law. While the doctrine is intended to be a fact-intensive determination,42 and thus variable in each case, sweeping pronunciations by the courts can be regarded as determinative of a broad class of similar situations.43 In this manner, courts can use a powerful subjective doctrine such as fair use to alter the balance of rights granted to creators in a broad class of situations.

C. Balancing The Rights Granted By Federal Intellectual Property Law With Those Accorded By State Law

The federal intellectual property laws are based upon a utilitarian public policy rationale that necessitates balancing the rights granted to creators with those retained by the public. The statutory framework of the copyright and patent laws, and their corresponding judicial doctrines, have evolved to adjust and protect this balance. However, the federal laws do not provide the only means to grant rights to creators or the public-both state laws and private contractual agreements can alter this balance of rights. Since our legal system grants significant deference to both state and private action, federal law will preempt such action only if necessary for an important federal purpose. Courts have recently faced issues of state and private actions that significantly alter the balance of rights granted under federal law, and have been forced to decide whether to preempt these actions. In resolving the preemption issue, courts have relied on the constitutional authority of federal intellectual property laws and the Supremacy Clause44 of the Constitution, which dictate that state action, including state-enforced private action, may not encroach on that authority.

In the 1964 companion cases of Sears, Roebuck & Co. v. Stiffel Co.45 and Compco Corp. v. Day-Brite Lighting, Inc.,46 the Supreme Court gave a broad reading to the constitutional preemption of state laws relating to intellectual property. In both cases, state unfair competition laws prohibited the copying of products that were neither patented nor copyrighted. In Sears, the court stated that "[w]hen state law touches upon the area of these federal statutes, it is 'familiar doctrine' that the federal policy 'may not be set at naught, or its benefits denied' by the state law. [citation omitted.] This is true, of course, even if the state law is enacted in the exercise of otherwise undoubted state power."47 The court recently reaffirmed this policy in 1989 by striking down a Florida law that prevented the copying by direct molding of unpatented boat hulls in Bonito Boats, Inc. v. Thunder Craft Boats, Inc.48 The "state regulation of intellectual property must yield to the extent that it clashes with the balance struck by Congress in our patent laws."49 Thus, there are clearly constitutional limits on the laws which states may enact and enforce. At a minimum, state laws will be preempted if they provide protection to uncopyrighted or unpatented works in such a manner as to give the author or creator of those works a property right equivalent to that granted by patent or copyright law.

Although some state laws alter the balance of federal intellectual property rights sufficiently to be preempted, the Supreme Court found that state trade secret law does not. In upholding an Ohio law of trade secrets, the Court held that a state law is void under the Supremacy Clause only if it stands as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress.50 However, the Court noted that limits existed on how far trade secret law could be enforced. "If a State, through a system of protection, were to cause a substantial risk that holders of patentable inventions would not seek patents, but rather would rely on the state protection, we would be compelled to hold that such a system could not constitutionally continue to exist."51

Thus, it is difficult to determine whether a particular state or private action should be allowed to alter the balance of federal intellectual property rights. The Supreme Court has stated that while state regulation of intellectual property, which clashes with the balance struck by Congress, must be preempted, such regulation will clash in this manner only if it stands as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress.52 It is clear, however, that consideration of the purposes and objectives of Congress will be the touchstone to maintain the careful balance of federal intellectual property rights in a preemption analysis.

III. The role of intellectual property Misuse in enforcing utilitarian public policy

As previously noted, the utilitarian goal of intellectual property is accomplished by granting property rights in a limited monopoly to creators, and the underlying public policy necessitates the careful balancing of the rights and powers granted to the creator and the rights and powers retained by the general public. The judicial doctrine of intellectual property misuse was created to address situations in which the owner of an intellectual property right used his or her legal monopoly to create such an asymmetry in the balance of rights that the courts refused to enforce the normal intellectual property rights.53

A. The History Of The Patent Misuse Doctrine

The doctrine of intellectual property misuse first arose in the early 1900s in conjunction with the use of patents.54 In the 1917 case of Motion Picture Patents v. Universal Film Mfg. Co.,55 the patentee licensed its patented movie projector on the condition that the film used in the machine must be purchased from the patentee (a type of tying arrangement).56 The Court found that:

[s]uch a restriction is invalid because such a film is obviously not any part of the invention of the patent in suit; because it is an attempt, without statutory warrant, to continue the patent monopoly in this particular character of film after it has expired, and because to enforce it would be to create a monopoly in the manufacture and use of moving picture films, wholly outside of the patent in suit and of the patent law as we have interpreted it.57

In short, the Court denied relief to the patentee because the licensing restrictions attempted to extend the scope of the film projector patent into the unpatented area of film.

The Court extended the patent misuse doctrine in the seminal 1942 case Morton Salt Co. v. G. S. Suppiger Co.,58 and in doing so addressed the relationship of the misuse doctrine to antitrust law. In Morton Salt, the plaintiff held a patent on a machine that placed salt tablets into canned goods, and required licensees of its machine to purchase the unpatented salt tablets for the machine exclusively from the plaintiff.59 The defendant produced its own salt tablet-depositing machine that was modeled after the plaintiff's machine, and the plaintiff sued for direct infringement of its patent.60 Among other defenses, the defendant argued that the plaintiff misused its patent through its licensing agreements. The Court, without ruling on validity or infringement, found that the plaintiff's tying arrangement constituted patent misuse. In so doing, the Court prevented the plaintiff from enforcing its patent,61 despite the lack of any connection between the defendant's infringing machine and the plaintiff's tying actions.62 In addition, the court found that an antitrust violation was not necessary for a finding of misuse.63

The courts and Congress continued to develop and refine the doctrine of patent misuse after Morton Salt. While patent misuse has been analyzed in a number of factual settings,64 two paradigmatic cases emerged where misuse was typically found-tying arrangements where the patentee requires the purchase of unpatented goods or services along with a patented product or process, and non-compete clauses that prevent a patent licensee from producing or selling competing goods.65 The analysis was changed for tying arrangements in 1988 when Congress passed the Patent Misuse Reform Act66 which modified 271(d) of the Patent Act, and held that patent tying arrangements are no longer per se misuse.67

Thus, the patent misuse defense continues to be an important judicial doctrine, with some recent statutory modifications for certain types of misuse cases. However, the misuse doctrine has not been limited solely to the patent arena. It has also been applied in trademark law and has recently emerged in the area of copyright.

B. The Recent Emergence Of The Copyright Misuse Doctrine

Although the copyright misuse doctrine was articulated by the courts only recently, the doctrine has been implicitly recognized since the time of Morton Salt. In the 1948 case of United States v. Paramount Pictures, Inc.,68 the Supreme Court affirmed the lower court's finding that Paramount's block-booking of movies was an antitrust violation by relying on the reasoning of the patent misuse cases.69 The Court was more explicit in a later block-booking case, United States v. Loew's Inc.,70 in analogizing the patent misuse cases to the area of copyright.71 Furthermore, statements by other courts indicated that copyright misuse might have been recognized under different sets of facts.72

Despite this long history, the copyright misuse doctrine was not applied by an appellate court until 1990.73 In that year, the Fourth Circuit decided Lasercomb America, Inc. v. Reynolds,74 in which the plaintiff, Lasercomb, had developed and licensed software for the computer-aided design and manufacture of steel rule dies used to score paper for folding into cartons.75 The defendant, Reynolds, had purchased four licenses for the software. However, rather than purchasing additional licenses, Reynolds chose to circumvent the copy protection devices included with the software and made three additional unauthorized copies which Reynolds used on its own systems.76 Reynolds then created an almost exact copy of Lasercomb's software, and marketed this program in competition with Lasercomb's product.77 Because the defendant purposefully copied the software and made misrepresentations to Lasercomb, there was no question of copyright infringement.78 However, in addition to limiting copying of licensed software, Lasercomb also included a provision in its standard form contract that prevented a licensee from independently developing a competing software program for 100 years.79 Although the defendants were not themselves subject to this licensing provision, other licensees of Lasercomb had agreed to this provision,80 and the defendants argued that Lasercomb had misused its copyright through the use of this non-competition provision.

After a detailed analysis, the Fourth Circuit accepted the argument of the defendants, and found that the plaintiff's license agreements amounted to copyright misuse because the license provision inhibited independent creation for a lengthy period of time (the 100 year term was potentially longer than the life of the copyright itself). 81 In so ruling, the court held that Lasercomb's copyright was unenforceable until the plaintiff had purged the effects of its misuse.82 In its analysis, the court acknowledged that the existence of the copyright misuse defense was unclear, and proceeded to conduct an extensive analysis of the history of the misuse doctrine and the public policy underlying the patent and copyright laws. The court finally concluded that the patent misuse doctrine first enunciated in Morton Salt should be incorporated into copyright law,83 and found copyright misuse despite the fact that the actions of the plaintiff had no direct effect on the defendants84 and despite the fact that no antitrust violation had been found.85 Since the Lasercomb decision, four other circuit courts have recognized the doctrine of copyright misuse.86

IV. The proper Test To ENhance Utilitarian Public Policy: Antitrust "Rule of reason" Or Misuse "Scope of the grant"?

Although most courts have accepted that some version of the copyright misuse doctrine should exist, questions remain over its application. One view holds that misuse should be found only when an antitrust violation occurs. In general, the antitrust laws are intended to prohibit situations that unreasonably harm competition, and can apply to various uses of intellectual property rights. Another view argues that the misuse doctrine should be applied when the scope of the intellectual property grant has been exceeded. This view asserts that the intellectual property laws have balanced the rights granted to creators with those granted to the public, and seeks to prevent the extension of creators' rights beyond the scope granted by the intellectual property laws. As such, this "scope of the grant" view of misuse is independent of antitrust criteria such as market power.

The extension of copyright to computer software significantly alters this analysis. Software possesses features that provide it greater market power than other copyrighted works. Moreover, the grant of copyright protection to computer software can create additional disruptions in the normal balancing of rights granted to creators. Due to these factors, the software arena needs a copyright misuse doctrine unencumbered by antitrust analysis in order to preserve the public policy rationales of copyright law.

A. The Misuse Analysis Provides Advantages Over An Antitrust Analysis In Balancing The Rights Granted

The decisions in both Morton Salt and Lasercomb endorse a strong version of intellectual property misuse which measures the actions of the intellectual property owner against the scope of the grant of intellectual property rights accorded to the owner. This scope of the grant view of misuse asserts that the intellectual property laws have established a ceiling on the level of rights granted to creators, and seeks to prevent any extension of creators' rights beyond this ceiling. Since this analysis is not related to restraints on competition, this version of misuse is independent from antitrust. In addition, those opposed to an antitrust-based approach to misuse have argued that the goals of the intellectual property laws and the antitrust laws are different, and that a misuse offense can occur even if antitrust laws are not violated.87

1. An Antitrust Analysis For Intellectual Property

While the explicit use of the copyright misuse doctrine is relatively recent,88 questions about the application of antitrust principles to patent and copyright situations have a long history. As was the case in Morton Salt,89 it is not unusual for an antitrust law claim to accompany a misuse claim. In general, the antitrust laws are embodied in the Clayton Act90 and the Sherman Act,91 and are intended to prohibit situations that may harm competition.92 It is generally necessary to show that the accused has market power in the relevant market, and courts subsequently use a "rule of reason" to analyze whether the contested practice has "unreasonably" restrained competition in light of the circumstances.93

Since the enforcement of intellectual property rights restrain the abilities of others to act, these rights can be considered to be limited, legally-imposed monopolies. However, despite these monopoly-like results, these aspects of intellectual property rights are part of the careful balance of rights, and should not implicate antitrust laws. Thus, while the enforcement of a patent or copyright will always involve some restraint of competition, it will not be an unreasonable restraint if applied within the normal scope of the intellectual property right.

2. A Comparison Between An Antitrust Analysis And A Misuse Analysis

While court decisions like Morton Salt and Lasercomb endorse the view that misuse can occur even without a violation of the antitrust laws, this view is not universal. The circuit courts are split on this point.94 Congress also examined this issue when it enacted the Patent Misuse Reform Act in 1988.95 In that year, Congress considered limiting patent misuse to those actions which violated the antitrust laws in 1988, but declined to take such action.96 Adherents to the Chicago School of Law and Economics,97 including Judges Posner and Easterbrook of the Seventh Circuit, also argue strongly for an antitrust-based approach to misuse analysis98 and they have been joined by other commentators.99

Proponents of the scope of the grant approach have countered several of the antitrust proponents' arguments. Two typical arguments raised in support of an antitrust-based approach to misuse stem from a belief that antitrust law and intellectual property public policy are both based on similar goals (that of promoting consumer welfare through a combination of free competition and innovation),100 and that antitrust principles provide more certainty than those of misuse.101 Several commentators, however, have pointed out that the intellectual property goal of stimulating the creation and distribution of creative works is different from the antitrust goal of encouraging marketplace competition,102 and that there are instances in which an antitrust inquiry alone fails to prevent all abuses of the intellectual property grant that may harm the public.103

An antitrust analysis can also be particularly difficult to apply to the area of software. While an antitrust analysis is often considered to be more objective and predictable than a "scope of the grant" determination, commentators have argued that the antitrust analysis is no more certain than a misuse analysis in the area of software.104 In particular, it can be extremely problematic or impossible to define the relevant market for software goods. Since antitrust law depends on the concept of market power, it can thus be difficult to apply to the software industry. For instance, software goods are frequently bundled with other products (including other software, hardware, or with services),105 making measurement of market share difficult. Producers of software often leverage products in other seemingly unrelated markets, and use the desire for compatibility to achieve market power in seemingly unrelated areas.106 Because these interconnections between unrelated markets in the software arena are difficult to define, the traditional antitrust analysis can lead to unpredictable results.

Thus, it has been argued that an antitrust-based approach to analyzing misuse is appropriate because of the common goals of antitrust law and intellectual property law, and the certainty provided by an antitrust analysis. However, many commentators have refuted these arguments, particularly as they apply to the software industry. Therefore, the justification for an antitrust-based approach to misuse is at best uncertain.

3. Antitrust Law Does Not Apply Well To Network Economic Situations

While certainty and common goal arguments have been mentioned by commentators as reasons for an antitrust-based analysis for misuse, the primary thrust for an antitrust approach arises from the belief that antitrust's economic analysis allows for a more efficient utilization of the intellectual property rights granted.107 For example, this argument (based on the Chicago School of Economics) asserts that an intellectual property right conveys only a limited amount of power, and that normally tying arrangements cannot extract more than that amount.108 Moreover, ties can serve as a form of metering by allowing the seller to profit based on the amount of use (by extracting most or all of the profit from the tied item if it is consumed during use), and this metering can allow price discrimination which simultaneously maximizes access to the public and profits to the intellectual property owner.109

However, a relatively recent "network theory" of economics attacks these traditional antitrust economic assumptions. Adherents to network economic theory argue that traditional antitrust assumptions do not apply to certain industries (such as the software and Internet areas) and that traditional antitrust analysis is therefore inapplicable to these areas.110 These adherents assert that in certain industries, systematic tendencies exist for inefficient technologies to become established and to resist replacement by superior alternatives.111 These tendencies are particularly powerful if there is a strong need for compatibility between different purchasers or users of the products supplied. While it has long been clear that this phenomenon existed in situations where users were physically connected (e.g., a telephone system has little use if different people have different types of phones which cannot inter-communicate), commentators have recently realized that the same phenomenon can apply in the software industry where users need compatibility (e.g., users need to share files with other users and software programs need to interact with other software programs).112 As users increasingly become interconnected through the Internet and intranets,113 the problem in the software industry is exacerbated. These types of network situations do not fit well within the framework of antitrust law.

4. The Antitrust Analysis Does Not Address The Proper Scope Of Rights For Creators

If this network theory of economics is correct, an antitrust analysis based on traditional economic theory will not identify many misuse situations in the software industry. But even if the analysis of the Chicago School is correct, allowing the full utilization of a granted right is only beneficial and efficient for society if that granted right conveys the appropriate degree of power. If the intellectual property owner is trying to enforce rights that go beyond those which they should rightfully possess based on public policy considerations, then allowing the full utilization of the owner's desired scope of rights will be counter-productive to the societal goals and public policy underlying the original granting of the right. Therefore, if enforcing a judicial doctrine such as intellectual property misuse is perceived to maintain the appropriate balance of rights between those granted to the creator and those retained by the public, then arguing for full utilization of granted rights is premature because the scope of the rights granted to a creator must be defined before the rights can be enforced.

However, using a judicial doctrine to define the appropriate scope of rights for a "scope of the grant" analysis has its own problems. In particular, such a scheme would create debilitating uncertainty if intellectual property owners do not know the extent of the rights granted to them until after judicial intervention in an enforcement action. In this situation, any benefits achieved by a careful balancing of rights would likely be overshadowed by the transaction costs associated with the system. Such a system would thus be unworkable if each individual creator needed to rely on a judicial pronouncement for a definition of his rights.

One solution, however, is to balance the rights for a class of works rather than for each individual work.114 A small number of judicial decisions could then define the scope of the rights granted to creators of works in each class, and the resulting uncertainty would be no more than exists in any common law system of jurisprudence, including that of antitrust law. Of course, such a system would be practical only if the determination of class membership was simple, and would be reasonable only if the works in the class had shared features which altered the normal balance of rights and thus justified a judicial restriction on the standard statutory grant.

B. The Unique Status Of Software As A Copyrighted Work

As discussed above, some commentators argue that a judicial doctrine of intellectual property misuse is appropriate only if an antitrust-based analysis is used, primarily because an intellectual property owner should be able to engage in a full utilization of the intellectual property rights granted. Even if the underlying theory of economics behind this argument is valid as applied to the software industry, the argument is preempted if the misuse doctrine is needed to redefine the appropriate extent of rights that are granted in the first place. But if a judicial determination of rights engenders excessive uncertainty on the part of intellectual property owners and others, then the system will be unworkable. This can only be avoided if the scope of the rights granted can be generalized to a class of works, which then provides the necessary predictability for other works in that class. For the reasons outlined below, computer software is such a class of copyrighted works.

1. The Uneasy Balance Of Software And Copyright Law

An examination of the history of copyright protection for computer software demonstrates a tumultuous road. Because software does not fit well within the traditional confines of copyright law, many difficult issues have arisen, and many still remain to be resolved.

While some would argue that computer programming first began in the nineteenth century with Charles Babbage and his Analytical Engine, modern computer software traces its origins to the creation in the 1940s of the first electronic computer. In response to copyright issues raised by the increasing prevalence of computer software in the 1960s and 1970s, Congress established in 1974 a National Commission on New Technological Uses of Copyrighted Works (CONTU). Based on the 1978 CONTU final report, Congress amended the Copyright Act in 1980115 to make explicit that copyright protection extended to computer software.

In the seventeen years since the 1980 amendments to the Copyright Act, the following major issues are among those that have arisen related to software copyrightability: whether software object code (as distinct from software source code) is copyrightable;116 whether software copyright protection violates the Baker v. Selden117 prohibition on protection for a system described in a copyrighted work;118 whether software copyright protection violates 17 U.S.C. 102(b)'s prohibition on protection for ideas, procedures, systems or processes;119 whether the structure, sequence and organization of a program are copyrightable (as opposed to the textual computer code);120 whether computer menu command hierarchies are copyrightable;121 whether the "look and feel" of computer screen displays are copyrightable;122 whether temporary copies of software in Random Access Memory (RAM) are infringing;123 whether reverse engineering of software is infringing;124 and whether attempts to deny software purchasers the statutory rights granted to "owners" (such as the right to resale or to make backup copies) can be enforced through mass-market shrink-wrap licenses.125

For many of these issues the law is not yet clear, and new issues continue to arise in abundance, particularly with relation to software on the Internet. Although these examples do not suggest that copyright protection for software is inherently wrong, they do illustrate that fitting software into previously existing copyright doctrines has been akin to fitting a square peg in a round hole.126 At a minimum, these problems illustrate that copyrighted software works share features that distinguish them from other classes of copyrighted works.

2. Software Is A Functional Work And Copyright Provides It Market Power & Patent-like Protection

Computer software embodies a group of works that is distinct from other copyrighted works.127 One such distinction arises because software is a functional work. For other forms of copyrighted works such as books or paintings, perceiving the expression is the goal of the work. This is true even for a work such as a movie or a sound recording in which the item embodying the work (a film, a phonorecord, a compact disk, etc.) requires a machine to allow the user to perceive the work.

On the other hand, software's inherent value lies in its ability to perform a function or task when its instructions are executed on a machine.128 The user almost always lacks interest in the particular expression that the software engineer used to accomplish the result (e.g., while another software engineer might comment on the aesthetic beauty of how Sally used a tail-recursive subroutine rather than a looping construct or how Bob structures his object classes to promote reusability of code, most users merely care whether the program accomplishes its task in a minimum amount of time without crashing the computer). While a computer screen is often needed to perceive the code for a computer program, only the very rare user acquires software out of a desire to see the expression that is present in the lines of software code. Thus, while users of other copyrighted works are interested in perceiving the expression, a software user is interested in the results that are accomplished when this expression is executed and performs a function.

The extension of copyright protection to a functional work is a factor that grants greater power to a software copyright owner than to other copyright owners, and necessitates an adjustment of the scope of the rights granted to these copyright owners so that an appropriate balance between the owner and the public is maintained. Copyright protection for software can be used to withhold access to the ideas underlying the expression,129 and can also prevent others from building compatible products thereby leveraging existing technology.130

In a network industry such as software, market conditions often dictate the need for a single standard, thus effectively destroying competition.131 Even a small difference in initial market power can tip the balance of power between competing products in a new market segment. And once a product has larger market share than competitors, the network externalities that stem from the desire for compatibility with the installed user base act to perpetuate and increase the market share (i.e., software developers will increasingly support the standard to the exclusion of others, independent service organizations will arise to train and repair problems with that software, hardware vendors will preload the software on the computers that they sell, etc.). If an existing dominant software vendor can tie a new entry in the market to its existing dominant products or standards, new entrants in the marketplace have little opportunity to compete.132 To the extent that software copyright is used to enhance and maintain this type of "monopoly", it can confer significant market power133 and make software copyright protection more akin to that provided by patent law (which protects functional ideas).134

This software copyright market power creates a difficulty for those who argue that the protection afforded by copyright will not generally afford sufficient market power to justify a copyright misuse doctrine.135 To the extent that copyrighted software holds such market power, this argument falls short.

3. Copyright Law Can Be Used To Extend Patent-like Protection To Software Ideas And Trade Secrets

In addition to the functionality aspects of software which distinguish it from other copyrighted works, the distribution of software in object code format allows the copyright owner to prevent access to important parts of the work. These excluded parts include not only the original creative expression of the software engineer (which is embodied in the source code), but also the ideas that underlie the expression and to any trade secrets that are contained in the software.136 In fact, computer software may be the only product that simultaneously receives patent, copyright, trademark and trade secret protection for the same aspects of the product. Even if a user is allowed and is able to reverse engineer the object code, the resulting reverse engineered end-product will contain less information than the original creative work for which the copyright owner obtained protection.137 Courts and commentators alike have criticized the use of copyright law to prevent reverse engineering and thereby prevent access to expression and protect both underlying ideas138 and trade secrets.139

C. A "Scope Of The Grant" Copyright Misuse Doctrine Should Be Applied To Computer Software

As discussed above, the scope of the rights granted for intellectual property works is determined by a combination of statutory and judicial doctrines, and is designed to provide a balance between the rights granted to creators and those retained by the public. If a class of copyrighted works shares features that provide an excess of power to the copyright owner, a judicial determination is needed to realign the extent of rights that are granted. As discussed earlier, computer software is such a class of copyrighted works and owners of computer software copyrights thus possess an excess of rights relative to what is appropriate based on public policy. The misuse doctrine is the appropriate vehicle to realign the rights granted because it focuses on the scope of the rights granted and upholds the public policies underlying the intellectual property laws.

Adherents of an antitrust-based approach to copyright misuse rely on arguments that do not fit well within the domain of software copyright misuse. The copyright goal of promoting creative works is different from the antitrust goal of promoting competition, and copyright and antitrust laws achieve their goals through different mechanisms. In addition, defining the relevant markets for an antitrust-based approach to copyrighted software is extremely difficult, thus making the application of antitrust law uncertain. More importantly, the argument that copyrights do not convey sufficient market power to necessitate a misuse doctrine does not apply to software. Moreover, the argument to allow full utilization of the rights granted is preempted if the misuse doctrine is needed to define the appropriate level of rights that should be granted.

Because software is a class of copyrighted works that can provide greater power to the owner than is true of other classes of copyrighted works, the copyright misuse doctrine is a necessary vehicle to correct the imbalance of power between owners of copyrighted software works and the general public. As the Lasercomb court determined, the copyright misuse doctrine should be applied without the need for an antitrust-based approach, and courts should instead rely on the scope of the copyright grant as determined by public policy.

V. The Past Use of other judicial doctrines to limit software copyright in misuse contexts

This paper has argued that software is a unique form of copyrighted work that confers significant market power to the copyright owner. But if the copyright misuse doctrine is truly needed to mitigate the power of software copyright, the question arises as to why the first appellate decision enforcing misuse did not occur until 1990,140 with only a gradual acceptance of the misuse doctrine in the time since. The answer to the question lies both in the recent emergence of the software industry itself and in the past use of other judicial doctrines by courts to address misuse-like situations.

A. Recent Growth Of The Software Industry Creates A Need For Copyright Limitations

The software copyright misuse doctrine's recent emergence can partially be attributed to the relatively recent ascent of the computer software industry itself. While high school students can probably not conceive of a world without computers (including their Sega Saturn(tm) and Nintendo-64(tm) game machines), it is important to realize that the first personal computer was not introduced until the late 1970s,141 less than 20 years ago. Despite a current $150 billion a year packaged software global market that continues to grow at a rate near 15% a year,142 this market was only $30-35 billion a year in 1990143 and merely $18 billion a year in 1985.144 More strikingly, the World Wide Web on the Internet did not even come into existence until the early 1990s. Yet Internet market revenue has grown to over $5 billion a year in 1995, and is projected to reach over $40 billion a year before the turn of the century145 and $200 billion a year by 2006.146 Thus, it is not surprising that a 1990 case, based on actions that occurred beginning in 1983,147 was the first case in which the software copyright misuse doctrine was applied.

B. Employing Fair Use To Prevent Misuse

While relatively young, the software industry is highly competitive. This competitiveness dictates that at least some producers of software will use every tool at their disposal (including intellectual property rights), even if their use is questionable use or a misuse of an intellectual property right. Due to the paucity of precedent for applying the software copyright misuse doctrine, and perhaps limited by the pleadings of the parties, courts have in the past used other judicial doctrines to address instances of copyright misuse. In particular, the subjective and equitable doctrine of copyright fair use148 has proven to be a powerful tool.

The most visible examples of the courts' use of fair use in the area of software arise with respect to reverse engineering149 of software150 in the Sega151 and Atari152 cases. Since the legal analysis of the Federal Circuit in Atari is similar to that of the Ninth Circuit in Sega, only Sega will be discussed here. In Sega, the plaintiff, Sega, was a video game manufacturer which produced both game consoles and game cartridges that ran in the consoles, and the defendant, Accolade, was a game manufacturer who independently produced game cartridges that were designed to be compatible with the Sega consoles.153 While Sega licensed some vendors to produce compatible games, Accolade was not a licensee. Instead, Accolade reverse engineered Sega game cartridges to discover the necessary software interfaces to make their games compatible with the Sega consoles. After discovering the interfaces, Accolade included the required 4-letter initialization code in their own game cartridges so that they could interact with the Sega console.154 Sega attempted to use copyright law to prevent competition from Accolade in two ways-it claimed that Accolade's copying of software that occurred incidental to the reverse engineering process was infringement, and that the use of the 4-letter initialization code in Accolade's game cartridges was infringement.155 The court rejected the proposition that these actions were copyright infringements, and instead found that Accolade's actions were protected by fair use.156

While the case was thus decided under the doctrine of fair use, there is a strong argument that the application of fair use was inappropriate, and that the misuse doctrine would have been more appropriate.157 The first indication that the fair use doctrine was the incorrect vehicle to use is that the traditional fair use analysis did not fit well with the defendant's commercial actions,158 and so the court needed instead to rely heavily on public policy to justify its conclusion.159 The second indication that the fair use analysis was inappropriate was the fact that the court found it necessary to make a broad pronouncement as a matter of law that would affect software copyright owners as a class,160 a result that was seemingly inconsistent with the fact-intensive nature of the traditional fair use analysis.161 But the most telling indication that the fair use doctrine should not have been used was that the court went out of its way to denounce the actions of the plaintiff,162 an unnecessary departure from the focus on the defendant in the normal fair use analysis. Since the court found it necessary to restrict the rights available to software copyright owners as a class and thus adjust the balance of rights, the fair use doctrine was inappropriate. With its focus on public policy and limiting over-reaching by copyright owners, the misuse doctrine instead should have been applied.

Thus, the Sega decision illustrates that when the court focuses on the actions of the plaintiff rather than those of the defendant, appeals to underlying public policy, treats software as a special class of copyrighted works,163 and makes broad pronouncements of law, the court is misapplying the fair use doctrine. Since the court is effectively performing a misuse analysis that limits the rights of an entire class of copyrighted works, direct application of the misuse doctrine would provide more consistent and predicable results.

C. Employing Other Copyright Doctrines To Prevent Misuse

While the judicial doctrine of copyright fair use provides a powerful tool for courts to use in preventing misuse of software copyrights, the courts have other tools at their disposal. In Bateman v. Mnemonics, Inc., a recent case related to reverse engineering of software, the Eleventh Circuit articulated this well:

Whether the [copyright] protection is unavailable because these factors render the expression unoriginal, nonexpressive per 17 U.S.C. 102(b), or whether these factors compel a finding of fair use, copyright estoppel, or misuse, the result is to deny copyright protection to portions of the computer program. Thus, we today join these other circuits in finding that external considerations such as compatibility may negate a finding of infringement.164

An example of another judicial doctrine that has been used to limit the rights of software owners who attempt to leverage their copyrights beyond allowable limits is the copyright preemption doctrine.165 However, copyright misuse principles often provide a superior framework to restrain plaintiffs who abuse their copyright privileges. In ProCD, Inc. v. Zeidenberg,166 a software manufacturer attempted to contractually restrict the uses of uncopyrightable telephone directory information167 which could be downloaded from its product. While the Seventh Circuit court found that the shrink-wrap contract contained inside the product box was not preempted (in a ruling that was led by Judge Easterbrook and which relied heavily on a Chicago School economic analysis),168 the district court judge had conversely found preemption due to the invalid extension of copyright power.169 If the plaintiff truly was trying to extend the power of its copyright beyond the bounds allowed by public policy, the misuse doctrine provides a better framework with which to analyze these activities.

While the Seventh Circuit did not believe that the actions of the plaintiff in the ProCD case were an impermissible extension of copyright law, other courts have found differently under analogous situations. In the Vault case,170 the Fifth Circuit found several provisions of the Louisiana Software License Enforcement Act (SLEA)171 to be subject to preemption by the Copyright Act because the SLEA authorized contractual duplication or extension of rights granted under copyright law.172 The SLEA allowed producers of software to impose various contractual terms upon software purchasers, provided that the terms were included with the software in a license agreement that comported with SLEA. The district court identified four areas in which the SLEA could be preempted by the copyright law, both constitutionally and under section 301.173 It found that SLEA provided rights which exceeded those of the copyright law by allowing for prohibition of copying for any purpose, creating a bar against copying that was unlimited in duration, and allowing protection for any computer program, regardless of its originality.174 In addition, the district court found that SLEA provided equivalent rights to those of the copyright law by allowing the prohibition of the creation of derivative works.175 The Fifth Circuit, while affirming the reasons set forth by the district court, added yet another reason for constitutional preemption of SLEA-that the prohibition against reverse engineering does not allow the owner of a computer program to exercise the 117 right to adapt the program as an essential step in its utilization, and thus gives rights to the owner of the copyright that exceeds those granted under the copyright law.176 Thus, the circuit courts are split on what constitutes an impermissible extension of copyright under a preemption analysis. Since the misuse analysis focuses more directly on public policy, it is more appropriate than preemption to address such extensions of copyright law.

In addition to the use of the preemption doctrine, some courts have followed the Bateman suggestion to find works "nonexpressive per 17 U.S.C. 102(b)" when allowing copyright protection would unduly extend the reach of copyright law.177 For example, the West Publishing Company has long attempted to prevent competitors from launching similar software products by claiming copyright protection on the "star pagination" used in its compilations of uncopyrightable judicial opinions. In two recent cases, the courts have found the West star pagination to be undeserving of copyright protection,178 again focusing on the inappropriate actions of the alleged copyright owner.

Thus, in only a short period of time, owners of software copyrights have used various techniques to extend their rights. Courts have addressed these actions using a variety of copyright doctrine analyses, including fair use, copyright preemption and section 102(b) "non-expressiveness." The erection of such disparate barriers to prevent abuses of software copyrights demonstrates the existence of a significant problem, and furthermore, the use of a single uniform approach would provide greater consistency and predictability. Since copyright misuse focuses on the actions of copyright owners and relies on underlying public policy considerations, a uniform approach to future examples of misuse under the penumbra of the software copyright misuse doctrine is a solution to the problem.

VI. Applying Software Copyright Misuse To Traditional Misuse Contexts

Having argued that software copyright misuse is a necessary and useful doctrine, the question next arises as to the proper application of the doctrine. Although the proper scope of the doctrine will ultimately be decided through a series of cases that develop underlying principles, it is appropriate to consider possible areas of application. One starting point arises from the established case law in the patent misuse area. As described previously, the two paradigmatic patent misuse cases involve tying arrangements (where the patentee requires the purchase of unpatented goods or services along with a patented product or process), and non-compete clauses (that prevent a patent licensee from producing or selling competing goods).179

A. Software Copyright Misuse And Tying Arrangements

Due to the complicated nature of software tying arrangements, a general resolution of the issues involved is beyond the scope of this paper. Although it was fairly easy to find an illegal tying arrangement between unpatented salt tablets and the patented salt-depositing machinery in Morton Salt (because it is easy to treat the salt tablets and the salt-depositing machinery as distinct products), it is more difficult to determine whether a computer system sold with the hardware and operating system bundled together is such a tying arrangement. In many past business situations, one company produced both the computer hardware and the only operating system that functioned on the hardware, thus requiring both parts to achieve a minimally useful machine. In the software industry today, it is common practice to bundle productivity applications together into "office" suites. However, this is bundling only because we conceptually conceive of the different parts of the suite as independent applications-as we increasingly move toward a computing world where data is king, the notion of distinct applications may disappear.180

These "tying" situations may be acceptable (or not tying at all), because it is easy to treat these tied products as a single product. At a minimum, a strong functional relationship between the tied products provides a necessity-based justification for the tying. However, in other types of software tying situations, the software copyright owner attempts to forbid incidental copying of software to prevent competition in goods or services that are unrelated to the software. These situations will be addressed below in relation to restrictions on competition.

B. Software Copyright Misuse And Restrictions On Competition

1. Use Of Non-Competition Agreements In Software Licenses

Not surprisingly, the first software copyright misuse case arose due to an attempted restraint on competition. As described above, the plaintiff in the Lasercomb case had included a provision in its standard form contract that not only limited copying of the licensed software, but also prevented licensees from independently developing a competing software program for 100 years.181 The Fourth Circuit held that this restriction constituted copyright misuse. Other courts have found similarly,182 and little argument has been advanced that an alternative policy would be desirable.

2. Enjoining Incidental RAM Copies To Prevent Development Of Products

Since the 1993 decision in MAI (which held that temporary copies of software made in RAM while the computer is turned on and running are sufficiently fixed to be infringing),183 companies have tried to use copyright law to control various third party activities involving temporary copies, including the development of competing products and competing services. Since copyright law is designed to protect only the creative expression embodied in the source code, the functional effects of the object code should not be included within that protection unless they are integrally tied to the creative expression. Arguably, enjoying the functional results of an executing RAM copy of software is sufficiently related to the copy itself that the creation of the RAM copy is copyright infringement. However, when the creation of the RAM copy is incidental to the activity being controlled, a strong argument exists that it would be misuse to use the RAM copy to restrict the activity. For most computers, merely pressing the power switch to turn it on causes the operating system software to be loaded into RAM, thus creating an infringing copy.184 Furthermore, in order to use a computer program it must be loaded into RAM,185 whether the stored copy of the program normally resides on the computer's hard drive, on a floppy disk, on a CD-ROM, or on another machine on a computer network. Even if these RAM copies are infringing under a MAI-type rationale, the question remains whether the activities of the copyright owner constitute copyright misuse.

In the 1996 case of DSC Communications Corp. v. DGI Technologies, Inc.,186 the Fifth Circuit found one such example likely to constitute copyright misuse. The plaintiff, DSC, manufactured telephone switching systems which were a specialized form of computer, including an operating system, at least one central computer processor and removable microprocessor cards.187 When the cards are inserted into the system, they must "boot up", and in doing so they download a copy of the operating system into their RAM.188 The defendant, DGI, was developing compatible microprocessor cards to run in DSC systems. In order to test its product, DGI needed to download the operating system into the card in the same manner that would occur during normal operation of the system.189 In the context of a preliminary injunction requested by DSC to prevent DGI from making these copies, the Fifth Circuit found that DSC was unlikely to succeed on the merits because it was likely that their activities constituted copyright misuse.190

The principles advocated in this paper support the Fifth Circuit analysis. DSC attempted to extend the rights granted to one work to prevent the creation of other independent works. Moreover, the removable add-in cards were unrelated to the creative aspects of the operating system software. A contrary holding would thwart copyright law's underlying goal of fostering creation and innovation. The development of the add-in card by DGI required substantial creativity and innovation, and its use of the operating system for compatibility purposes was incidental. Granting a monopoly over products such as the add-in cards was not the type of incentive envisioned by the framers of the copyright laws for rewarding the creativity expressed in creating a software literary work.

3. Use Of Incidental RAM Copies To Prevent Development Of Services

Although the Fifth Circuit held that extending copyright in functional RAM copies to prevent the creation of products unrelated to the copyright was misuse, several courts have found otherwise when services unrelated to the copyright were obstructed. However, this distinction between products and services is not warranted,191 particularly as our economy becomes increasingly information-based and service-based rather than product-based.

The MAI case illustrates a blatant example of extending copyright in temporary RAM copies to unrelated services. In MAI, the Ninth Circuit found that the defendant had committed copyright infringement by merely turning on a MAI-produced computer at the direction of a MAI customer who had purchased the computer. 192 The plaintiff, MAI, manufactured and sold computers that included operating system software and other diagnostic software developed by MAI.193 The defendant, Peak, provided servicing of computers, including routine maintenance and emergency repairs, that included turning on the computer and viewing the system error log (which is part of the operating system).194 Turning on the computer was found to be sufficient to constitute copyright infringement, since a copy of the MAI operating system was made in RAM as part of the boot process of the computer, even if the diagnostic software was not used by Peak.195

A consideration of slight variations in the fact pattern illustrates the absurdity of this holding. Under the scenario posited by the court, if the Peak repairman had arrived at the customer's site and the computer was fortuitously on, no copyright infringement would have occurred.196 Alternately, infringement would have been avoided if the repairman had stood over the shoulder of an authorized user while the user pressed the on button. A legal analysis that determines whether a computer service can be provided or whether copyright infringement has occurred based on who presses the computer "on" button seems ridiculous. In a paraphrase of the DSC court, MAI is attempting to use its copyright to obtain a patent-like monopoly over unpatented maintenance services, and the defense of copyright misuse forbids the use of the copyright to secure an exclusive right or limited monopoly not granted by the Copyright Office.

The absurdity of the MAI result exists even if the copyrighted software was more closely related to the service of performing maintenance on a computer. If the defendant had made a RAM copy of diagnostic software rather than operating system software, then at least there would be some logical relationship between the copy being made and the service being performed. However, while this modification to the MAI scenario would at least provide a tenuous relationship between the activities of producing the software and providing the service, the result is no less absurd. The on-site repairman will now have to stand over the shoulder of an authorized user until the user starts up the diagnostic software (rather than the operating system software), thus creating an authorized RAM copy which the repairman can use.

The problem with the MAI ruling is that it discriminates against different classes of users. A customer should be authorized to use its software in a normal fashion and a service provider should be entitled to act as an agent for the customer and take the same actions.197 Instead in MAI, an inane distinction occurs based on the status of the repairman as an employee (if the company is big enough to afford a full-time system administrator) or an independent contractor. If a software producer has provided diagnostic software to a customer for the customer's use, later attempts to enjoin incidental copying in order to create a monopoly in servicing the hardware is an undeserved extension of the copyright.

Despite the obvious attempt on the part of MAI to extend its exclusive rights to an unrelated service, an important distinction can be made between computer servicing cases198 and the DSC case. The competition being restrained in the computer servicing cases involves the provision of a service rather than the production of a good. Arguably, however, the distinction between goods and services should be irrelevant in today's society. As our society becomes increasingly specialized and the portion of our economy devoted to providing services grows, it is difficult to argue that an innovative service is less valuable from a public policy perspective than an innovative good. For example, the innovative service from Federal Express to provide overnight delivery of packages at economical prices has revolutionized how business is done. In addition, there is no principled reason why the creation of services cannot be covered under the constitutional authorization to promote "writings."199

Thus, attempts to use copyright protection to stifle competition should be met with judicial disapproval, regardless of whether the competition comes in the form of goods or services. If the underlying copy of the software from which the RAM copy was created is itself a legitimate copy, then the incidental copy created in RAM should not be treated as infringement for the purposes of restricting competition. Attempts to alter this result through license agreements with customers should be treated as an undue restraint that constitutes misuse of the rights granted by copyright law.

VII. Applying Software Copyright Misuse To current Software practices

While considering the paradigmatic patent misuse cases is an appropriate starting point in defining the necessary scope of software copyright misuse, it is also necessary to consider those aspects of software which make it unique as a copyrighted work and to impose some limits where the unique nature of software grants excessive power to software copyright owners. The intent of this section is not to provide an exhaustive list or an in-depth analysis, but rather to suggest areas that deserve special judicial attention.

A. The Licensing Rather Than Sale Of Software

The widespread practice of "licensing" software to users rather than selling it is troubling in the context of mass-market software products.200 By definition, such "licensing" occurs because software copyright owners believe that it grants them greater rights than they receive under copyright law. If this were not true, companies would not invest the time and energy to develop and propagate the licenses. However, copyright misuse restrictions on this practice would collide head-on with one of the basic premises of our capitalistic economy-that contract law provides the ability to achieve the most efficient ordering of assets and rights through private agreements in the free marketplace.201

Although it is unclear whether mass-market202 form licenses are generally enforceable,203 software companies are undoubtedly attempting to alter the careful balance of rights granted between copyright owners and the public. In particular, by characterizing purchasers of off-the-shelf (or off-the-net) software as licensees rather than owners, the copyright owners are explicitly attempting to alter the rights granted to "owners" of software products under the "first-sale" provisions of section 109204 and under the computer-program specific rights granted to users in section 117.205 Allowing this practice would thwart the careful balance of rights struck by Congress.

B. License Restrictions On Reverse Engineering

While the licensing of software can generally lead to abuses of power and the alteration of the basic rights granted under the copyright law, mass-market contractual attempts to prevent reverse engineering are even more bothersome. As discussed earlier, courts have generally allowed reverse engineering in the absence of contractual restrictions.206 Therefore, the question becomes whether this balance of rights can be altered contractually.

Although the courts have not yet decided this issue, they have discussed the related question of the patent-like protection that copyright law can provide to ideas and trade secrets when software is distributed in object code format.207 This question involves competing principles. On one hand, a software copyright owner should not be able to receive the benefits of copyright law without having to disclose the underlying ideas and expressions in the source code. On the other hand, trade secrets have commercial value and businesses have a legitimate desire to protect them.

In Kewanee, the Supreme Court held that trade secret law was not preempted by patent law208 and emphasized the important role that reverse engineering played for trade secrets embodied in products.209 The court in Bonito Boats recently returned to this theme:

[I]n essence, the Florida law prohibits the entire public from engaging in a form of reverse engineering of a product in the public domain. This is clearly one of the rights vested in the federal patent holder, but has never been a part of state protection under the law of unfair competition or trade secrets ¼ . The protections of state trade secret law are most effective at the developmental stage, before a product has been marketed and the threat of reverse engineering becomes real. During this period, patentability will often be an uncertain prospect, and to a certain extent, the protection offered by trade secret law may 'dovetail' with the incentives created by the federal patent monopoly.210

Despite the doubts of the Court in Kewanee that owners of patentable inventions will rely on trade secret law instead of patent law, that is exactly what happened in the early days of software patents when it was unclear if a patentee had to disclose actual source code in a patent application in order to enable the invention and describe its best mode. In this uncertain situation, many software owners recognized that a combination of trade secret law and copyright law provided stronger protection than patent law. The combination of trade secret law and copyright law allowed the software owners to use their software without fear of reverse engineering. Presumably, the widespread use of copyright law to enhance the power of trade secret law could serve to eviscerate the use of software patents, and if so, should be preempted by the patent laws.

Like the protection provided by copyright law, contracts may be used to protect ideas and trade secrets embodied in software. Restrictions in mass-market licenses can routinely prevent reverse engineering by the general public, and thus can greatly enhance the power of trade secret law. If so, the balance between trade secrets and patents will have shifted sufficiently that these clauses will be an "obstacle to the accomplishment and execution of the full purposes and objectives of Congress"211 and should be struck down as misuse.212

As a final consideration, in an era of increasing international harmonization of intellectual property law, it is important to note that the European Union has rejected the use of contract law to abrogate various legitimate rights of software users.213 Software copyright owners cannot contractually restrict the right to make back-up copies,214 the right to attempt to determine the ideas underlying the program,215 or the right to reverse engineer software in order to create competing or compatible software.216 It is instructive that the Europeans, who normally grant greater rights to authors than does the United States, do not allow contractual restrictions on reverse engineering when it is necessary for interoperability.

VIII. Conclusion

The United States' federal patent and copyright laws are based on a common utilitarian public policy of benefiting society through the encouragement of the creation, discovery and dissemination of novel ideas and creative expression. The grant of property rights in a limited monopoly to creators achieves these goals, and the underlying public policy necessitates the careful balancing of the rights granted to the creator with those retained by the general public. Enforcement of the intellectual property misuse doctrine prevents reordering of these rights, whether through state law or by individual owners of intellectual property.

While the misuse doctrine can be used in a general manner as described above, it can also be used to redefine the rights granted to creators when a class of protected works provides excessive power to the owners of such works. The misuse doctrine is the appropriate vehicle to be used in such a situation because it focuses on upholding the public policies underlying the intellectual property laws. However, a judicial redefinition of rights is justified only if it can avoid creating undue uncertainty on the part of intellectual property right owners as to the rights that they possess.

Computer software is a class of works that is unique among copyrighted works. Because copyright law provides excessive power to software copyright owners, the copyright misuse doctrine is needed to retain the balance of rights necessary to accomplishing the public policy underlying the intellectual property laws. The excessive power received by software copyright owners arises as a combination of the functional aspects of computer software, the ability to protect underlying ideas and expression when distributing software in object code format, the ability to protect trade secrets in a matter akin to patent law when distributing software in object code format, and unique aspects of the software industry which provide network externalities to some producers of software. For related reasons, a software copyright misuse doctrine separate from antitrust law is needed.

Due to the recent emergence of the copyright misuse doctrine, courts have applied other doctrines to address situations in which software copyright owners have attempted to unduly extend their deserved monopolies. Whether through fair use, intellectual property preemption, non-copyrightability or findings of misuse, the abuses by software copyright owners should be equitably limited. While the courts can generally arrive at the fair result through other doctrines, application of these doctrines creates suspect legal reasoning that does not always extend well to novel problems that the courts must face. Thus, a uniform software copyright misuse doctrine will provide a more appropriate vehicle for defining the scope of rights accorded to software copyright owners.

In developing a law of software copyright misuse, courts should be especially wary of explicit attempts to alter the statutorily-defined balance of rights granted, and should disallow attempts to restrain competition. In particular, the use of non-competition agreements and copyright restrictions on incidental RAM copies to prevent the development of unrelated goods or services should be unenforceable due to their inherent misuse. Similarly, the use of contracts to significantly limit the rights retained by purchasers of mass-market software warrants judicial limitations. While software is an incredibly useful product that deserves some protection, the rights granted must be limited so that the public retains its fair allocation.