KEPNER-TREGOE, INC. v. LEADERSHIP SOFTWARE, INC.

WIENER, Circuit Judge:

Plaintiff-appellee Kepner-Tregoe, Inc. (K-T) won a permanent injunction and damages against Defendant-Appellant Leadership Software, Inc. (LSI) in a suit complaining that LSI's computer program infringed K- T's exclusive license to copyrighted management training materials. LSI appeals, raising issues touching on both copyright law and the Federal Rules of Evidence. We modify in part the ruling of the district court and, as modified, affirm.

I

FACTS AND PROCEEDINGS

K-T sued LSI, alleging that LSI's computer program called "Managing Participation in Organization" (MPO) infringed copyrighted materials (Licensed Materials) that Professor Vroom, a 50% owner of LSI, had exclusively licensed to K-T in a 1972 agreement (the Agreement). Prior to trial, K-T secured a temporary restraining order and a preliminary injunction prohibiting LSI from manufacturing, distributing, or selling the MPO computer program. In a bench trial, the district court found in favor of K-T and entered a permanent injunction against LSI's distribution of the MPO program. The court also awarded K-T damages in the amount of $46,000.

Significantly, the court enjoined distribution of (1) the original MPO program, (2) a modified MPO program that LSI produced in an effort to remove just the infringing language, and—seemingly—(3) all future modifications and revisions of MPO. The court also awarded K-T attorneys fees, costs, and interest. LSI timely appealed, asserting that the court erred with respect to both its application of copyright law and its evidentiary rulings.

K-T, a Maryland corporation with its principal place of business in Princeton, conducts management training seminars. LSI, a Texas corporation with its principal place of business in the Houston area, sells the MPO computer program, which was developed by Professors Vroom and Jago, the equal owners of LSI. Jago was a named defendant in the district court proceedings, but K-T could not compel Vroom's attendance at trial in Houston because he resides in Connecticut. Neither Vroom nor Jago is a party to this appeal.

In 1972, K-T signed the Agreement, by which it acquired an exclusive international license to copyrighted materials entitled Leadership and Decision Making Cases and Manuals for use in Leadership Training. In consideration for this exclusive license K-T paid the authors—Professor Vroom and his former colleague, Professor Yetton—more than $400,000 in royalties over several years. K-T ultimately bought out the license for an additional, one-time payment of $100,000. The Licensed Materials comprise a management decision-making model called the Vroom-Yetton Model (V-Y Model) and include descriptions of management problems, explanations of management decision-making styles, flow charts presenting decision-making possibilities, and rules designed to help managers make the best decisions.

Among the copyrighted materials that were exclusively licensed to K-T are eight "questions" or "problem attributes" and five "definitions" or "processes." These questions and processes are the core of the V-Y Model. The questions prompt managers to evaluate their decision-making landscape. For example, one question asks if the manager has adequate information to make a good decision himself. The five definitions or processes describe various decision-making approaches, from autocratic (the manager alone) to more democratic (decisions requiring consensus). Presumably, the V-Y Model rules tell a manager which decision-making process to use in a given context, based on the manager's responses to the eight questions.

In 1983, Professor Jago—working closely with Vroom—created the MPO computer program. The MPO program retains the same eight questions and five processes that comprise the V-Y Model, but adds four additional questions. The MPO program also evidently processes the information generated by the questions differently, and presumably tells managers which decision-making process to adopt, without forcing them to consult flow charts and elaborate decision-making rules. The organization and language of the five processes and eight questions is virtually identical in K-T's Licensed Materials and the MPO program.

Vroom and Jago kept key K-T personnel apprised of their work on the MPO program. In 1987, Vroom and Jago incorporated LSI to market the MPO program. Shortly thereafter, K-T initiated negotiations to reach a licensing agreement covering the program; these negotiations failed.1 In 1990, K-T learned that LSI had been selling copies of the MPO program. K-T commenced this lawsuit almost immediately, alleging that MPO infringed the copyright covering its Licensed Materials. Although K-T became entitled to copyright protection through an exclusive licensing agreement, this appeal is not about licensing arrangements. Rather, this is a straightforward copyright case.

II

ANALYSIS

On appeal LSI contends that the district court erred both in its application of copyright law and in its evidentiary rulings. We consider each topic in turn.

A. Copyright Issues

Concluding that LSI's MPO program infringed K-T's Licensed Materials, the district court awarded K-T damages and enjoined LSI from "further copying, producing, distributing, and/or selling the MPO program." The court also concluded that a modified version of the MPO program infringed K-T's Licensed Materials. This modified MPO program resulted from LSI's efforts to remove all infringing language from the original MPO program. The injunction covers this modified program too. Finally, the court enjoined all future modifications and improvements of the MPO program. For clarity, these programs—(1) the MPO program, (2) the modified MPO program, and (3) all future modifications of the MPO program, are discussed separately.

1. The MPO Program

The district court concluded that LSI's MPO program infringed K-T's Licensed Materials. To reach that conclusion, the court had to find that (1) K-T owned a valid copyright over the Licensed Materials, (2) LSI copied portions of the Licensed Materials when it made the MPO program, and (3) among the portions copied were substantial protectable elements of the Licensed Materials.2 LSI does not contest that K-T's Licensed Materials are covered by a valid copyright. Indeed, Vroom—a 50% partner in LSI—applied to register the V-Y Model materials that were subsequently licensed to K-T, evidently reflecting his belief that those materials were the proper subject of copyright protection.3 Rather, LSI insists that (1) K-T failed to prove that LSI had actually copied K-T's licensed materials, and (2) the court erred in extending copyright protection to inherently unprotectable elements of K-T's materials.

a. Proof of Actual Copying

LSI argues that the district court erred in finding that it had actually copied K-T's Licensed Materials, rather than copying other materials—possessed by third parties—that contained the same information. We find this argument to be without merit.

As direct evidence of copying is uncommon, plaintiffs generally demonstrate copyright infringement indirectly or inferentially by proving that (1) defendants had access to the copyrighted works, and (2) there is a substantial similarity between infringed and infringing works.4 These copyright issues of access and substantial similarity are findings of fact and are consequently reviewed under the clearly erroneous standard.5 In this case, the district court found that both Vroom and Jago had access to the Licensed Materials. As Vroom helped develop and write those materials, and as Jago relied on them in developing the MPO program, that finding is not clearly erroneous. Indeed, it is factually correct.

LSI, however, insists that K-T never demonstrated that LSI literally copied the specific materials that were licensed to K-T. But LSI's dogged insistence is nonsensical. Even if LSI did lift the offending expression from third party sources, its reproduction of that expression for commercial purposes may be infringing. Language copied from those third party sources was itself copied or derived from K-T's Licensed Materials, and its legality depends on copyright law. In other words, even if LSI copied a copy of K-T's Licensed Materials, such copying may still constitute infringement. Copying a copy of copyrighted materials is a cognizable contravention of the Copyright Code.6

LSI does not dispute that there is substantial similarity between the MPO program and K-T's Licensed Materials. Indeed, LSI admits that the MPO program incorporates the same eight questions and five processes that the district court characterized as the "heart and soul" of the V-Y Model, which was licensed to K-T. Thus, the district court's finding that MPO is substantially similar to K-T's Licensed Materials is not clearly erroneous. In summary, neither the district court's finding that LSI had access to K-T's Licensed Materials, nor its finding that MPO program was substantially similar to those materials is clearly erroneous. Consequently, the district court's finding that LSI copied K-T's Licensed Materials is not clearly erroneous.

b. Protectable and Unprotectable Elements of K-T's Copyrighted Materials

LSI argues nonetheless that the portions of K-T's licensed materials that it allegedly copied are but unprotectable ideas or facts, and that the district court therefore erred in holding that LSI infringed K-T's copyright. LSI is correct that the mere fact that K-T's Licensed Materials are copyrighted does not mean that all aspects of those materials are automatically protected.7 Specifically, LSI rightly observes that copyright law protects tangible, original expressions of ideas, not ideas themselves.8 "[N]o author may copyright facts or ideas. The copyright is limited to those aspects of the work— termed `expression'—that display the stamp of the author's originality."9 Thus, if we conclude that LSI only copied unprotectable elements of K-T's materials, we must reverse the district court's judgment.

Unfortunately, the line between idea and expression is hard to draw. Additionally, when an idea can be expressed in very few ways, copyright law does not protect that expression, because doing so would confer a de facto monopoly over the idea. In such cases idea and expression are said to be merged.10 To determine the scope of copyright protection in a close case, a court may have to filter out ideas, processes, facts, idea/expression mergers, and other unprotectable elements of plaintiff's copyrighted materials to ascertain whether the defendant infringed protectable elements of those materials. In this case, however, we disagree with LSI's contention that such a filtration process leads to the conclusion that the district court erred in finding that LSI's MPO program infringed K-T's Licensed Materials.

Although there is no evidence that the district court undertook a rigorous "abstraction-filtration-comparison" analysis of the sort approved by courts for sophisticated treatment of copyright cases,11 such an analysis was not absolutely necessary here. The district court carefully juxtaposed selections from K-T's Licensed Materials with selections from the MPO program, thereby demonstrating a damning similarity—nay identity—of organization and language. This comparison of literal language or expression provided strong evidence for the court's finding that the MPO program infringed K-T's Licensed Materials.

Seizing upon the court's statement that the questions and processes of the Vroom-Yetton model are its "heart and soul," LSI argues that these elements are "inherent in the leadership management theory ... of the Vroom models," implying that questions and processes that comprise the V-Y Model are unprotectable ideas. LSI contends that there is no protectable expression remaining in K-T's licensed materials, once all unprotectable elements are filtered out. But this is absurd.

Each question and process in the V-Y Model is presented in a paragraph of text. There are countless ways of expressing the content of each paragraph,12 so there was no need for the MPO screen text to copy exactly the language of K-T's materials. Even if each of the eight questions and five processes conveys unprotectable ideas, the specific words, phrases, and sentences selected to convey those ideas are protectable expression under any reasonable abstraction analysis. As LSI's MPO program copied those words, phrases, and sentences verbatim, we conclude that—far from being clearly erroneous —the district court's finding that the MPO program infringed K-T's Licensed Materials was correct and must be affirmed.13

2. The Modified MPO Program

The district court also concluded that LSI's modified MPO program infringed K-T's Licensed Materials. As noted above, the modified MPO program was the end result of LSI's surgical efforts to remove only the infringing language from the original MPO program. As this effort was partially successful, we cannot affirm the district court's conclusion that the modified MPO program infringes K-T's copyright merely by making a cursory comparison of the modified MPO program's language with that of K-T's materials.

As with the original MPO program, there is no doubt that LSI copied K-T's materials in creating its modified MPO program. As noted above, evidence of LSI access to K-T's materials and the substantial similarity of the original MPO program over-whelmingly suggest copying. The same is true for the modified MPO program, which is just a post factum rearrangement of the original MPO program, itself a wholesale plagiarism of the definitions and processes that were licensed to K-T. Our finding of copying does not, however, lead automatically to the conclusion of infringement: the question remains whether LSI's modified MPO program copied any protectable elements of K-T's Licensed Materials.14 We conclude that it did.

The main purpose or function of K-T's Licensed Materials is to teach managers how to analyze their own decision making, and how to make the best decision in each decision-making context. Clearly this basic idea of a management training program is unprotectable.15 Likewise unprotectable is the more specific idea of training managers by asking them a series of questions about their decision-making landscape, and then— based on their answers—suggesting a preferred decision-making process.

At the other end of the abstraction spectrum, the specific words, phrases, and sentences used to formulate the questions and processes clearly constitute protected expression.16 But the intermediate levels of abstraction, consisting of such factors as the structure, sequence, and organization of a copyrighted work, are more problematic: courts' judgments about the protectability vel non of such elements are inevitably ad hoc and fact specific.17 In this case, we conclude that the modified MPO program copied substantial protectable elements of K-T's Licensed Materials and thus infringed K-T's copyright.

Like the district court, we are unimpressed by LSI's shallow efforts to remove infringing language from the MPO program. Although it is true that the modified MPO program does not identically trace the language of the definitions and processes contained in K-T's Licensed Materials (as did the original MPO program), the modified MPO program's language is still substantially similar to that of K-T's materials. True, "you solve the problem yourself" has been replaced with "you reach a solution alone." And "you consult one-to-one with those that report to you" replaces "you share the problem with relevant subordinates." But such modifications do not completely dispel the similarity of expression shared by infringed and infringing materials.

LSI argues that the modified MPO program is merely substantially similar to unprotectable conceptual elements of K-T's materials, that the modified MPO program communicates the same concepts as K-T's Licensed Materials, but with different expression. We disagree.

As an analogy, consider the familiar quote from Romeo and Juliet: "O Romeo, Romeo! wherefore art thou Romeo?"18 Reformulating this quote in a manner analogous to LSI's modification of K-T's copyrighted language, we might write: "Ah Romeo, Romeo! Why did you have to be born Romeo?" Are these two quotes alike only in conceptual substance? Obviously not! Yes, they both express the same concept: why did Juliet have to fall in love with Romeo, scion of Montague—her family's bitterest foe? But they are also quite alike in expression. They are not identical, but they are alike. Both sentences embrace the dramatic repetition of Romeo's name. Both sentences are phrased as questions. Both have the quality of a sigh: the gasping resignation of a woman marvelling at the fateful irony of life. Although the two sentences are not identical, they are manifestly similar in expression, as well as in conceptual content.

The language of LSI's modified MPO program is likewise similar to that of K-T's materials: their paragraphs are about the same size, their phrases are similar, their ideas are presented in the same order; in short, parts of the modified MPO program are but a transparent, syntactic rearrangement of portions of K-T's copyrighted materials. While no longer identical to those materials, the modified MPO program still bears many telltale signs of its origins. It is still a copy—still a child of infringement.

Additionally, we conclude that the modified MPO program infringes upon elements of K-T's materials, which—although existing at a higher level of organizational abstraction—are nonetheless protectable under copyright law. As we have noted, generalizing about the degree of copyright protection owed to intermediate levels of abstraction, such as the structure, sequence, and organization of copyrighted works, is difficult. A fairly broad consensus has emerged, however, that such non-literal elements of computer programs and other copyrightable works may be protected.19 Generally speaking, we join that consensus.20

LSI argues that the V-Y Model, which was described in the materials licensed to K-T, amounts to a law of nature like Newton's Law of Gravitation, the constant W (for discussing the geometry of circles and spheres) or Einstein's E = MC2. Specifically, LSI contends that the eight definitions and five processes that comprise the V-Y Model are fundamental, ineluctable aspects of a managerial relationship—that intelligent discussion of managerial decision-making is impossible without specific reference to these universal definitions and processes. Although there is a kernel of truth in this assertion, it is a small kernel: in the main, we disagree with LSI's self-aggrandizing characterization of the V-Y Model.

Some unprotectable fundamental concepts are undoubtedly buried in the definitions and processes of the V-Y Model. For example, the idea that a manager can make a decision without consulting his subordinates is unprotectable: it is simply one of the relational possibilities that exist between managers and subordinates. As noted above, however, the specific ways that the questions and processes are formulated—the exact words, phrases, and sentences used to describe decision-making processes or questions—are protectable expression. And other, more abstract organizing principles of the V-Y Model are protectable as well.

In dissecting the problems of management decision-making into five processes and eight questions, Vroom and Yetton unquestionably originated a useful model of managerial decision-making. Yet, they obviously did not discover the single, unique, unavoidable description of human managerial relations. In creating the MPO program, for example, Jago added four more questions to the eight that were part of the V-Y Model. Such an expansion of the original theory indicates that the first eight questions did not exhaustively and uniquely portray human management relations.

Close analysis of the V-Y processes and questions reveals that the ideas they encapsulate can be packaged in different ways. For example, one process (designated AII) instructs a manager to gather information from subordinates, then make the relevant management decision himself. The process states that the manager "may or may not tell subordinates" the nature of the problem in getting information from them. This makes two separate logical possibilities, so this process could—in a different model—be broken into two different processes. In short, not only does the V-Y Model fail to describe all conceivable features of the management decision-making landscape, it also fails to organize and package its managerial truisms in a single, unique, ineluctable way: there are many ways of organizing those same insights.

This point is driven home by examining the way different processes and questions are designated in the V-Y Model. The five processes of that model are designated respectively AI, AII, CI, CII, and GII. How did LSI designate the five processes of its modified—and putatively non-infringing— MPO program? Not surprisingly, AI, AII, CI, CII, and GII. Are these designations fundamental constants supplied by nature, like W for a circle or Planck's constant in quantum physics? Clearly not, they are arbitrarily selected characters: the V-T Model would work just as well if its processes were designated a, b, c, d, and e! These features of the V-Y Model are thus original, protectable expression, not fundamental constants of nature; and LSI's copying of these features —and other related features—is thus technically infringing.21

LSI obviously uses V-Y Model designations in its modified MPO program, and does so intentionally. The V-Y Model has been wildly successful, and the MPO program could benefit from that success by incorporating recognizable, original expression from the V-Y Model, whether that expression is the verbatim formulation of questions and processes, the organization of the model, or the arbitrary designations of the model's constituent parts. But the MPO program is not supposed to benefit from such incorporation. It was precisely the right to benefit from such copying that Vroom and Yetton licensed exclusively to K-T for half a million dollars.

LSI attempts to obscure this point by noting that the V-Y Model is reproduced, discussed, and described all over the United States, suggesting that the V-Y Model is in the public domain and thus unprotectable. But protected expression does not lose its protection simply because it is widely disseminated. If the V-Y Model is widely discussed, described, and reproduced, it is presumably with the permission of the copyright holder. Alternatively, it is because such discussion, description, and reproduction constitutes fair use.22 "[F]air use of a copyrighted work ... for purposes such as criticism, comment, new reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright."23 Conspicuously absent from the list of fair uses is use for commercial purposes, which is exactly the sort of use LSI wants us to approve.

When Vroom and Yetton sold K-T an exclusive license to copyrighted materials, which included the V-Y Model, they signed away the right to copy, at least for commercial purposes, protectable elements of the V-Y Model. LSI may not now incorporate substantially similar expression into computer programs for commercial sale: it was precisely the right to make such commercial use of the V-Y Model that Vroom and Yetton sold to K-T for hundreds of thousands of dollars.

In summary, we conclude that the district court did not err in holding that the modified MPO program infringed K-T's Licensed Materials. As LSI was partially successful in removing infringing expression from the MPO program, the judgment that the modified MPO program infringes K-T's copyright is a fairly close one. Nonetheless, we agree with the district court that, although the modified MPO program does not identically trace the language of the definitions and processes delineated in K-T's Licensed Materials, it infringes substantial portions of the protected expression contained in those materials.

3. All Future Modifications of the MPO Program

We are uncertain whether the district court really intended to enjoin all future modifications of the MPO program, and if so, what it meant by such an injunction. The court's language in its Findings of Fact and Conclusions of Law is certainly broad enough to suggest that it enjoined all future modifications and improvements of the MPO program. But the Judgment obscures the court's intentions by enjoining all "modifications [of MPO] ... which are the subject of the Temporary Restraining Order (TRO) issued March 15, 1990," because no MPO modifications were expressly the subject of the TRO. Indeed, no modifications of MPO existed at the time the TRO was entered: no legal action by K-T had yet compelled such modifications. We hold, however, that whatever the court intended in that regard, it lacked the authority to enjoin generically all future modifications of MPO. Rather, the most that it could enjoin were future modifications and improvements of MPO that are substantially similar to K-T's copyrighted Materials.24

Under copyright law, the district court could enjoin only those future versions of MPO that are substantially similar to K-T's Licensed Materials. LSI is free to continue its efforts to devise a non-infringing management training program, notwithstanding any expansive language in the district court's opinion to the contrary.25

B. Evidentiary Issues

LSI also raises a handful of evidentiary issues, none of which is meritorious. LSI contends that the district court erred in failing to recognize LSI's right to use K-T's Licensed Materials. This contention actually subsumes three arguments made by LSI: (1) that Vroom retained the right to continue work on the V-Y Model when he signed the Agreement with K-T, (2) that K-T's conduct indicates that it granted LSI a non-exclusive license to use its Licensed Materials, and (3) that K-T is estopped from claiming that LSI infringed its license because it encouraged and supported LSI's development of the MPO program. None of these arguments is persuasive.

Both the Licensing Agreement and doctrines of copyright law (e.g. the fair use defense) give Vroom the right to continue his theoretical work on the V-Y Model.26 But they emphatically do not give Vroom (or LSI) the right to exploit, for his commercial purposes, expression that Vroom or LSI exclusively licensed to K-T. As, this is all that the trial court held, our foregoing resolution of the copyright issues should ensure the proper balance between K-T's exclusive license and Vroom's right to continue his research.

LSI refers to various expressions of encouragement and support volunteered by key K-T personnel and suggests that these expressions were "congruent with the granting [to LSI] of an implied license" to use K-T's Licensed Materials. The court rejected this argument and found that "there is no evidence that K-T had any knowledge that the program [MPO] was being sold to the public ..." at the time that its employees offered support and encouragement for LSI's efforts. K-T's encouragement of LSI's efforts makes sense, given K-T's exclusive right to all future iterations of its Licensed Materials.

The court also observed that a transfer of copyright is not valid unless in writing. LSI quibbles with the court's observation by suggesting that K-T transferred a non-exclusive oral license. But at trial LSI never characterized its putative license as non-exclusive. And LSI ignores the court's finding that K-T never intended to grant LSI any kind of license at all. In view of these observations, the court's finding that LSI had no license to use K-T's materials is not clearly erroneous.

Finally, LSI argues that these same expressions of support estop K-T from asserting infringement of its license. The court's finding that those expressions of support pre-dated K-T's realization that MPO was being sold to the public is relevant in this context as well, and—again—that finding is not clearly erroneous. Moreover, LSI neither cites any applicable law, nor elaborates the legal elements of its estoppel argument. We therefore reiterate the admonition pronounced by Justice Holmes when he wrote: "We see what you are driving at, but you have not said it, and therefore we shall go on as before."27

LSI also insists that the court abused its discretion by excluding evidence of K-T's knowledge, support, and encouragement of the MPO program. K-T correctly responds that LSI did not make an offer of proof, which is required for error to be predicated on the exclusion of evidence. But even if the court had thus erred, which it did not, the error would have been harmless because, again, LSI offered no evidence that K-T knew LSI was selling copies of the MPO program at the time K-T personnel were expressing their support and enthusiasm for the program's development.

Finally, LSI argues that the court abused its discretion by admitting the 1972 Agreement over LSI's hearsay objection. LSI made its hearsay objection while K-T was trying to introduce the Agreement into evidence during its cross-examination of Jago. LSI could have objected that no proper foundation had been laid for the admission of the Agreement, but it objected on the grounds of hearsay instead. Surprisingly, both K-T and the court were thrown by this objection: the district court ultimately admitted the Agreement, saying that it did not think that the document was being "offered for the truth of the matter stated."

The objection was—in fact—inapposite. "Signed instruments such as wills, contracts, and promissory notes are writings that have independent legal significance, and are nonhearsay."28 A contract is a verbal act.29 It has legal reality independent of the truth of any statement contained in it. Under the objective theory of contracts, the fact that two parties signed a contract is enough to create legal rights, whatever the signatories might have been thinking when they signed it. The admission of a contract to prove the operative fact of that contract's existence thus cannot be the subject of a valid hearsay objection.30 To introduce a contract, a party need only authenticate it. Thus, in this case, LSI's objection to the admission of the 1972 Licensing Agreement on the grounds of hearsay was inapt.

III.

CONCLUSION

We affirm the district court in all of its evidentiary rulings and in its conclusion that the MPO program infringes copyrighted materials that were exclusively licensed to K-T under the 1972 Agreement. We also affirm the court's conclusion that the modified MPO program infringed those materials. We clarify the court's judgment, however, to the extent that it seems to enjoin all future modifications and revisions of the MPO program, regardless of whether they are substantially similar to K-T's copyrighted materials. Copyright law is the measure of whether LSI's future efforts will be infringing, and copyright law limits infringement to modifications that are substantially similar to protectable elements of infringed materials. To the extent that the district court's judgment could be read more broadly than that, we modify that judgment. As thus modified, the judgment of the district court is in all respects

AFFIRMED.