Memorandum and Order
March 20, 1992
In this civil action, the plaintiff, Lotus Development Corporation ("Lotus"), seeks damages and equitable relief for alleged infringement by defendant, Borland International, Inc. ("Borland"), of the Lotus copyright in its computer software program, Lotus 1-2-3. This is the same copyright for the infringement of which Lotus has obtained relief under this court's decision in Lotus Dev. Corp. v. Paperback Software Int'l, 740 F. Supp. 37 (D. Mass. 1990) (holding that Paperback's VP Planner was an infringing software product).
Lotus contends that on undisputed facts Borland's Quattro and Quattro Pro (herein, collectively, the "Quattro programs") are likewise infringing. Borland responds that its products are materially different from both Lotus 1-2-3 and VP Planner and that the court should determine on undisputed facts that the Quattro programs do not infringe the Lotus copyright. Borland argues that this result follows under the rules and standards of law applied by this court in Lotus v. Paperback. In the alternative, Borland contends that this court should reconsider its rulings of law in the Paperback case and now hold that the elements of Lotus 1-2-3 it determined previously to be copyrightable are not copyrightable. Also, Borland contends that in any event it has valid defenses that Paperback did not have.
Pending before the court are cross-motions for summary judgment, fully briefed and with factual submissions complete except for confidential materials to be presented to the court pursuant to a stipulation and protective order.
For the reasons explained in this Memorandum, I conclude that neither party's motion is supported by the record now before the court.
The primary contention of Lotus that Borland copied the Lotus user interface as a whole fails because there is a genuine dispute of fact as to whether Borland did so. Borland's primary contention that it is entitled to a summary judgment on copyrightability fails on legal grounds for reasons explained in Part VI of this Memorandum.
Each party has advanced alternative contentions for summary judgment. Each of the submissions before me, however, fails to meet the movant's burden of identifying both a clearly stated legal theory and a clear statement of undisputed facts sufficient to demonstrate an entitlement to summary judgment. In these circumstances, I conclude that it is appropriate to deny the cross-motions for summary judgment. Explicitly, however, I do not preclude the filing of another motion by either party if one can be presented consistently with the rulings stated and explained in this Memorandum.
I. Redacted Submissions
Redacted submissions (Docket Nos. 141 and 147 and certain affidavits, declarations, and exhibits) now before the court were filed pursuant to this court's practice (for the protection of parties, counsel, and court personnel from risks of inadvertent disclosure) not to allow protected materials to be filed with the Clerk before the judge (not any other person acting under a delegation of authority) has approved in writing the specific filing.
In this instance, I have concluded that the description of the protected materials, appearing in the redacted filings, is sufficient for me to determine that they are not material to any of the rulings stated and explained in this Memorandum. If counsel think otherwise, however, they may, with notice to the court and opposing counsel, bring the protected materials to the next scheduled conference to be presented to the court for examination during that conference. In light of this ruling, the parties' motions regarding the submission of confidential documents (Docket Nos. 140, 156, and 160) are dismissed as moot.
II. Disposition of the Cross-Motions for Summary Judgment
This court held in Paperback that the 1-2-3 user interface, taken as a whole, was copyrightable. That ruling was never appealed or vacated and continues to have the limited precedential force of a district court decision. The applicability of that precedent to this case is disputed, however, because Borland contends that its allegedly infringing products are materially different from VP Planner (the computer work involved in Paperback). VP Planner has been described, both in and outside this court, as an imitation (or look-alike, or clone) of 1-2-3. The conclusion that the user interface of 1-2-3, as a whole, was copyrightable was outcome determinative as to a computer work like VP Planner because, without dispute, the 1-2-3 user interface as a whole was copied. Lotus has maintained, however, both before this court and elsewhere that there is a distinction between imitations of 1-2-3 and other products such as Microsoft Excel.
In the present case, unless other issues are dispositive, it will be necessary to determine on which side of an unmarked boundary, between imitations of 1-2-3 and products such as Excel, the Quattro programs fall. I cannot determine on the record before me that the "1-2-3 interface" (also called the "emulation interface") of the Quattro programs is indisputably an imitation of the 1-2-3 user interface. Thus, for Lotus to meet its burden of showing entitlement to a summary judgment of infringement it must (a) identify expressive elements in 1-2-3 that were indisputably copied in the Quattro programs, (b) establish that those expressive elements, either separately or together, are as a matter of law copyrightable, and (c) establish that the copied expressive elements of the Quattro programs' emulation interfaces are substantially similar to copyrightable elements of the 1-2-3 interface.
The conclusion that the user interface as a whole is copyrightable (which this court reached in Paperback) does not resolve the further questions that may now have to be resolved regarding the copying and copyrightability of individual parts or a sum of parts less than the user interface as a whole. The Lotus motion and supporting submission fail to focus precisely, even in the alternative, any claim of copyrightability and undisputed copying of something less than the user interface as a whole. In these circumstances, having concluded that the claim of copying the user interface as a whole has not been demonstrated, I must conclude also that Lotus has failed to demonstrate it is entitled to summary judgment.
Similarly, Borland has failed to demonstrate support for its contention that, even if it copied something from 1-2-3, nothing Borland copied was copyrightable. As far as I am now able to determine, a factfinder may find on disputed evidence that Borland copied the 1-2-3 user interface as a whole. I have previously determined, and as explained in this Memorandum now adhere to the view, that the 1-2-3 user interface as a whole is copyrightable. Moreover, the present record suggests that a factfinder surely could find that Borland copied some expressive elements of the 1-2-3 user interface. The outcome of this case may depend on a more precise focus than Borland has presented as to what elements Borland copied and whether one, or more, or some set or sets of those elements are copyrightable.
One effective way of focusing contentions and considering whether they present genuine disputes of fact that are material under the court's resolution of contested issues of law is to consider how any fact questions might be framed for the factfinder's consideration. Especially is this so when a demand for jury trial has been filed. Questions to be submitted to the jury must at some point during full preparation for trial be framed precisely. The trial judge and trial lawyers may well commence this task early, to aid as well the focusing of arguments on motions for summary judgment. A lawyer who contends that a genuine dispute exists as to some material fact should be able to frame clearly a "written question susceptible of categorical or other brief answer," proposed for use in a "special verdict," Fed. R. Civ. P. 49(a). If, when challenged to do so, counsel cannot state precisely any proposed question of fact that is disputed and material, summary judgment is almost certainly appropriate. If, on the other hand, counsel can frame even a single disputed and material question of fact, summary judgment is inappropriate.
On the present submissions, I am not able to conclude that either party has met its burden of showing a basis for summary judgment in its favor. I therefore deny the motions at this time; however, I conclude that it is appropriate to allow the parties an immediate opportunity to focus their arguments more precisely, on the chance this may lead to disposition without the high costs of trial, and in any event will better focus issues for trial if a trial is necessary. Therefore, I deny the motions now before me, but will allow each party, on the schedule stated below, to file a new, better focused motion if the party concludes it can surmount the serious obstacles identified in this Memorandum.
The claims and defenses in dispute in this case present issues of law and issues of fact that are independently complex. The relationships among the law and fact issues dramatically enhance complexity. A jury demand and the resulting need for clear and explicit explanations to jurors of the law they need to understand in order to identify and resolve material disputes of fact produce still more consequences of significance, and perhaps more complexity as well. Complexities and their consequences have significant implications for fundamental issues of fair adjudication. They present also some pragmatic issues of case management in both pretrial and trial proceedings. The Order entered at this time is aimed at creating a context in which the parties and their attorneys will have incentives to use their adversary presentations in ways that will better focus and illumine issues for the decisionmakers, judge and jury, and will save the parties, their counsel, and the public from needlessly excessive litigation costs. Suggestions of counsel in aid of this aim are invited and encouraged.
III. Entanglement of Law and Fact as to Copyrightability and Substantial Similarity
A. Basic Methods of Separating Law and Fact
In general, an issue of law is decided by a court -- not by a jury in a jury trial and not by a trial judge as factfinder in a nonjury trial. Also, in general, if reasonable persons can differ about how a fact question should be decided on the evidence received in the trial, the jury decides it in a jury trial, and the jury decision is final; in a nonjury trial, the judge as factfinder decides, and the decision is subject to review under a deferential standard, Fed. R. Civ. P. 52(a) (findings sustained unless "clearly erroneous").
These clear and straightforward rules sometimes require supplementation, however, when the issues of law and fact are "mixed." If a court can clearly separate law from fact, it may use either of two basic methods of submitting questions to a jury. One method asks the jury to answer a mixed law-fact question; the other, a strictly factual question.
Under the first method, the court submits a combined law-fact question to the jury. To use this method properly the court must first decide the issues of law so it can give the jury clear instructions on the law. The jury's answer, if they discharge their responsibility well and truly, decides only fact questions even though their answer is in form an answer to the mixed law-fact question.
This first option may be used either for a "special verdict" under Rule 49(a), or for a "general verdict accompanied by answers to interrogatories" under Rule 49(b) of the Federal Rules of Civil Procedure.
The second option open to the trial judge is to frame "special questions" for the jury under Rule 49(a), or "interrogatories" under Rule 49(b), that ask only fact questions, not mixed law-fact questions.
A charge that directs the jury to return only a general verdict cannot use the second method because the general verdict is necessarily a mixed law-fact finding (the type required under the first option). Thus, a verdict "for the plaintiff" or "for the defendant" does not provide any basis for a reasoned determination by the court that the verdict rests on findings of fact supported by evidence and not on a different view from that of the court about the legal elements of mixed law-fact determinations.
If only a "special verdict" under Rule 49(a) is to be returned, and the questions are well framed, the jury need not have any explanation of the legal rule or rules the court will apply to determine what judgment to enter in view of the jury's findings. The judge may choose to explain anyway, but giving the explanation is a matter of choice rather than necessity.
When two or more questions of fact must be submitted, the court may submit one or more questions by one of these methods and one or more by the other. Also, the court may use a mixture of general verdict and "interrogatories," Fed. R. Civ. P. 49(b).
When issues of law and fact are easily separable, the court, when it perceives a need to do so to avoid jury confusion or bias, may use a "special verdict" form under Rule 49(a) that strictly submits fact questions; also, the court may cleanly separate the hearings before the court on the legal issues from those before the jury on the fact questions. A key purpose of trying a case in this way is to reduce the risk that the jury will be improperly influenced by evidence and arguments not admissible in relation to the strictly factual issues submitted to them.
When law and fact are not easily separable -- regardless of how deep the entanglement may be -- it nevertheless remains exclusively the responsibility of the court to decide the legal questions.
Ordinarily, it is the responsibility of the factfinder (jury or, in a nonjury case, trial judge) to decide the fact questions, but exceptions exist, as will be noted in Parts IV, VI, and VII below.
B. Problems of Entanglement
In order to use a strictly factual special verdict form, a court must cleanly separate law from fact. Not all mixed questions of law and fact can be easily separated. Indeed, there may be instances in which law and fact are so deeply intertwined that, at least as a practical matter if not strictly in principle as well, total separation cannot be achieved. In copyright cases, is "substantial similarity" such an instance? Is "copyrightability" such an instance?
1. Substantial Similarity
"Substantial similarity" is not consistently used with a single, settled meaning. In the law of copyright this phrase is used in at least two distinct senses. For this reason, the use of the term "substantial similarity," absent some contextual guidance, may be very misleading.
As part of its prima facie case, a plaintiff must prove substantial similarity between copied copyrightable elements of the copyrighted work and expressive elements of the allegedly infringing work. "Substantial similarity," as used in this statement, indicates a degree of similarity between the allegedly infringing material and what is copyrightable (that is, the copyrightable part or parts). It is not enough for a plaintiff to prove great similarity of the allegedly infringing work to uncopyrightable parts of the copyrighted work.
The similarity to copyrightable parts, it is said, must be such that an "ordinary observer" would find that there has been "unlawful appropriation." Concrete Machinery Co. v. Classic Lawn Ornaments, 843 F.2d 600, 608 (1st Cir. 1988). Thus, the copying must be extensive enough to be "substantial." This is "substantial similarity" in a mixed law-fact evaluative sense. Thus, the test for determining whether there is "substantial similarity" in this sense involves not merely a ministerial task of measuring by a yardstick or word count, but a judgmental task of weighing the quantitative measurements along with other relevant factors and coming to an overall evaluation that applies a legal test for "unlawful appropriation" to the facts.
The plaintiff must also demonstrate copying of the copyrighted work. Unlike patent law, copyright law never determines that an infringement occurs merely because two works are similar, or even identical, so long as the works are independently created. Therefore, to be successful, a plaintiff must demonstrate that the defendant had access to plaintiff's work. Such access is not disputed in this case. Nevertheless, to prove copying, plaintiff must show more than access. Copying can, of course, be proved directly, and there is some direct evidence of copying in this case. However, once access is proved, copying can also be proved by demonstrating "substantial similarity." In this context, "substantial similarity" simply means sufficient similarity of a given element of a work to an element in the allegedly infringing work to support a reasoned inference that more probably than not the element was copied from the copyrighted work. This is not similarity in a mixed law-fact sense that includes being similar enough to constitute "unlawful appropriation." Rather, the elements of the copyrighted and allegedly infringing work must be shown to be substantially (i.e., notably) similar in a purely factual sense. This is "substantial similarity" in an evidentiary sense. "Substantial similarity" in this sense is one kind of circumstantial evidence of copying.
In this Memorandum, I will try in each instance to indicate clearly (by explicit statement, by context, or by both) to which usage of substantial similarity I am referring.
2. Consequences of Entanglement
The complexity introduced by the entanglement of law and fact, bearing on "substantial similarity" (in the mixed law-fact evaluative sense) and on copyrightability has important consequences, including distinctive problems of adjudication and case management both during pretrial proceedings and during trial. Some of these consequences are relevant to the present case.
First, fashioning appropriate verdict forms and instructions to the jury is more difficult when law and fact are so entangled that the usual mode of stating an issue is in the mixed law-fact form.
Second, determining the scope of evidence that is admissible in relation to questions to be decided by the factfinder (jury or trial judge in a nonjury case) is far more difficult. Moreover, failure of the trial judge to resolve problems of admissibility, especially in relation to proffered expert opinion testimony, may substantially impede fair trial because of risks of jury confusion and the incentives to counsel and experts to try to influence the jury with arguments on the law cast in the form of opinions on the mixed law-fact questions.
Third, when the applicable law requires the application of a standard for decisionmaking that is derived from a congressionally mandated accommodation among conflicting public policy interests and thus "reflects a balance of competing claims upon the public interest," Sony Corp. v. Universal City Studios, Inc., 464 U.S. 417, 431, 78 L. Ed. 2d 574, 104 S. Ct. 774 (1984), consequences of the first and second types are enhanced, sometimes beyond measure. Difficult as the task of applying that balance in particular cases may be, however, it is not open to a court to ease the task either by revising that balance, Feist Publications, Inc. v. Rural Telephone Service Co., 113 L. Ed. 2d 358,. . . U.S. . . ., 111 S. Ct. 1282 (1991), or by leaving the jury free to revise that balance. A congressionally mandated balance must be respected by courts and juries alike.
The mixed law-fact evaluative "substantial similarity" issue in copyright law inevitably produces the first and second of these types of consequences, and in particular cases may produce the third as well unless effective measures of judicial control are used. For example, consider whether or not expert opinion evidence about the ways in which two computer software programs are similar and dissimilar should be heard by a jury. One may argue for a negative answer under the authority of Concrete Machinery Co. v. Classic Lawn Ornaments, Inc., 843 F.2d 600, 608 (1st Cir. 1988) (substantial similarity is to be judged under the "ordinary observer" test "unaided by expert testimony"). Even if it may be appropriate and useful to a jury in understanding purely historical facts about characteristics of computer programs to hear expert opinions bearing on historical facts, still opinions that explicitly or implicitly depend on policy premises inconsistent with the law (as determined by the trial judge, subject to correction on appeal) should not be heard by the jury.
Affidavits of experts and arguments of counsel before me in this case illustrate the risks of contentious and potentially unfair proceedings incident to the strong tendency of witnesses and lawyers to intertwine every statement about factual similarity with opinions and arguments about what is, or even should be, the kind of similarity that is legally significant. These opinions are, of course, opinions on the law. Allowing a jury to hear them expressed by witnesses and lawyers is fundamentally inconsistent with the rule that jurors take their instructions on the law solely from the court.
I will return to the issue of "substantial similarity" in the mixed law-fact evaluative sense in Part IV, below.
Issues of "copyrightability" in computer software programs present the third type of consequence of entanglement of law and fact in a dimension so extreme that the problem is appropriately viewed as one different in kind and not merely degree. This is one reason, though not the only one, for concluding that at least in the present case, and perhaps more generally, all questions of law and fact bearing on copyrightability of elements of a computer software program are to be decided by a court, not a jury. That issue is considered more fully in Parts IV, VI, and VII, below.
IV. Entanglement of Copyrightability and Substantial Similarity
The complexity of entanglement of issues in this case extends not only to issues bearing only on copyrightability, and to issues bearing only on substantial similarity in the mixed law-fact evaluative sense, but also to the interdependence of copyrightability and substantial similarity.
The point is illustrated by Borland's argument that Borland is entitled to summary judgment on the issue of copyrightability, Docket No. 141, at 1 (emphasis added), because the elements of Quattro and Quattro Pro that are alleged to be substantially similar to elements of Lotus 1-2-3 concern only the way certain parts of the programs function, not copyrightable expressions. Id. at 4-8.
Borland argues that Concrete Machinery, 843 F.2d at 600, requires that issues of substantial similarity be addressed first, to determine whether "there are sufficient articulable similarities to justify a finding that the defendant has copied from the protected work." Borland's Memorandum, Docket No. 141, at 114 (quoting Concrete Machinery, 843 F.2d at 608 (citing Arnstein v. Porter, 154 F.2d 464 (2d Cir. 1946)). (In this argument, one may note, Borland is using "substantial similarity" in the evidentiary sense.) Then, Borland argues, "the court must determine whether the copying is sufficiently substantial to constitute 'unlawful appropriation' ('illicit copying')." Id. at 114 (quoting Concrete Machinery, 843 F.2d at 608). (This argument, it may be noted, focuses both on a strictly factual question about copying and a mixed law-fact question about whether the copying is sufficient to constitute "unlawful appropriation.") "Assuming copying of protected aspects is established, the trier of fact can then assess pursuant to the 'ordinary observer' test whether there is substantial similarity between the protected expression and the accused work." Id. (quoting Concrete Machinery, 843 F.2d at 609).
The first step formulated in the Concrete Machinery opinion may be perceived as itself composed of two parts -- determining what are the "protected aspects of the copyrighted work" (that is, the "copyrightability" issue) and determining whether there has been copying of one or more of those aspects. It is not clear that one of these two need precede the other in an analysis of the evidence and arguments presented to a court. Indeed, they may be sufficiently interrelated that in most cases they are best addressed together -- a possibility at least suggested by the First Circuit's combining the two into the first "step" of its "two-step" test.
In any event, the argument that First Circuit precedent requires a particular order of analysis of a given problem is contrary to a substantial body of precedent in the circuit. For example, the First Circuit spoke quite early to make clear, in the context of discrimination cases, that the factors analysis set forth in McDonnell-Douglas Corp. v. Greer, 411 U.S. 792, 93 S. Ct. 1817, 36 L. Ed. 2d 668 (1973), is intended as a useful guideline and not a mandatory straitjacket. See, e.g., Dance v. Ripley, 776 F.2d 370, 373 (1st Cir. 1985) ("This circuit, along with other circuits, has rejected the argument that McDonnell Douglas and [Texas Dept. of Community Affairs v.] Burdine [, 450 U.S. 248, 101 S. Ct. 1089, 67 L. Ed. 2d 207 (1980)] set forth a 'rigid, three-step proof process in Title VII cases.'") For like reasons, I conclude that the pattern of analysis and decisionmaking presented in Concrete Machinery is not meant to be a straitjacket. Thus, I conclude that the order of addressing the various elements of a prima facie case fashioned in Concrete Machinery need not be applied to a case in which there are compelling practical reasons for a different order of proceeding.
Perhaps the procedure discussed in Part V, below, fits the Concrete Machinery model because the "first" step of that model is itself two-fold, as noted above. In any event, regardless of whether the procedure discussed in Part V precisely fits the Concrete Machinery model, I conclude, for reasons explained in Part V, that it is appropriate to consider an order of proceeding in this case that places issues of substantial similarity (in the evidentiary sense) ahead of issues of copyrightability.
V. Sharpening the Focus
In this case, if it is determined that neither a motion for summary judgment nor a motion for directed verdict should be sustained, how might issues involving substantial similarity (in the evidentiary sense) be submitted to the jury in a verdict form in the trial of this case?
To establish infringement, Lotus must prove, either directly or circumstantially, that Borland copied. Circumstantial proof of copying is made by showing access to the copyrighted work and substantial similarity (in the evidentiary sense). In the present case, I conclude that a reasonable trier of fact could find that the Quattro programs' user interfaces were not copied in their entirety from Lotus 1-2-3. However, because access is admitted, the record before me contains both circumstantial proof of copying (e.g., all of the commands in the Lotus 1-2-3 menu structure appear verbatim in the Quattro programs' menu structures in similar hierarchies) and direct proof of copying (e.g., Borland named certain files involved in the 1-2-3 emulation "123.rsc" and "123.mu"). It does not necessarily follow, however, that a jury must on this record find that some particular element was copied from Lotus 1-2-3 rather than being either independently created or copied from some other source.
For any elements of 1-2-3 that Lotus wishes to argue were copied, Lotus must demonstrate (either beyond dispute on motion for summary judgment, or else at trial by a preponderance of the evidence) that the alleged copying from 1-2-3 occurred. Lotus may do so either directly or circumstantially. To demonstrate that the coying was illicit Lotus must also show substantial similarity in the mixed law-fact evaluative sense and copyrightability of those elements.
Returning to the question posed above (How might issues involving substantial similarity in the evidentiary sense be submitted to the jury?) I take note that one option to be considered is to begin with the less complex issue of copying and with a broad focus, then moving toward narrower questions concerning what elements of Quattro or Quattro Pro were copied from 1-2-3.
Would it be appropriate to frame the first question on a Verdict Form in a way such as this (with a similar question to follow for the other allegedly infringing work)?
(a) Do you find that the Quattro Pro user interface as a whole was copied from the Lotus 1-2-3 user interface as a whole? 1"
(b) Do you find that the part of the Quattro Pro user interface called the "emulation interface" (also called the "1-2-3 compatible interface") as a whole was copied from the Lotus 1-2-3 user interface as a whole?
(c) If NO, do you find that some part, and, if so, which of the following part or parts of the Lotus 1-2-3 user interface were copied into some part of the Quattro Pro "emulation interface" (also called the "1-2-3 compatible interface")? (1) The menu commands ____) ES _____NO (2) The menu structure ____) ES _____NO (3) The command sequence ____) ES _____NO (4) The long prompts ____) ES _____NO (5) The macro facility ____) ES _____NO