PROVING COPYING.

• Author: Balganesh, Shyamkrishna

TABLE OF CONTENTS INTRODUCTION 302 I. THE EVIDENTIARY LOGIC OF THE INVERSE RATIO RULE 310 A. Copying as Act and Result 311 B. Inferring Factual Copying 314 C. The Similarity-Access Relationship 319 1. Probative Value and Probability 319 2. Conditional Probative Value and the Combinatorial Approach to Inference 321 3. Cascaded Inferences from Similarity 328 II. THE EVOLUTION OF THE INVERSE RATIO RULE 333 A. Origins and Crystallization 335 B. Early Confusion: The Myth of "Rejection" 339 C. Inter-Circuit Diffusion: The Ninth Circuit and Beyond 343 1. The Ninth Circuit's Error 343 2. Reception and Validation in Other Circuits 349 D. The Ninth Circuit Correction and Repudiation 352 1. The Correction 352 2. Public Disagreement 356 3. The Repudiation/Overcorrection 358 a. The Skidmore Trial Decision 358 b. The Ninth Circuit Panel Decision 360 c. The En Banc Proceeding 363 d. Aftermath 368 E. Nuanced Application: The U.K. High Court 369 III. INTEGRATING COPYRIGHT AND EVIDENCE LAW 372 A. Rule-Based Integration 374 1. The Exclusion of Expert Testimony (Rule 702) 375 2. Limits on Limiting Instructions (Rules 105 and 403) 376 B. Principle-Based Integration 378 1. Understanding Circumstantial Proof 378 2. The Best Evidence Principle 379 CONCLUSION 381 INTRODUCTION

The test for copyright infringement has long been plagued by confusion. The modern framework was formulated three-quarters of a century ago when creative, technological, and legal conditions were considerably different. (1) It has nevertheless withstood the test of time and continues to guide courts around the country. (2) Distilled to its basics, the framework embodies two prongs. In the first, commonly referred to as "factual copying," the plaintiff must show that the defendant appropriated (that is, copied) expression from the protected work. (3) In the second, variously described as "improper appropriation," "wrongful copying," or "substantial similarity" of protected expression, the judge or jury assesses the substantiality of that appropriation to determine whether it rises to the level of an infringement. (4) As in some other areas of intellectual property law dating back to the Founding Era, Congress never set forth an infringement test, leaving this vital aspect of copyright law to judicial development. (5)

The improper appropriation prong has attracted the bulk of judicial and scholarly attention. (6) Originally conceptualized as an unguided, intuitive assessment to be made by the jury, courts have struggled to align the law with copyright's overarching goals.' This task has only been rendered more difficult by copyright law's institutional setting, namely, the role of the jury in assessing whether the appropriation was improper or substantial, (8) and perceptual differences between lay and expert observers, especially in specialized and technical fields such as computer software and music. (9)

In contrast to improper appropriation, the factual copying prong of the test has received far less attention from both courts and scholars. This is despite the reality that as a factual determination, it implicates matters of proof and thereby intersects with the basic principles and rules of evidence that inform the adjudication process. (10) Insofar as this prong has invited greater scrutiny; it has been in relation to its intersection with the more contentious improper appropriation determination, and rarely ever on its own terms. (11)

As with most matters of proof, courts recognized early on that factual copying could be established through either direct evidence or circumstantial proof. (12) Direct evidence usually involved a defendant's admission or on rare occasion, an eyewitness account of the appropriation. (13) When factual copying was disputed by a defendant, courts invariably resorted to circumstantial evidence and, applying common sense, assessed the possibility of inferring such copying from the circumstances surrounding the defendant's conduct in relation to the protected work. (14) The jurisprudence soon came to recognize two such points of inference for circumstantial proof of copying: (i) proof that the defendant had access to the protected work, and (ii) similarities between the defendant and the plaintiff's works. (15)

Recognizing the inferential nature of the inquiry based on circumstantial evidence, courts developed a sliding scale of proof between the two points of inference, which came to be known as the "inverse ratio rule." (16) Under this sliding scale approach, greater evidence of access required lesser proof of similarity, and vice versa, for a court to draw an inference of factual copying from the circumstances. (17) In situations where the evidence showed a defendant to have had unequivocal, easy, or extensive access to the plaintiff's work, a lower degree of similarity was needed to infer factual copying. (18) Further, when the similarity between the works was extensive, courts lowered the requisite amount of proof needed for access, and at its extreme dispensed with looking for proof of access altogether when the similarity was "striking." (19)

Over time, this basic set of evidentiary rules needlessly generated confusion in its application. Some courts misunderstood the very nature of the circumstantial evidence inquiry and applied the sliding scale erroneously to allow a lower level of improper appropriation, the second infringement prong, upon a showing of greater proof of access. (20) This confusion was undoubtedly caused by courts' misunderstanding of the role of similarity in the two prongs of the infringement test, a misunderstanding fueled by some courts' use of the term "substantial similarity" for both types of similarity. (21) Whereas in establishing factual copying, the similarity was merely to enable an inference of copying in improper appropriation it was to judge the legality of such copying once factual copying was found to exist. In a similar vein, some other courts thought it absurd that the sliding scale could--in theory--allow a case to move forward when there was no evidence of similarity at all but based entirely on heightened proof of access--again without a fuller recognition of the way in which the inference was meant to operate. (22)

The matter came to a head in 2020 when the Ninth Circuit issued an en banc opinion in Skidmore v. Led Zeppelin "rejecting the inverse ratio rule" and purporting to "join the majority of circuits" that had already done so. (23) Misperceiving the confusion and uncertainty that the rule had brought to the question as well as the holdings of other circuits, the Ninth Circuit abrogated the inverse ratio rule entirely. (24) Some lawyers and copyright scholars have since celebrated the Ninth Circuit's repeal of the rule, reasoning that it put to rest several decades of analytical inconsistency, which had done little more than make it easier for copyright plaintiffs to establish factual copying and move to the improper appropriation prong of the infringement inquiry. (25) Unfortunately, the Skidmore decision and its advocates misunderstand the factual copying prong.

This Article shows that the rejection of the inverse ratio rule in its entirety is based on a fundamental misunderstanding of the manner in which inferential proof and circumstantial evidence operate. Unlike direct evidence, circumstantial evidence operates by requiring the fact-finder to draw an inference from a fact that is contextually related to the conclusion sought. (26) Being inferential in nature, circumstantial evidence is therefore strongly tied to a probability or likelihood of a conclusion (rather than to a certainty). (27)

Central to such likelihood is the concept of probative value in evidence law. (28) Probative value refers to the degree to which a piece of evidence makes a conclusion--through the inference--more or less likely. (29) Circumstantial proof of a conclusion works by combining inferences from a sequential chain of facts that can account for the conclusion. (50) Each of these facts and circumstances obviously has a different probative value with regard to the final conclusion being sought. (31) Even when each is independent vis-a-vis the conclusion being proven, the probative values of the sequential facts often depend on each other (referred to as "conditional probative value") reflecting the sequential and analytical logic underlying the inferences themselves. (32) When this occurs, the cumulative inference towards the final conclusion derives from the combination of the probative values of the individual facts presented as proof and the inferences drawn from them. (33) A fact that may be of minimal probative value on its own as an inference towards the conclusion may thus gain significant probative value towards that same conclusion in the presence of other facts. (34) In other words, circumstantial proof often depends on a combination of inferences from facts that may each independently be of minimal probative value in isolation but become more probative and possibly conclusive when assessed in combination with the other facts.

It is this combinatorial approach to inferences and probative value that explains the inverse ratio rule as a mechanism of circumstantial proof of actual copying. Access and similarity represent sequential points of inference from the chain of behavior ordinarily associated with copying. A defendant must have access to the protected work, thereupon engage in an act of reproduction, thereby resulting in the production of a copy that is similar. Thus, access and similarity have their individual probative values conditionally related to each other, and it is the strength of their combined inference that allows a court to assess whether actual copying by the defendant was more likely than not. By combining two circumstantial inference points (access and similarity), the law strengthens inferential accuracy by allowing for the strength or weakness of one to...