Google v. Oracle VII: Google’s Gaslight Defense, A Lesson From the Age of Melodrama

Google and its supporting amici rely substantially on painting a landscape of a contemporary software industry that is highly dependent on what they describe as “reimplementation” of certain types of computer code. Specifically, Google maintains, that the “declaring code” they copied for the production of the Android platform is so exemplary of standard practice that Oracle’s claim of infringement threatens the manner in which the industry has learned to operate over the last two or three decades.

That’s one version of the story, and it’s a narrative that has partly been written by the ten-year Google v. Oracle litigation itself. The other (one might say simpler) version is that in 2005, Google saw that the freight train called Mobile had already left the station and was about to squash its dominant position in online search and advertising, and the company raced to build a platform that could scale very quickly, in an effort to maintain said dominance. Achieving that aim required Android to attract developers to rapidly build apps for the platform, and that, according to Oracle, is the reason Google copied 11,000+ lines of “declaring code” without license, rather than writing its own code to perform the same functions.

Google counters that the “implementing code” component of the Java methods is where the creativity lies, and that because they wrote their own implementations in conjunction with Oracle’s declarations, they did not infringe anything. This bifurcation of the Java components, along with the reimplementation argument, and even the metaphor of a train about to smash into Google’s ad business, all brings to mind a seminal copyright case from 1868 called Daly v. Palmer. The infringed work in that case was a single scene in a stage play, and the matters raised, including the context of the theater industry at the time, offer an interesting analogy for considering Google v. Oracle.

As a matter of copyright statute, we know of course that computer code is squarely in the category of “literary works.” A colleague once explained it to me by saying, “Think of code as writings that a computer can read.” Fair enough. But for the purposes of this post, let’s compare software to “dramatic works” because the infringed work in Daly v. Palmer, I would argue, is analogous to “declaring code,” and the infringer in that case (just like Google) wrote what we might call his own “implementing code,” as well. But this did not distract Judge Samuel Blatchford of the District Court for the Southern District of New York from seeing where “a piracy” had been committed. First, some background …

Daly, Boucicault, and the Age of Spectacle

The sluggishness with which the United States approached the protection of foreign authors—roughly sixty years of dithering on what was typically summarized as the “international copyright question”—contributed to a post-war theater industry rife with plagiarism and unlicensed adaptations. In particular, New York producers entertained the city’s growing middle class in a frenzied, ad hoc, process that did not resemble anything like modern theater production. A show (often some unlicensed adaptation of a European story) might be on the boards for little more than a week, and a show that ran for a month was a hit. In this hodge-podge market, creative poaching—call it the reimplementation of its time—was not only common practice, it was essential for any theater manager who hoped to remain in operation.

In this atmosphere of constant churn, it is important to understand (with regard to Oracle) that when John Augustin Daly thrilled audiences in 1867 with the exciting railroad rescue in his play Under the Gaslight, this scene was the invention that made his melodrama a megahit for its time. Now a cliché, of course, the sequence is as follows:  a young man is tied to the tracks; the train is fast approaching; a heroine, who is locked in a nearby shed, breaks free and releases the man just before the train rushes across the stage, leaving the pair unharmed.

When Daly’s “railroad scene” was soon reimplemented in the London play After Dark, dramatist Dion Boucicault set the same sequence of events in the Underground with different characters, different dialogue, and slight variations on the action—and, of course, in the context of an entirely different story. But Daly believed his creation had been appropriated and that Boucicault had copied the most valuable component of his play, especially in a time when such spectacles were box-office gold. Thus, when After Dark was arranged for production in New York, Daly filed suit, principally to enjoin the use of his “railroad scene,” at least in America.

Stage Directions are Like Declaring Code & Computers Are Like Observers

Daly was seminal for three main reasons. First, it helped clarify the meaning of the “public performance” right, added to the bundle of copyrights in 1856. Second, it affirmed that part of a whole work can be protected by copyright. And third, it is generally viewed as the source of the “ordinary observer” doctrine, which considers whether a regular individual, perceiving an allegedly infringing work, will reasonably find that it is a copy of (i.e. “substantially similar” to) the allegedly infringed work.

Judge Blatchford, who would be appointed to the Supreme Court in 1882, found that Boucicault’s “railroad scene” had infringed Daly’s creative expression. He affirmed that the stage directions were individually and sufficiently expressive, regardless of the words used to instruct the dramatic action and separate from any dialogue, or other elements, in each scene. He then held that Boucicault had copied what we would today call the “heart” of Daly’s work.

Boucicault’s play contains everything which makes the “railroad scene” in the plaintiff’s play attractive, as a representation on the stage. As, in the case of the musical composition, the air is the invention of the author, and a piracy is committed if that in which the whole meritorious part of the invention consists is incorporated in another work, without any material alteration….

Daly’s “declaring code” (if you will indulge the analogy) breaks down thus:

Character A is confined in a space behind a locked door. Character B is helpless on the tracks. Lights and sounds herald a fast-approaching train. Character A uses an available tool to break free of confines. Train gets closer. Character A sprints to remove Character B from tracks. Train races across the proscenium. Nobody dies.

These instructions, regardless of the words used, are commands that the performers and stage hands must follow in a specific manner in order to effect the desired result—fear followed by relief—in the observers. And because these commands are the structural elements of the dramatic scene (i.e. not colored by dialogue, production design, or even the overall story), we can analogize them to “declaring code” in a single API. They are unquestionably part of the whole expression the audience experiences, but they are creative on their own and, again, the heart of Daly’s mental conception. Now, compare Daly’s stage directions to this example from Oracle’s brief describing one of its declarations:

Give me a security key (which I’ll call “verificationKey”) that you want me to use to verify a signature that you previously gave me. Also tell me the algorithm I should use to verify the key (I’m calling that “verificationEngine”). Caution: You can’t just give me any algorithm. The algorithm must meet specified requirements that you can find elsewhere (a class I call “Signature”). There are two ways this might not work (“exceptions”)—the key might be wrong (“InvalidKeyException”)or the algorithm might be wrong (“SignatureException”). If so, I’ll give you an error message. If the signature is valid, I’ll say, “True” (that’s what “boolean” means).

One need not have a clue what any of that means to consider that declaring code (stage directions), directing certain operations in a certain sequence (performers), will yield a certain result from a computer (audience). And because software experts tell us that there is creativity in how one designs, expresses, and organizes declaring code, just as a dramatist will say the same thing about stage directions, we can assume the creativity is present, even if we are not programmers. Judge Blatchford correctly identified that Daly’s creativity lay in the design, organization, structure, and timing of the “railroad scene,” and held that it was sufficiently expressive.     

Google, with its emphasis on the “implementing code” code component asks the Court to come to the opposite conclusion, to retroactively hold that Daly’s stage directions merely constructed a non-creative environment in which the “real” expression took place. Like Google, Boucicault also based part of his defense on the creativity he brought to the table, emphasizing the ways in which he had transposed, or colored, his “railroad scene” differently from Daly’s. Judge Blatchford rejected this premise and articulated the “ordinary observer” doctrine at the same time, writing:

“A mere mechanic in dramatic composition can make such adaptation, and it is a piracy, if the appropriated series of events, when represented on the stage, although performed by new and different characters, using different language, is recognized by the spectator, through any of the senses to which the representation is addressed, as conveying substantially the same impressions to, and exciting the same emotions in, the mind, in the same sequence or order.”

While it may be odd to think of a computer as an “ordinary observer,” it is in fact the most ordinary of observers, one that responds predictably—not by displaying emotions like relief or pathos, but by displaying results like data, text, images, sounds, calculations, functions, etc. Daly’s “railroad scene” directions (his declaring code) elicits a distinct kind of thrill from the audience (computers), without barring any other playwright (code developer) from producing a similar response from audiences by means of equally dramatic, but distinct, rescue scenes.

Protecting Expression, Not Results

The same generalized concern was raised in Daly as it is in Google v. Oracle. If Daly were granted his injunction, it was feared, this would be tantamount to granting him a monopoly on an idea—perhaps extending him the right to license all scenes involving trains nearly hitting people or all nick-of-time rescues. In fact, Daly did try to overreach in this regard, but importantly, “…even with respect to its core holding, subsequent cases narrowed Daly rather than expanding it,” writes copyright scholar Bruce E. Boyden.[1]

The Second Circuit did not even allow Daly to extend his complaint to a production of After Dark with a modified railroad scene. Specifically, that court held, “…in all except the rescue by a third person, the complainant was not the first to conceive the story,” stated the opinion. So, even in those nascent years for both the dramatic arts and modern copyright law, Daly’s protection was almost immediately limited to the “reimplementation” of his particular rescue sequence for use in stage plays. He was not granted copyright on all nick-of-time rescues nor indeed all narrow escapes from oncoming trains. Thus, it seems fair to consider, at least by analogy, that protecting the creativity of Oracle’s declaring code (if it is indeed analogous to Daly’s railroad scene) is not coextensive to a monopoly on the desired results.

Don’t Get Stuck on the Wrong Analogy

At first glance, Daly might appear to offer the wrong defense for Oracle. One could get hung up on the fact that Daly and Boucicault each used different words (code) to instruct performers in order to produce the same general result for audiences, and that this is the relevant distinction in Google v. Oracle. In fact, Oracle has averred that if Google had used its own code to achieve the same result, this litigation would have no foundation; and one might be tempted to argue that this is where my whole analogy breaks down. But only if we misread Daly. Judge Blatchford states …

“Boucicault has, indeed, adapted the plaintiff’s series of events to the story of his play, and, in doing so, has evinced skill and art; but the same use is made, in both plays, of the same series of events, to excite, by representation, the same emotions, in the same sequence. There is no new use, in the sense of the law, in Boucicault’s play, of what is found in the plaintiff’s ‘railroad scene.’ The ‘railroad scene’ in Boucicault’s play contains everything which makes the ‘railroad scene’ in the plaintiff’s play attractive, as a representation on the stage.”

Stagecraft (declaring code) is creative. Moreover, Judge Blatchford’s reference to the “attractiveness” of Daly’s scene is a soft allusion to the market-based motive for Boucicault’s copying so valuable a scene, which brings us back to the other reason Daly is an interesting framework for reviewing Google v. Oracle. One could, of course, critique the analogy because Daly entails non-literal copying while Google entails literal copying, but Google probably would not want to go there, since case law recommends stronger protection where literal copying has occurred.

A Reminder that Reimplementation is Another Word for Copying

In addition to Judge Blatchford’s opinions in Daly v. Palmer, the history surrounding that case reminds us that the claims of any industry’s “need” to copy should always be viewed skeptically. Despite the fact that chronic appropriation (including by Daly and Boucicault) unquestionably catalyzed an abundance of theater activity in the nineteenth century, few of the plays from that period would be viewed by modern audiences as terribly “original.”

As was the case with American publishing in the nineteenth century, investment in new dramatic works—and Boucicault was among the preeminent dramatists to suffer in this regard—was stifled until the 1891 copyright law adopted reciprocal rights with foreign nations and relieved the theater industry from a vicious cycle of chronic appropriation. Hence, those who, like Google, tend to overemphasize “reimplementation,” and regard copyright as an intolerable obstacle to their business models, seem to reflect a lack of imagination that would have failed to anticipate the explosion of creativity that occurred in the twentieth century.

In fact, as both creativity and copyright law evolved since 1868, we came to recognize that endless variations on scenes a faire, nick-of-time rescues coexist within in a rich tapestry of dramatic works. Likewise, it is at least plausible—if not irrefutable—that if Google’s copying in this case is held “a piracy,” that somehow creative variety will manage to prosper in the software industry. Therefore, it will be interesting to see whether the Court is persuaded by Google’s allegedly urgent need to copy, or whether it concludes that Google simply dashed onto the tracks to rescue its ad business. 


[1] See Bruce E. Boyden, Daly v. Palmer, or the Melodramatic Origins of the Ordinary Observer. https://ssrn.com/abstract=3157782

© 2020, David Newhoff. All rights reserved.

Follow IOM on social media:

One comment

  • “In fact, Oracle has averred that if Google had used its own code to achieve the same result, this litigation would have no foundation”

    For good reasons. SQL has been developed by IBM and Oracle created a re-implementation. If such technical interfaces can be copyrighted and fair-use does not apply, they would be in trouble.

    “And because software experts tell us that there is creativity in how one designs, expresses, and organizes declaring code, just as a dramatist will say the same thing about stage directions, we can assume the creativity is present, even if we are not programmers.”

    An inventor has to be creative, too, but an invention is uncopyrightable. It can only be protected by a patent which expires after a much shorter time than a copyright.

    Copyright from a legal standpoint is about “expression”, not “ideas”. Though I suspect this distinction is an illusion, since all ideas which we share must necessarily be expressed somehow.

    “Likewise, it is at least plausible—if not irrefutable—that if Google’s copying in this case is held “a piracy,” that somehow creative variety will manage to prosper in the software industry.”

    In art or scholarship variety is good. In technical interfaces it is bad.

Join the discussion.

This site uses Akismet to reduce spam. Learn how your comment data is processed.