Joshua L. Simmons* A pdf version of this article may be downloaded here.  

Introduction

Despite the persistent notion that modern invention is performed by individual inventors in their garages, few would disagree that today most patentable inventive activity occurs in corporate and university settings and that most individuals who would be labeled “inventors” in the twenty-first century are employees of corporate entities.[1] In the corporate setting, if an employee creates a copyrightable work within the scope of his or her employment, the Copyright Act not only grants ownership of the work to the employer but actually considers the employer the “author” of the work.[2] By contrast, under the Patent Act, one who creates an invention is its inventor, and ownership will only pass to another, including an employer, through a written assignment.[3] In other words, unless there is an agreement to the contrary, an employer does not have any rights in an invention “which is the original conception of the employee alone.”[4] Given the close relationship between copyright law and patent law,[5] it is puzzling that employees’ works and inventions would be treated so differently under the two disciplines. Yet, when one considers their historical foundation, it becomes clear that there were pivotal differences between the needs of the two disciplines in the nineteenth century that led to their modern formulations.[6] In particular, whereas the type of copyrightable works created in the nineteenth century transitioned from individual labors to collaborative labors among multiple individuals working together, which necessitated the collecting of rights in order to make use of the resulting copyrightable work, patentable inventions continued to be perceived during the nineteenth century as the work of individuals. Moreover, patent law developed other, more limited doctrines that provided some rights to inventors’ employers. Today, however, most patentable inventions are invented by multiple inventors in a collaborative environment.[7] In fact between 1885 and 1950, the percentage of U.S. patents issued to corporations grew from 12% to at least 75%.[8] These numbers have only continued to increase in the past decade.[9] Moreover, the recently passed Leahy-Smith America Invents Act partially recognizes this change by permitting an “applicant for patent” to file a “substitute statement” instead of an inventor’s oath or declaration in certain circumstances.[10] Nevertheless, patent law remains stuck in the past. The failure to modernize how patent law handles employee invention has led to a string of significant court opinions, including from the Supreme Court and the Federal Circuit, holding that employers had not received adequate assignments to their employees’ patented inventions and thus could not bring suit based thereon.[11] In order to resolve these issues and bring patent law into the twenty-first century, the Patent Act should be amended to borrow from the Copyright Act and adopt a principle similar to the work made for hire doctrine that would grant employers the rights to their employees’ inventions made within the scope of their employment.[12] This article proceeds in four parts. Part I discusses the current state of copyright and patent law vis-à-vis employers’ rights in their employees’ intellectual labor. This part will compare copyright’s work made for hire doctrine to three patent law doctrines: the shop rights doctrine; the hired-to-invent doctrine, and the employee improvements doctrine. Part II describes the evolution of copyright law during the nineteenth century from the former regime, which vested employees with presumptive ownership of their work, to the current regime, which grants employers presumptive rights to their employees’ efforts. Part III describes patent law during the time that copyright law was changing so dramatically—including the development of the patent law doctrines described in Part I—and posits potential reasons that patent law did not make a similar move. In particular, (a) patent law’s development of the doctrines described in Part I provided similar benefits, although in the form of different rights, to those provided by copyright law’s work made for hire doctrine, and (b) the perceived nature of invention in the nineteenth century did not call for a unification and codification of those doctrines the way copyright law’s did. Finally, Part IV argues that patent law should modernize and develop an “inventions made for hire” doctrine.

I. IP and the Employer-Employee Relationship

It is well settled in the United States that copyright law and patent law treat legal title in the intellectual labor of employees differently.[13] Section A discusses the copyright law doctrine of work made for hire and its benefits for employers. Section B discusses doctrines that provide some of those benefits in patent law.

A. Copyright Law

Under the Copyright Act, title in a copyrightable work initially vests in the author or authors of a work.[14] However, “author” is a term of art with meaning beyond the creator of or “person who originates or gives existence” to a work.[15] In particular, where a “work made for hire”—another term of art—is concerned, the employer is “considered the author” unless the parties agree otherwise.[16] A work may be considered made for hire if (1) it was prepared within the scope of an employee’s employment, or (2) it is a certain type of commissioned work and the parties have so agreed previously.[17] Whether the creator of a copyrightable work is an employee for purposes of the work made for hire doctrine is determined by looking to the common law doctrine of agency.[18] In Community for Creative Non-Violence v. Reid, the Supreme Court identified thirteen factors that it considered relevant to whether an individual was considered an employee as a matter of agency law.[19] The Supreme Court declared that “[n]o one of these factors [was] determinative,”[20] but at least one circuit has held that not all factors are created equal and some will be important in “virtually every situation” while others “will often have little or no significance.”[21] The more important factors are to be given “more weight in the analysis.”[22] Once a work is considered made for hire, the employing or commissioning party enjoys several benefits over a work that is not considered made for hire, but rather is transferred to the employing party. First, unlike in patent law, it immediately becomes the owner of a legal right to the work.[23] In patent law, an inventor must first file an application with the United States Patent and Trademark Office (“PTO”), which is then examined by a PTO employee to determine if the alleged new invention is entitled to a patent.[24] Copyrights, on the other hand, vest immediately once an original work of authorship is fixed in a tangible medium of expression.[25] For a company to gain an ownership interest in a copyrighted work not made for hire, the interest must be negotiated for and transferred in a signed written document.[26] When a work qualifying as a work made for hire is created, however, no negotiation or written instrument is required for that specific work because, for copyright purposes, as long as agency would hold that the work’s creator was an employee working within the scope of his employment, his employer is considered the work’s author.[27] There are two subsidiary benefits to the immediate grant to the employer of a work’s copyrights. First, no court intervention is required to transfer the rights. As will be described below, in certain cases, patent law will grant employers rights in their employee’s patents, but to vest those rights the employee must still sign a document transferring ownership to his employer. If he refuses, a court order may substitute for the transfer. Second, the work made for hire doctrine allows a grant of rights without needing to define what is being granted. Whereas when a grant of rights is negotiated, the transferor may come back to the transferee to argue about which rights were actually transferred and the scope of the intended transfer.[28] Second, an employer holding the rights in a work made for hire may register the work in the Copyright Office in its own name.[29] This is an optional but important step, because without registration a copyright infringement action may not be initiated,[30] and generally, statutory damages are not available for infringements prior to registration.[31] Third, the work made for hire doctrine grants employers all of the rights associated with copyright ownership. Thus, they can exclude others from using the work, and leverage that right to extract rents in exchange for a license authorizing someone else to use the work.[32] Further, employers using works made for hire are authorized to then transfer their ownership of the works’ copyrights to another party, either in whole or in part.[33] They can also, of course, exercise any of the rights authorized under the Copyright Act itself, including creating derivative works.[34] One set of rights an employer is not granted under the Copyright Act, however, is the right to attribution and integrity granted for works of visual art.[35] Fourth, the work made for hire doctrine also protects employers from future rights granted to authors. For example, the Copyright Act of 1976 granted authors the right to terminate transfers, but those new rights did not apply to creators of works made for hire because their employers were considered the author.[36] In addition, after the 1976 revisions, employers remained shielded from the exercise of the transfer termination provisions because no transfer is considered to have taken place under the work made for hire doctrine;[37] the employer was the work’s original “author.” Fifth, the work made for hire doctrine grants employers certainty with regard to the duration of their copyrights. Works not made for hire subsist for the life of the author plus 70 years.[38] This means that those to whom the original author transfers must know the duration of the original author’s lifetime in order to calculate the duration of the copyrighted work’s term. Works made for hire, however, are merely granted a term of 95 years from publication or 120 years from creation, whichever expires first.[39] This time frame serves two functions. First, it simplifies the requirements of an employer to keep track of the duration of its copyrights. Second—grim though it may be—if the employee were to die within 25 years of publication or 50 years of creation, the employer would benefit from a longer copyright term than if the work were considered made for hire.[40] Finally, the work made for hire doctrine applies to all works universally, regardless of the type of work or the employee’s level of contribution. In other words, a work created and conceptualized entirely by only one employee is treated the same as a work created by hundreds of employees, each of whom contributed minor improvements and expression to the work.

B. Patent Law

Some of the benefits described in the previous section are mirrored by certain patent law doctrines—or are irrelevant with regard to patents—but others have no patent law equivalent. Section A described eleven benefits associated with copyright law’s work made for hire doctrine: (1) grant of title without negotiation;[41] (2) grant of title without transfer/assignment; (3) grant of title without court intervention; (4) registrability in the employer’s own name; (5) exclusion of others from using the work and licensing rights in the work to others; (6) use of the work; (7) transfer of title to others; (8) protection from future rights granted to work’s creator; (9) shield from transfer termination; (10) benefits with regard to copyright term duration; and (11) universal application regardless of level of contribution.[42] It is immediately apparent that the benefits of the termination shield and duration are not relevant to patent law because patents are granted for a set term of years—whether the base of 20 years or an extended term due to regulatory or PTO review[43]—not tied to the inventor’s lifetime. Therefore, there would be no special duration benefits should an inventor’s employer be considered an “inventor” for purposes of the Patent Act. In addition, unlike copyright transfers, which as discussed above may be terminated upon notice, patent assignments are not subject to termination.[44] The other benefits, however, require additional consideration. The three following sections describe modern patent law doctrines that give employers some, but not all, of the benefits of copyright law’s work made for hire doctrine.

1. Employer Use and Shop Rights

One of the benefits of the work made for hire doctrine is that an employer is guaranteed the right to make use of an employee’s creative work. Employers may receive a similar benefit from patent law’s shop rights doctrine.[45] However, because patent law describes negative rights, it is a protection against infringement actions by the employee patent holder and her assigns, and not an affirmative grant of use of the patented invention. Shop rights arise when an employee, “during his hours of employment, [and] working with his [employer’s] materials and appliances, conceives and perfects an invention for which he obtains a patent.”[46] The courts have held that in such circumstances, the employee by force of law must give his employer a nonexclusive, royalty-free right to practice the invention.[47] The doctrinal basis behind the shop right remains fuzzy,[48] as courts have described it being based on (a) a license implied in fact,[49] (b) estoppel,[50] and (c) equity and fairness.[51] Shop rights do not have the same scope as work made for hire, however. A shop right is not an ownership interest in the patent, which means that an employer does not have either the right to exclude others by threatening to sue for infringement or the right to file a patent application.[52] Similarly, a shop right is non-exclusive, which means that the employee patent holder is free to assign rights to others by granting them a patent license. In addition, a shop right is not transferable other than with the sale of the entire appurtenant business.[53]

2. Commissioned Invention and the Hired-to-Invent Doctrine

As mentioned in section A, under copyright law a work will be considered for hire if it was made by an employee within the scope of his or her employment. Similarly, patent law’s hired-to-invent doctrine grants an employer rights to the inventions of its employees if the employee was hired to invent them. Under this doctrine, an employee hired to solve a particular problem or to invent in a certain field will forfeit his patent rights even without a written contract.[54] Just as in the work made for hire context, courts have established a number of factors to indicate whether an inventor has been hired to invent.[55] However, unlike work made for hire, the hired-to-invent doctrine requires the employee to assign any patent obtained; it does not vest title immediately in the employer upon invention.[56] Although, failure to assign after a court order to do so may result in the order having the same effect as an assignment.[57] At first blush, this patent law doctrine would appear to accomplish much of what the copyright law work made for hire doctrine does. However, there are differences. First, work made for hire vests title in an employer immediately. The hired-to-invent doctrine merely obligates the inventor to assign the invention to his or her employer. True, no negotiation is required after the inventive activity occurs, but until an assignment is signed, the invention’s creator continues to hold the patent on the invention—albeit in trust for his employer. Furthermore, if the invention’s creator refuses to assign his patent rights, then a court order is required. Second, the doctrine is only effective against employees who are actually hired to invent. Employees who create an invention within the scope of their employment but who were not specifically directed to do so retain their patent rights.[58] Third, unlike the work made for hire doctrine, which permits an employer to file a copyright registration in its own name, the hired-to-invent doctrine still requires the patent application to be filed in the inventor’s name, even if the employer, as assignee, files a patent application on his behalf. Finally, as the employee is still considered the “inventor” under the Patent Act, it is possible for future rights to be given to the “inventor” that would not be automatically vested in the inventor’s employer. For example, in the copyright context, when the right to terminate transfers of ownership was introduced as a right of “authors,” it did not apply to those whose works were made for hire, as their employers were considered the “authors.” Similar rights could be granted to “inventors” that would affect the assignees of the inventors’ patent rights.

3. Employer Inventions and Employee Improvements

In addition to the situation where an employer hires an employee to invent, an employer will also receive the benefit of insignificant improvements by his employees on inventions the employer has conceived of himself.[59] Under this doctrine, when an employer conceives of an invention, but receives ancillary suggestions or improvements from his employee, he retains all rights in the invention, including any employee improvements.[60] To qualify under the employee improvements doctrine, the employer must have had a “plan and preconceived design,”[61] and the improver must be an employee.[62] The employee improvements doctrine provides a minimal step toward the work made for hire doctrine, but does not come anywhere close to providing a substitute. First, the doctrine still requires there to be an inventor. All the doctrine does is subsume improvements by another person into the inventive exercise of the employing inventor; it does not operate to grant an employing corporation rights. In addition, the improvements that are granted must be ancillary, so any benefits from the doctrine will be minimal. Thus, if the employee creates a non-ancillary invention or improvement, the employer still needs to qualify under one of the previous doctrines or negotiate an assignment. At base, the employee improvements doctrine determines whether an employee should be considered a joint inventor or not, which does not come close to achieving the same benefits of the work made for hire doctrine. Given that patent law never developed a doctrine that defaulted patent rights to an employer the way copyright law did, one is left to wonder why. Why is it that copyright law in the nineteenth century began to apply an employer presumption? And why did the courts not feel the need to apply a similar presumption in patent cases? Part II discusses the development of the work made for hire doctrine in the nineteenth century, and Part III discusses potential reasons that a similar doctrine was not developed in patent law.

II. Development of the Work Made for Hire Doctrine

The work made for hire doctrine predated the Copyright Act of 1976. In fact, although it was first codified in the Copyright Act of 1909,[63] Professor Catherine L. Fisk has claimed that the doctrine “neither was invented by the drafters of the 1909 Act, nor was it well recognized in the cases before 1909.”[64] Instead, between the American Civil War and the passage of the Copyright Act of 1909 the default rule in copyright cases quietly switched from employee ownership to employer ownership. Professor Fisk argues that this move was justified by “cloaking” artists in an “aura of . . . romantic genius” to argue for greater legal protection, while simultaneously “downplaying or ignoring individual creative genius so as to assert corporate ownership over those copyrighted works.”[65]

A. Antebellum

Prior to the American Civil War, the default—and essentially the rule—was that the employee author was the owner of his works.[66] In the early days of copyright, protection was of limited application. The first copyright act—the Copyright Act of 1790—only protected maps, charts, and books.[67] Among the books that received copyright protection were the books in which court decisions were reported. Thus, it was only natural for the first copyright cases concerning employment to involve books containing those decisions and the authors that authored them. In the first case involving an employed author, the author of twelve volumes reporting the caselaw of the United States Supreme Court assigned his copyrights to a publisher. The publisher then sued a subsequent case reporter for copyright infringement because he had produced a volume titled “Condensed Reports of Cases in the Supreme Court of the United States,” which included the work of the prior reporter.[68] The Supreme Court, noting that the original reporter had failed to satisfy the formality requirements, held that it could not confer common law copyright on the initial reporter. However, what is telling about the opinion is that it appears to assume that the original reporter—and not the publisher—would be entitled to the copyrights in his work, even though his status as an employee was uncertain at best.[69] Later cases in the antebellum era involved school books,[70] theatrical works,[71] and cartography.[72] What is particularly interesting about this period of time is the courts’ clear pro-employee stance. In later decades, courts’ rhetoric would be inconsistent with the holdings of their decisions. While courts would claim to celebrate the genius and romanticism of the independent author, they would simultaneously grant first publishing houses and later employers the rights to that labor. During the antebellum nineteenth century, however, courts’ rhetoric praising the individual was consistent with the holdings of their decisions, which largely gave employees ownership of the copyrights in their works. The only cases where employers were held to have been granted copyrights in their employees’ works involved expressed contracts and cartography which, by its very nature, requires coordination among various individuals. True, smaller scale maps could be created by one person on their own, but for anything of a greater scale, coordination among various parties would be required.

B. The Nineteenth Century Postbellum

In the period during and after the American Civil War, courts began to hold that employers had been granted copyrights in their employees’ works, not by operation of contract, but based on the employment relationship. Particularly important to the discussion in Part III is that while cases originally held that the employment relationship granted an employer rights in his employees’ work by virtue of the involvement of a corporate president, later cases created the legal fiction of the company as author.[73] In the postbellum period, like the antebellum period, the relevant cases generally involved legal publications,[74] as well as theatrical works.[75] One of the theatrical work cases is particularly important for the purposes of this article. In Keene v. Wheatley, the court ruled against the plaintiff-employer on many of her arguments, but with regard to ownership of one of her employee’s works, the court stated:

Here, [the employee], while in the general theatrical employment of the [employer], engaged in the particular office of assisting in the adaptation of this play; and made the additions in question in the course of his willing performance of this duty. [The employer] consequently became the proprietor of them as products of his intellectual exertion in a particular service in her employment.[76]

For this proposition, the court did not rely on copyright law. Instead, the court looked to patent law, and determined that since “[w]here an inventor, in the course of his experimental essays, employs an assistant who suggests, and adapts, a subordinate improvement, it is, in law, an incident, or part, of the employer’s main invention,” the plaintiff in Keene was entitled to the “literary proprietorship” of her employee’s work.[77] In the eyes of the Keene court, the employee was merely adding minor enhancements akin to the patent doctrine of employee improvement. Yet, over time the doctrine of work made for hire found increasing judicial support and eventually became a default rule in favor of employers.[78] Over the next several years, this new doctrine was mentioned in dicta, but in no case did a court rule that an employer possessed rights over its employee by virtue of the default rule.[79] Instead, the presumption was switched surreptitiously. In Callaghan v. Myers, for example, while holding that a court reporter-employee held the copyrights in his work instead of his government employer, the Supreme Court determined that this was the case only due to “a tacit assent by the government to his exercising such privilege,”[80] This shift in emphasis made employer ownership a presumption that was only rebutted by tacit agreements and expressed contracts. Why did the courts make this shift? Professor Fisk posits three potential factors:

[1] Courts might have felt that a default rule of employer ownership was more likely to reflect the intent of most parties and wanted to save the parties the trouble of negotiating for employer ownership. . . . [2] [C]ourts might have begun to see employers as possessing a stronger moral claim and believed that any employee who planned to assert copyright ownership ought to be forced to disclose that intent and negotiate for it. . . . [3] [A]s changing assumptions about the nature of authorship strengthened the rhetorical force of the employer’s claim, a default rule of employer ownership might have seemed more intrinsically appealing, irrespective of whether the parties might negotiate around it.[81]

In response to the shifting sands of copyright law doctrine of the nineteenth century, one might expect that parties would have begun contracting around the default rules, but two factors may have made contracting prohibitively difficult:

First, the costs of transacting might be high when the parties have to discuss something as touchy as authorship. Employers might have been afraid to alienate employees by demanding assignment of the copyright, preferring to run the risk of litigation later. Employees may have lacked legal sophistication to realize that it was necessary to contract for copyright ownership. Second, the instability of the law may have made enforcement of any contract they did reach highly uncertain.[82]

In any case, by the nineteenth century, the work made for hire doctrine was in the wind, whether because of changes in the parties themselves, or merely the impressions of those sitting on the bench.

C. Corporate Ownership and Codification

The work made for hire doctrine was codified by Congress in 1909, but its foundations had existed prior to that time. As described in section B, the concept of authorship and ownership of copyright were shifting over the last half of the nineteenth century. However, it was not until 1885 that a court held that a corporation, and not an individual employer, held the copyrights in a work.[83] It was at this time that the legal fiction arose that a corporation could be an author. This idea makes sense when viewed one way: corporations are merely collections of individuals working toward, generally, hierarchically determined joint goals. However, it flies in the face of the entire notion of the individual romantic author. The concept behind the expression embodied in a copyrightable work can be created by any number of different people as it moves through the corporate process. This is particularly clear today when one looks at the large corporations behind music, television, and film.[84] These corporations are made up of hundreds of thousands of employees, and even more independent contractors. To produce one blockbuster film may require hiring and coordinating thousands of individuals, each of whom may contribute to a portion of the film. Without the work made for hire doctrine, these companies would have to rely on a messy web of contracts and the employee improvements doctrine, and even then, it is unlikely that they would be assured that all the relevant rights had been secured. By 1899, however, the courts had come to realize the need for a doctrine that allowed corporate employers to control the copyrights in the works of their employees by default. In Collier Engineer Co. v. United Correspondence Schools, a salaried employee hired “to compile, prepare, and revise . . . instruction and question papers” had moved to a new employer and prepared similar materials.[85] His former employer made a motion for a preliminary injunction, which the court denied, but in so doing determined that the employer was entitled to any copyright in the original materials.[86] This idea of protecting employers when their employees move to other employers will resurface in Part III as a justification for employer ownership of patent rights. Later cases embraced the idea that when an employer hired individuals to produce copyrightable works, the copyrights in those works were granted to the employer.[87] In this way, by the turn of the century, copyright law and patent law had developed fairly parallel doctrines with regard to employee works and inventions. In 1909, however, the Copyright Act was revised to codify the work made for hire doctrine by adding the following language:

[T]he word ‘author’ shall include an employer in the case of works made for hire.[88]

By including employers of those that create works made for hire within the definition of the term “author,” Congress continued in the direction the courts had been heading, but went farther by creating an explicit employer presumption. Professor Fisk identifies three reasons for this change: (1) an ease in drafting the Act, (2) avoiding constitutionality challenges, and (3) ensuring that copyright ownership would vest initially in employers so they could benefit from copyright renewals.[89] And that, ladies and gentlemen, was history. Since 1909, work made for hire has remained a mainstay of American copyright law. Yet, despite over one hundred years of the work made for hire doctrine in copyright law, patent law remains based on the rule that only the individuals that create an invention can be considered inventors. Moreover, for an employer to claim patent rights, he must either negotiate an assignment, or rely on one of the doctrines discussed in Part I.B.: shop rights, hired-to-invent, or employee improvements. Barring one of these options, an employer holds no rights in a patent absent a written assignment. Part III will explore the patent cases of the late nineteenth century, and posit potential reasons that patent law did not take a similar route.

III. Employment in Patent Law

As previously described, patent law never developed a doctrine that provided all of the benefits of copyright law’s now well-established work made for hire doctrine. This divergence between the two disciplines can be explained by two factors. First, inventive activity in the nineteenth century was perceived to be individualized and not to require significant coordination among individuals. Second, patent law developed the three doctrines described in Part I.B., which provided satisfactory protection to employers who wanted to continue using their employees’ inventions.

A. Antebellum

In order to understand the nature of inventive activity during the nineteenth century, one must understand that fundamental shifts were occurring in society at the time that changed both who the inventors were and what they were inventing. The United States has always been a country of inventors. In fact, both George Washington and Thomas Jefferson are credited with numerous inventions, including plows, adjustable writing desks and swivel chairs.[90] However, at the end of eighteenth century, there were “probably no more than 300 people who we would now class as scientists in the entire world,”[91] By 1800, there were about a thousand, and by 1900 there were approximately 100,000.[92] Due to the growth in the number of scientists, science began to shift from “a gentlemanly hobby, where the interests and abilities of a single individual [could] have a profound impact, to a well-populated profession, where progress depend[ed] on the work of many individuals who [were], to some extent, interchangeable.”[93] Similarly, the number of non-scientist inventors blossomed throughout the nineteenth century leading to an increased ability to appreciate the “interplay between science and technology, particularly in the fields of electricity and engine building,” which “led to a host of new practical machines that changed communications, transportation, the home, the workplace, and the farm,”[94] For example, at the beginning of the nineteenth century, America remained a nation of farmers who used the same hand implements that had been used for centuries to harvest their crops and who would travel primarily by horse or boat.[95] In the time before the Civil War, however, non-scientist inventors made many advances, including the invention of the steamboat, farming machinery, the telegraph, and synthetic materials.[96]  Such advances were only possible due to the increased interplay between science and technology that continued to grow throughout the nineteenth century.[97] These developments also led to societal changes, including increased prosperity and the end of slavery as new equipment was developed that could perform the work better and more cheaply.[98] Despite the democratization of science and invention in the nineteenth century, however, inventive activity continued to be considered the work of individuals. In fact, there are “no reported cases before 1843 in which an employer claimed, as against an employee, ownership of a patent because the inventor had been working for him at the time of the invention,”[99] Thus, in the nineteenth century, the general perception was that inventorship was the work of certain individuals, who were considered the great men of the time. One visual example of this mentality is Christian Schussele’s 1857 painting titled Men of Progress, which depicts Benjamin Franklin overlooking nineteen nineteenth century inventors:[100]

Despite this public perception, however, the reality was that even these great men had assistance in creating their famous inventions. Among those depicted in Schussele’s painting is Charles Goodyear, who is remembered for his invention of vulcanized rubber in 1839.[101] He received a patent on his process in 1844,[102] and almost immediately became embroiled in patent litigation.[103] Not only was Goodyear forced to litigate against his competitors, however, but also against his own employees, one of whom attempted to patent one of Goodyear’s inventions.[104] Fortunately for Goodyear, despite recognizing a presumption in favor of the employee—who actually made the machine—the courts considered Goodyear’s employee’s activities to have been at Goodyear’s direction, and thus, they determined that it was Goodyear, and not his employee, that was the true inventor.[105] Despite the favorable result for Goodyear, the case only underscored the presumption that employees owned their inventions, as the court continued to articulate employee-ownership as the general presumption in such cases. Thus, time and time again, courts were required to find exceptions in order to protect employers from the presumption’s consequences. For example, in 1843 the Supreme Court, in McClurg v. Kingsland, determined that a license or grant from the employee to the employer may be presumed by virtue of the fact that the employee:

was employed by the defendants . . . receiving wages from them by the week; while so employed, he claimed to have invented the improvement patented, and after unsuccessful experiments made a successful one . . . the experiments were made in the [employer’s] foundry, and wholly at their expense, while [the employee] was receiving his wages, which were increased on account of the useful result.[106]

This implied license, or shop right, prevented the assignees of the employee’s patent from bringing suit against the former employer for infringement when all the employer was doing was continuing to use the process it had been using prior to the employee’s departure.[107] The Court reached this result without requiring an assignment of the patent. Rather, the Court held that the plaintiffs took their assignment from the employee subject to his license to his former employer.[108] As such, McClurg is generally thought of as the first shop rights case.[109] It can also be seen as a fairly pro-employee case, for the employer received no rights to the invention other than the right to use it despite the fact that the employee would not have been able to make his improvement but for his employer’s time, money and equipment.[110] During this same time period, other inventors were creating major innovations in both science and industry. For example, in 1844 Samuel Morse—also depicted in Men of Progress—connected the first intercity line between Washington and Baltimore and sent the message, “What hath God wrought!”[111] Moreover, scientists were making significant advances in electric theory, evolution, and thermodynamics.[112] Similarly, in the decade that followed, solo inventors continued to be credited with major breakthroughs in both science and industry. For example, during the 1850s, Heinrich Geissler is credited with the leaps forward that resulted in the invention of the vacuum tube.[113] Further, the 1850s saw huge advances in telecommunications technology, including Morse’s testing of the first transatlantic telegraph cable.[114] Even Abraham Lincoln took the time to consider the discoveries, inventions and improvements that resulted from America’s willingness to observe, reflect and experiment.[115] Despite the attribution of these major advances to individual inventors, the reality was that even in the 1840s, teams of workers were working together to develop new technologies, which “somewhat undermine[s] the heroic mythology surrounding the great individualistic and entrepreneurial American inventors,”[116] The dichotomy between the myth of the sole inventor and the reality that inventive activity was being carried out by groups of individuals in many ways mirrors the shift that was occurring in copyright law at that time. However, instead of acknowledging that multiple individuals were involved in the inventive process, courts turned a blind eye to the significant contributions employees were making to their employers’ inventions and dismissed those contributions as small modifications.[117] In King v. Gedney, for example, the court disregarded the following evidence as “vague and equivocal,” and instead determined that the employer “might have” given the employee—a draftsmen and general foreman—general directions about the improvement:

[The employee] was the foreman, and directed the making the alterations in the factory. [The employer] gave him no directions in relation to those alterations; [another employee] asked him for directions during the time that he was making alternations, and his answer was, ‘I do not know anything about it; go to [the employee].’ He said it was [the employee’s] improvement, and he knew nothing about it; if it did not work it was [the employee’s] fault; [the other employee] must go to him for all instructions. [The other employee] did not hear [the employer] give directions to any of the workmen in the shop. He has seen [the employer] and [the employee] conversing together at the factory, when chalk marks were made upon the floor by [the employee] to convey to [the employer] the manner in which he was going to prosecute the work.[118]

The court determined that if the general idea was the employing inventor’s, then the employee “could only be considered as acting as [his] servant,” and held the employer to be the original and sole inventor.[119] Thus, by crediting employers with being sole inventors, courts were able to continue to articulate the presumption that those who do the actual inventing will own the rights to their inventions. However, as employees became more and more involved with the inventive process, courts would find it increasingly difficult to rely on the employee improvements doctrine to rebut the presumption of employee ownership.[120]

B. The Nineteenth Century Postbellum

American invention did not stop during the American Civil War. In fact, the war spurred increased invention, particularly in agriculture. For example, Cyrus McCormick—another famed inventor depicted in Men of Progress—developed a reaper in 1861, which allowed northerners to increase food production “to the highest levels of surplus ever known” even though they had fewer farm hands due to the war.[121] Similarly, in the mid-1860s, scientists were making major advances in chemistry and thermodynamics.[122] Then, in 1865, George Westinghouse received his first patent for a rotary steam engine, followed by patents for railroad frogs and air brakes.[123] A few years later, a “self-taught practical innovator” by the name of Thomas Alva Edison would apply for his first patent.[124] Edison is a particularly interesting example of a nineteenth century inventor, because it is well documented that he “regularly employed not only mechanics and craftsmen but also college-trained chemists and engineers in his invention workshop,”[125] In some ways, then, it is not surprising that the Supreme Court first recognized the employee improvements doctrine in its 1868 Agawam Co. v. Jordan opinion. In Agawam, an inventor named John Goulding hired a blacksmith, Edward Winslow, to help in his machine shop.[126] Goulding had been working on his invention and was at the point of experimentation when Winslow, having visited another factory, made a suggestion of replacing a piece of Goulding’s invention with a spool and drum.[127]  Goulding went back and forth about the value of the suggestion, but after making another modification to make the change practicable, he adopted Winslow’s suggestion.[128] The new invention was patented and eventually assigned to the plaintiff who then sued the Agawam Woolen Company for infringement.[129] Agawam defended, in part, by arguing that it was Winslow, and not Goulding, that invented the improvements for which the patent was granted.[130] The Court held that Goulding’s claim to invention of the patent was sustained, because all that Agawam could prove was that Winslow suggested the spool and drum—an “auxiliary” part of the invention.[131] In handing another victory to an employer, however, the Court continued to recite that the general presumption is that employees hold rights to their inventions.[132] Similarly, as discussed above, during this time employers in copyright cases began to see the presumption switch from employee ownership to employer ownership. For example, five years prior to Agawam, the influential case of Kenne v. Wheatley was decided in favor of the employer relying on an employee improvements-type argument to grant a theater proprietor the copyrights to a play adapted by her and her theatrical company.[133] Whereas in theater it was undeniable that multiple individuals were required to develop new materials and produce it for the stage, invention in the 1870s continued to be perceived as the work of individuals with minimal help from employees instructed to carryout manual labor. Significant changes, however, were on the horizon for those industries that employed workers to design and construct their products. The “United States began to develop a single price and market system,” which led to capital intensive industries that “employed workers who owned no tools and had few skills, leading to social inequalities that created new political crises,”[134] Those inequities led to the formation of labor unions, which employed a variety of tactics to effect the status of workers in this country. “The decades from the 1870s through the early 20th century saw one disruptive strike after another and the beginning of the appeal of ideologies and reform strategies that would characterize the labor movement—agrarian reform, populism, socialism, anarchism, and progressive reformism.”[135] Despite these changes, inventive activity in the United States soldiered on. For example, Edison—who, as discussed above, employed a variety of workers in his invention workshop[136]—had filed for patents on 21 inventions by 1871.[137] As inventive activity began to involve more complex technology, the proportion of inventions devised by solo inventors continued to decline.[138] Perhaps in recognition of these changes, the Supreme Court began to hint at a willingness to recognize employers’ ownership of patented inventions that they hired employees to create.[139] Nevertheless, there were limits to the Court’s willingness to grant patent rights to one who employed another to invent. In Collar Co. v. Van Dusen, for example, the Supreme Court determined that the plaintiff, despite employing a set of paper-makers to create a paper collar that would be perceived as a real starched linen collar, was not entitled to a patent because only the paper-makers had invented something patentable.[140] The plaintiff had argued that the paper-makers had received inspiration and direction, but the Court determined that the paper-markers had provided the whole improvement and therefore the plaintiff had no rights to their invention under Agawam’s ancillary discoveries rule.[141] Despite its use of the terms “employer” and “employee,” however, the Collar case is better understood as concerning, using our modern parlance, independent contractors, who, even if Collar was a modern copyright case, would be unlikely to fall within the work made for hire doctrine. Rather than inspiring and directing the paper-makers’ improvement, the patentee derived all of the improvements he attempted to patent from the paper-makers. Similarly, other courts in the 1870s held that if an employee’s invention was outside the employee’s inventive arena, the employer would receive no more than a shop right to the invention.[142] By the 1880s, “farmers and industrial workers represented a force for political insurgency that began to coalesce into new movements that threatened to change the nature of society,”[143] These movements reflected a general anti-monopoly and big business sentiment at the turn of the century, as they began to lead to discussions of new government regulations “to control railroad rates and other large enterprises . . . to address the growing power of bankers and railroad magnates,”[144] At the same time, entrepreneur inventors, like Westinghouse, were beginning to build empires of their own from their own companies and inventions, as well as those of others.[145] These companies employed various individuals who developed patentable inventions during their employment. During the remainder of the nineteenth century, disputes over employees’ inventions became more frequent and led to cases that further developed the shop right[146] and employee improvements doctrines.[147] The courts, however, failed to develop a mechanism for dealing with employees that were explicitly hired to invent,[148] until the hired-to-invent doctrine was formally recognized by the Supreme Court in 1924.[149] Also, the rise of the corporate form introduced additional complications,[150] such as whether a shop right could be assigned.[151] Despite these developments, however, in not one of these cases was an employing company considered an “inventor” by virtue of its employees’ activity. Rather, for a company to receive patent ownership, an employee needed to assign it his or her rights. By the turn of the century, each and every person in the United States had been affected by the inventions of the nineteenth century. In particular, the poor were being put to work in factories using newly invented machinery and tools. Even the middle classes indirectly benefited from the benefits of the day:

For the middle classes, such as physicians, attorneys, engineers, academics, journalists, and managers, whose employment was less dependent on machinery and tools than on their education, the new technologies provided new sources of entertainment, comfort, and opportunities. By 1890, a person who worked in such a profession could come home, switch on the electric light, make a phone call, go to the refrigerator and take out a cold beer, light a factory-made cigarette with a safety match, then sit down to listen to a phonograph record while reading a newspaper with news of the day from all over the planet. It was a new world, for none of those experiences had been possible 40 years earlier, even to the wealthy.[152]

IV. Is Patent Law Stuck in the Past?

For inventors in the twentieth century, more and more inventive activity would occur within corporate settings. However, patentable inventive activity would continue to be credited to individuals, not groups, until late into the century. In fact, the average number of inventors per patent would not begin to change dramatically until the 1970s, when a “dramatic increase in the proportion of ‘highly collaborative’ inventions” would drive the average up:[153]

  Yet, despite the fact that today the “ordinary inventor is . . . a joint inventor who invents as part of a team,”[154] patent law continues to require that corporate employers receive assignments from their employees in order to effect a transfer of the ownership of the employees’ patent rights. In Part II, several reasons behind copyright law’s move toward the work made for hire doctrine were posited: (1) courts were attempting to reflect the intent of the parties; (2) courts felt that employees were in the best position to determine ex ante whether they intended to assert copyright ownership; (3) changing notions of the nature of authorship; (4) ease of statutory drafting; (5) avoiding constitutional concerns; and (6) concern about renewal terms. Three of these reasons simply do not apply to patent law. First, patents in the United States do not have renewal terms. Second, the copyright statute was being revised in 1909 for reasons beyond the codification of the work made for hire doctrine. The patent statute, by contrast, was not significantly revised until 1952. Without an independent push to redraft the Patent Act, the need to ease statutory drafting was absent, and constitutional concerns in granting patent rights to non-“inventors” would not need to be addressed. As for the remaining reasons for the shift in copyright law, patent law simply did not share those factors. First, the perceived nature of inventive activity in the nineteenth century did not lend itself to an argument for the need of a doctrine like work made for hire. As discussed in Part II, copyrighted works grew in complexity over the course of the nineteenth century. While copyrighted works were originally only embodied in single author books, they steadily grew to include complicated works involving the input of the multiple creative individuals. In particular, the creation of maps and theatrical works required coordination among multiple parties to insure that works could be exploited. Once the 1909 Act was passed, that complication grew exponentially as additional works were brought under the copyright statute. The perception of inventive activity in the nineteenth century, on the other hand, did not appear to require significant coordination among parties. Rather, credit for advances was given to individual inventors, whether or not they, like Edison, employed others in their invention workshops. Moreover, even in the nineteenth century, most patents were on incremental improvements of existing technologies, which made it even more likely that the number of individuals involved in each inventive push forward would be small. Further, even if an invention required multiple individuals to reduce it to practice, the individual credited with the invention could subsume any other inventive activity into his invention under the employee improvements doctrine.  In fact, in 1916, three-quarters of U.S. patents were issued to individuals.[155] Thu­­s, the need to coordinate among multiple individuals that existed in copyright law in the nineteenth century did not exist in patent law.[156] Second, the existing doctrines that had developed, including the hired-to-invent doctrine, provided sufficient ex ante protection to employers. The shop rights doctrine, which was well established in the early part of the century, provided employers with a way of ensuring that their expectations were met. Unlike today where many organizations that own patent rights merely license those rights to others to use, in the nineteenth century, most companies used patent rights merely to exclude competitors while they themselves developed products to sell to consumers. Without the need to license, companies merely needed the right to use, which the shop rights doctrine provided. Moreover, as described above, when an inventor did use assistance in the process of developing his invention, the employee improvements doctrine protected him from needing to secure rights from each assistant in his workshop. As long as the inventor couldn’t be seen as deriving his invention from someone else, he could be fairly content that his rights in the invention would be secure. Finally, by the end of the nineteenth century, employers were securing contracts with their employees that included provisions requiring the assignment of any patented inventions later developed.[157] In 1885, only 12% of U.S. patents were issued to corporations.[158] By 1950, it had grown to 75%, and by the 1980s, 84% of U.S. patents were assigned to corporations, “usually the employer of the actual inventor,”[159] In the past decade, those efforts have become increasingly successful.[160]

  In fact, it will come as no surprise that the Intellectual Property Owners Association’s list of top patent owners does not include the name of a single individual.[161] Further, by the turn of the century, when an employer hired an inventor, the hired-to-invent doctrine would require the inventor to assign their rights to the employer. Despite the protection of the existing patent law doctrines that give employers some of the benefits of the work made for hire doctrine, it is clear from Table 5 that not one doctrine captures all of the benefits from that doctrine:[162] Moreover, as recent cases make clear, companies cannot rely on the patent law doctrines and contracts with their employees to ensure ownership of the rights to their employees’ inventions.[163] For example, the Supreme Court recently held that although an inventor working as a research fellow at Stanford University signed a Copyright and Patent Agreement that stated that he “‘agree[d] to assign’ to Stanford his ‘right, title and interest in’ inventions resulting from his employment at the University” and written assignments of rights, which Stanford used to file several patent applications, Stanford did not receive the inventor’s patent rights because he signed a Visitor’s Confidentiality Agreement that provided that he “‘will assign and do[es] hereby assign’ to Cetus his ‘right, title and interest in each of the ideas, inventions and improvements’ made ‘as a consequence of [his] access’ to Cetus” between signing the CPA and the assignments.[164] Further, many modern employers intend to license, rather than use, their employees’ inventions, which a shop right would not permit. Similarly, in Abbott Point of Care Inc. v. Epocal Inc., Abbott filed a patent infringement suit against Epocal, a competitor in the diagnostic field, alleging infringement of patents that “cover systems and devices for testing blood samples,”[165] Epocal, however, was founded by Dr. Imants Lauks, the named inventor of the patents at issue and a former employee of Abbott’s predecessors.[166] Despite the fact that Dr. Lauks had entered into two employment agreements, one of which contained a provision requiring him to assign all of his rights to any inventions that he might conceive, the Federal Circuit determined that Abbot did not hold rights to the patents and, thus, lacked standing to bring the lawsuit.[167] The court determined that Dr. Lauks had resigned from his previous employment in 1999, and that while the consulting agreement that he entered into after his resignation stated that various provisions of Dr. Lauks’ employment agreement remained in effect, it did not address invention assignments or obligations.[168] Thus, the two patent applications Dr. Lauks filed after entering into the consulting agreement did not belong to Abbot.[169] Additionally, despite the high assignment rate discussed above, when a patent assignment is found to be ineffective, the consequences can be extreme. This is particularly true where the employer acts on its belief that an assignment is effective and incurs costs associated with the patented technology, only to learn that they do not, in fact, hold any patent rights. As such, it is time for patent law to come to terms with the fact that it is stuck in the past and adopt an inventions made for hire doctrine. To do so, Congress need only pass an act amending two sections of the Patent Act.  First, Congress would add a subsection (c) to 35 U.S.C. § 111 as follows:

(c) In the case of an invention made for hire, the employer is considered the inventor for purposes of this title, and, unless the parties have expressly agreed otherwise in a written instrument signed by them, owns all patents rights to the invention.

Second, Congress would add a subsection (f) to 35 U.S.C. § 100 as follows:

(f) An “invention made for hire” is an invention invented or discovered by an employee within the scope of his or her employment.

Whether an individual is an employee for purposes of the invention made for hire doctrine will be determined by looking to the same common law doctrine of agency that the Supreme Court identified in Community for Creative Non-Violence v. Reid.[170] Moreover, there is over two decades of caselaw interpreting agency doctrine in light of the Copyright Act and over a century of caselaw interpreting the work made for hire doctrine for courts to rely on in implementing this new provision.[171]

Conclusion

As this article has suggested, patent law remains stuck in the past by continuing to rely on the rhetorical device that modern invention is performed by individual inventors in their garages, when, in fact, most patentable inventive activity occurs in corporate and university settings. Moreover, contrary to popular perception, most individuals who would be labeled “inventors” in the twenty-first century are employees of a corporate entity and assign their patent rights to their employer. Nevertheless, patent law continues to require written assignments to give employers ownership of their employees’ inventions. While that formulation may have made sense when invention was credited to individuals and required little coordination among inventors, it no longer does. Rather, employers should be able to feel secure in the knowledge that any inventions invented or discovered by their employees within the scope of their employment are owned by the party that financed and supported them (i.e., their employers). As such, it is time for the Patent Act to be amended to borrow from the Copyright Act and adopt an inventions made for hire doctrine.

---

Aaron X. Fellmeth* A pdf version of this article may be downloaded here.  

Introduction

It is an axiom of patent law that utility patents protect “inventions” only; it is a paradox of patent law that the Patent Act does not define “invention” except tautologically: “The term ‘invention ’ means invention or discovery.”[1] Given that patents are designed to reward inventors for the disclosure of their inventions,[2] it would have been helpful had the Act settled just what constitutes an invention. Alas, it does not. After 220 years of progressively developed patent statutes and jurisprudence, the concept of invention, so central and fundamental to patent law, has become one of the most misunderstood. The Patent Act’s obscurity on this key point of law does not prevent the Act from employing the term “invention” ubiquitously. A person is entitled to a patent unless “the invention was known or used by others in this country” or the inventor “has abandoned the invention.”[3] Patentability may not be negated “by the manner in which the invention was made.”[4] A person infringes the patent by making or selling the “patented invention.”[5] Among the mysteries emanating from uncertainty about what kinds of technologies qualify as a patentable invention, identifying the proper inventors has developed into one of the most intractable. Because the U.S. Constitution allows only “inventors” to obtain patents on their inventions,[6] we must know what an invention is before we can decide who qualifies as its inventor. The inventor himself has a self-evident interest in certainty about his or her entitlement to a patent as well. In 2011, Congress supplemented the definitions section of the Patent Act through the Leahy-Smith America Invents Act (“AIA”), to take effect in March 2013, which defines the term “inventor” for the first time. However, that definition merely adds another point of confusion; it provides, “The term ‘inventor’ means the indiv­idual or, if a joint invention, the individuals collectively who invented or dis­cover­ed the subject matter of the invention.”[7] A definition stating that an inventor is the person who invents the invention would offer us nothing more than another tautology. However, the new amendment does not exactly do that; instead, it defines as the “inventor” the person who invents the “subject matter” of the invention.[8] It is not immediately clear how the subject matter of an invention differs meaningfully from the invention itself. Instead of resolving a difficult and core problem of law, the new definition merely adds a point of obscurity. The intuitive way to begin resolving this paradox is to define as an “invention” whatever satisfies the legal requirements for the grant of a utility patent. The Patent Act requires that, to qualify for a utility patent, an invention must be “novel,” “non-obvious” at the time of conception (after the AIA, the “effective filing date”), and “useful.”[9] But the Act and its implementing regulations also require the fulfillment of several more conditions for a patent to issue, including a written disclosure of the invention, an oath or declaration by the applicant, and the payment of a fee.[10] Surely, not all of these requirements speak to the definition of an invention. We can hardly say that a person has failed to invent something until he has signed an oath and paid a fee to the U.S. Patent and Trademark Office.[11] At the application stage, the doctrinal analysis for invention and inventorship is deceptively straightforward. If the application satisfies the requirements of the Patent Act and regulations, the invention will be considered whatever is described in the application, and the question of inventorship will be resolved by presumption. The PTO will not examine inventorship, but will usually accept without question the claims of the applicants that they originated the invention.[12] The trouble typically arises when the patent is enforced or challenged in litigation.[13] In the disputing context, the definition of the invention does not normally assume a pivotal role because the focus of litigation tends toward the “claims.” Claims are statements in the patent “particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.”[14] To infringe a utility patent, the accused infringer must generally have made, used, sold, or imported a product described in one or more of the patent’s claims. If the defendant challenges the validity of the patent, the strategy is to prove the invalidity of all allegedly infringed claims. Courts have similarly gravitated toward claims analysis to determine whether a member of a research team qualifies as an inventor. As consistently invoked by both the Supreme Court and the U.S. Court of Appeals for the Federal Circuit, the accepted patent law doctrine treats claims as if they were equivalent to the invention itself.[15] Yet, in inventorship disputes, the case for fixating on the claims is based more on pedigree than logic or policy. Claims are nearly always drafted by lawyers or patent agents after (sometimes years after) the invention was completed.[16] The claims may or may not express what the inventor conceived; they rarely express the full scope of the invention, and not infrequently capture more than what the inventor invented.[17] Despite the convenience of equating the claims with invention in infringement analysis, judicial opinions equating the two have propelled patent law down a treacherous path. The consequences of this wayward doctrine reach into the very processes of innovation. Most modern technology research is the product of collaborations between multiple scientists and engineers, usually in large research institutions.[18] Collaborations frequently involve teams from industry and universities or government agencies joining forces to advance technology despite divergent economic interests. Joint inventorship has accordingly come to assume a critical role in technological research and development. Approximately half of all patents now list more than one inventor.[19] As Congress has long recognized,[20] the patent system best achieves its goal of encouraging technological development by facilitating rather than hindering joint invention. To achieve this goal, inventorship doctrine needs to define its criteria with sufficient specificity to allow research teams to determine who qualifies as an inventor with some confidence. To achieve maximal efficiency and fairness, those criteria should be neither overinclusive nor underinclusive. All collaborators who contribute significantly to the conception of the invention, and only those collaborators, should qualify as inventors. The consequences for a bona fide inventor who is excluded from a patent application covering an invention to which he contributed significantly can be devastating for the omitted inventor. The financial rewards of being named an inventor may be substantial, indeed monumental, because the inventors are the presumptive owners of the patent to the invention as tenants in common.[21] In the absence of an agreement to the contrary, a joint inventor is entitled to exploit the patented invention commercially without the consent of, or owing a duty of contribution or accounting to, the other inventors.[22] In other words, exclusion from joint inventorship means that the putative inventor may lose all of the financial rewards of the invention, no matter how great that inventor’s investment of time, resources, or creativity in the invention.[23] Recognition as a joint inventor means not only a potential share of the economic rewards, but it may have career-altering effects. A scientist or engineer may invest months or years in a research project that ultimately results in a patent. He may expend institutional funds, exhaust grants, and stake his or her reputation on the success of the project. Frequently, hiring, promotion, or tenure depends on successful completion of the project. Exclusion from the inventorship on the resulting patent denies the researcher not only his or her just financial rewards, but career recognition that may dictate his future in academia or industry. Fixating on claims will in many cases have no effect on inventorship determinations. Frequently, a joint inventor’s contribution, if it appears anywhere in the issued patent, will be claimed expressly. But relying exclusively on claims to determine inventorship is like solving three adjacent sides of a Rubik’s Cube. From some viewpoints, the puzzle appears fully solved. But the apparent order masks a fundamental incoherence evident from other viewpoints. The full puzzle cannot be solved without twisting the ostensibly finished sides back into disorder. Inventorship doctrine suffers from the same fundamental incoherence. Fixating on the claims makes inventorship doctrine seem sound when the inventor’s contribution is recited expressis verbis in one or more claims. Other inventive contributions fall within the confused jumble of fragmentary judicial statements on the doctrine of invention. For reasons to be explained, there is no necessary correlation between the relative importance of an inventive contribution, or the effort or creativity involved in that contribution, and its expression in the claims. To deny such collaborators recognition as inventors is unfair to them, confuses basic patent doctrine, and undermines the patent law’s incentive structure. The purpose of this article is to disaggregate the patent claims from the invention and to restore the integrity of the concepts of invention and inventorship. This restoration will require abandoning the time-honored judicial fixation on claims. Giles Rich, frequently cited by reverent courts and patent bar alike,[24] famously said, “[T]he name of the game is the claim.”[25] That phrase reflects Rich’s zealous opposition to the concept of invention[26] as well as the key role of claims in limiting the patent monopoly. But the influence of his maxim has far surpassed its meager truth. As Oliver Wendell Holmes once observed, “It is one of the misfortunes of the law that ideas become encysted in phrases and thereafter for a long time cease to provoke further analysis.”[27] Claims are operationally important to patent law, but inventors do not invent claims, they invent inventions. And, to coin a rather different maxim: The prevention of any mention of invention causes misapprehension. This article will explore the causes of the confusion; its consequences for inventors; and its resulting effect on U.S. technological development. Part II of this article begins by summarizing the law governing invention and inventorship, and its development and interpretation by the Supreme Court and Federal Circuit. Joint inventorship receives special attention in this part because the case law in this area reveals especially well how fixating on the patent claims distorts patent law analysis more generally. Part II then relates how an idiosyncrasy in the standard of proof in patent litigation has reinforced distorting biases in inventorship analysis. Part III maps out how the broken link between inventorship doctrine and the public policy it is intended to promote can be repaired. Specifically, Part III.A analyzes the legislative history of the Patent Act provisions on joint inventorship to determine congressional intent and compares the current joint inventorship doctrine’s misdirected fixation on claims to the policy goals sought by Congress and discusses the policy consequences. Part III.B then evaluates how the question of inventorship affects the role of the patent system in technological development generally and proposes an interpretation of the Patent Act that better fulfills the goals of U.S. technology policy. Part IV concludes by summarizing the argument.

I. The Law of Invention and Inventorship

A. The Patent Law Concept of Invention

Although the 1952 Patent Act uses the term “invention” throughout, the closest the Act comes to defining the term nontautologically is in Section 101: “Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.”[28] This is not technically a definition, but it gives clues to one by specifying that an invention is a process or physical product that is new and useful.[29] The term “new” is almost universally interpreted as a term of art in patent law (equivalent to “novel”)[30], further defined in Section 102.[31] The term “useful” is not defined in the Act but has been construed in the case law to mean that the invention offers some immediate and practical benefit to the public.[32] While Sections 100 and 101 provides a starting point for defining invention, they are incomplete. As suggested in Part I, there are other requirements for patentability. Procedural requirements such as the inventor’s oath or the application fee obviously do not figure in the definition of the invention itself. But there are other substantive requirements for an invention to qualify for patent protection, the most important of which is nonobviousness. Despite its omission from Section 101 of the Patent Act, nonobviousness is in some ways the keystone of the definition of invention. The Patent Act, as amended by the America Invents Act, defines an invention as obvious:

if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. . . .[33]

The nonobviousness requirement lies so close to the heart of the definition of an invention because its purpose is to clarify the historically confused, and sometimes contradictory, judicial interpretation of the constitutional requirement of “invention.”[34] A new technology that would have been obvious to an ordinary scientist or engineer at the critical date may qualify as an innovation in the colloquial sense, but it is not an invention in the patent law sense.[35] A statutory analysis of the concept of invention might be expected to end here. Other requirements for obtaining a patent are usually considered procedural rather than substantive. The Patent Act itself classifies the novelty, utility, and nonobviousness requirements in a chapter of the Act (“Patentability of Inventions”) separate from the chapter specifying the procedural requirements (Chapter 11—“Application for Patent”). Yet, the definition of invention remains incomplete because nothing in Sections 101 to 103 obligates the inventor to understand and appreciate the significance of his own discovery.[36] An unrecognized product or process is not yet an invention at all in the patent law sense.[37] The relevant section of the Patent Act, Section 112, is found among the procedural provisions of Chapter 11. Section 112 codifies the disclosure requirement.[38] Specifically, it requires that every patent application include a “specification” section that sets forth a full written description of the invention, “and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use” that invention.[39] Section 112 also includes the requirement that the application terminate with “claims.”[40] Between the written description and the claims, the specification is supposed to alert the public to the outward bounds of the invention that will be protected by patent, and to teach the public how to make and use the claimed part of the invention. For this reason, Section 112 is commonly thought to embody the fundamental bargain of the patent system: the inventor discloses the invention to the public in exchange for a temporary legal monopoly over making, using, selling, offering to sell, or importing the invention.[41] To that extent, the enablement provisions of Section 112 are indeed procedural and not substantive. They establish a disclosure requirement that the applicant must fulfill in order to obtain a patent, but this requirement assumes that the inventor sufficiently appreciates what he has invented and how to reproduce it. In other words, an applicant who can fulfill the requirement to describe clearly how to make and use the invention proves that he understands the invention sufficiently, and that he “actually invented the invention claimed.”[42]

1. The Role of Conception

Section 112 thus implies an uncodified but essential aspect of the definition of an invention. Invention has not occurred until the inventor appreciates that he or she has invented something useful, functioning, worthwhile. This subjective component of invention is inherent in the idea of “conception,” which is integral to establishing the moment when the invention became entitled to a patent. Conception is considered the “touchstone” of invention,[43] and has been defined as “the complete performance of the mental part of the inventive act” and “the formation in the mind of the inventor of a definite and permanent idea of the complete and operative invention, as it is hereafter to be applied in practice.”[44] Conception is incomplete unless the idea of the invention “is so clearly defined in the inventor’s mind that only ordinary skill would be necessary to reduce the invention to practice, without extensive research or experimentation.”[45] In other words, even if the inventor has a complete mental picture of the invention’s physical configuration, the invention is not “conceived” in patent law until the inventor has also determined how to make and use the invention in practice.[46] Even when conception no longer defines the date at which the inventor is entitled to a patent, conception will still be necessary, because without it the applicant will lack the information necessary to complete the Section 112 disclosure. Therefore, while Section 112 is indeed a procedural provision, one of its other functions is to provide evidence of whether the application’s subject matter satisfies the substantive requirements of invention. As the Federal Circuit put it: “The conception analysis necessarily turns on the inventor’s ability to describe his invention with particularity. Until he can do so, he cannot prove possession of the complete mental picture of the invention.”[47] The relationship between the inventor’s understanding and the physical configuration of the invention can be illustrated through an example:

{1a}  Ira has processed several chemicals to create a novel and nonobvious compound through a complex series of steps. Ira believes the compound could prove to cure cancer, because its molecular structure is loosely related to the structures of other chemicals that have anticancer effects.

In example {1a}, Ira may be capable of fully describing how to make and use his new molecule, and so he can satisfy the procedural requisites of Section 112. However, if Ira’s belief in the molecule’s anticancer properties lacks any firm basis in science, the molecule does not satisfy the Section 101 utility requirement. Although under current law an inventor need not know with certainty or even probability that the invention will work for conception to be considered complete,[48] a mere wish or unsupported expectation does not suffice. In Hitzeman v. Rutter, the Federal Circuit held that a genetically-altered yeast intended to produce antigens having certain key size and sedimentation rates had not yet been conceived by the inventor, because the inventor had only a “hope,” not a reasonable expectation, that the yeast would perform in that manner.[49] In Scenario {1a}, then, unless Ira has some scientifically sound basis, however partial, for attributing anti-cancer properties to his molecule, there is not (yet) a patentable invention. Consider a slight alteration of example {1a}:

{1b}  On January 30, Ilsa processes several chemicals to create a novel and nonobvious compound through a complex series of steps. At the time, she has no idea whether the chemical has any effective use. On June 1, she discovers that the new chemical is extremely effective at treating cancer. Ilsa files a patent application on the chemical.

Ilsa has clearly created the precursor to a new cancer-fighting drug, but when was it invented? On January 30 or June 1? Example {1b} illustrates the subjective requirement of invention. A new product or process may meet the objective statutory patentability requirements (novelty, nonobviousness, and utility), but it is not a patentable “invention” until its creator appreciates the utility of the invention and understands how to reproduce it.[50] As a correlate, there have been cases in which the inventor created a new chemical substance unwittingly; the substance was not deemed to have been conceived until the inventor perceived its presence,[51] because conception has not occurred until the inventor gains the ability to describe the invention.[52] In Example {1b}, conception occurred on June 1. Another variation illustrates a concomitant principle:

{1c} On January 30, Ian conceives of a chemical compound that he has sound scientific reason to believe will be extremely effective at treating cancer. However, he has no expertise in synthesizing this class of chemicals. On June 1, Rowena assists Ian to synthesize the compound using new insights and techniques developed specifically for the purpose.

Example {1c} is in some ways the reverse of {1b}. In {1b}, the first inventor could visualize and make the compound but did not know how to use it. In {1c}, the first inventor could visualize and knew how to use the compound, but did not know how to make it. The outcomes are equivalent; in both cases, the compound is not “invented” until June 1; in scenario {1c}, Ian and Rowena are joint inventors.[53] The foregoing discussion, like any doctrinal restatement, gives an illusion of certainty about the law relating to invention. In practice, gray areas in the subjective prong of conception, coupled with frequent misunderstanding of the inventive process, plague the law and its application with inconsistencies. The explanation of example {1b} relies on the PTO’s or court’s ability to distinguish between sufficient and insufficient bases for attributing utility to the invention. From a scientific or engineering perspective, the question of whether an untested product or process will function as intended is not a “yes” or “no” one, but a sliding scale of probability. The Federal Circuit has adopted a murky test to determine where conception occurs on the scale between mere surmise and certainty of adequate functioning. For example, on one hand, in order to shift the burden of proof to the applicant, a PTO examiner challenging utility must produce evidence that a person having ordinary skill in the relevant art would reasonably doubt the alleged utility.[54] On the other hand, an applicant or patentee asserting the utility of a patented invention must produce at least some evidence, however inconclusive, that the invention works as intended.[55] This leaves the PTO, which bears the burden of proving an absence of utility,[56] without much guidance. To be safe, the PTO has effectively adopted the practice of virtually never rejecting pharmaceutical inventions on utility grounds unless the allegation of utility is incredible.[57] This is partly because the PTO does not test the utility of an invention disclosed in a patent application and so must rely primarily on the applicant’s representations and the (usually very general) background knowledge of the examiner. Such an “incredibility” test is inappropriate in judicial pro­ceedings, where the party challenging the patent may introduce evidence of inutility that was unavailable to a PTO examiner. Yet, the Federal Circuit and its predecessor have endorsed an exceedingly low standard of evidence to uphold an allegation of utility, especially in the active field of pharmaceutical research.[58] Congealing the uncertainty is the jurisprudence, discussed in Part II.C, creating a strong legal presumption of validity of the patent despite a complete or nearly complete vacuum of information about whether the invention was in fact useful at the time the patent was examined and issued. Another difficult doctrinal issue is raised in example {1b}. Suppose Ilsa did not discover the utility of the chemical; a new researcher, Ron, did. Does this not make Ron the inventor rather than Ilsa? Or does Ilsa remain the inventor? Or might they be considered joint inventors? From a logical and strictly doctrinal standpoint, the fact that conception was incomplete until Ron discovered the utility of the chemical implies that Ron is the sole inventor if he did not collaborate with Ilsa in any way, and a joint inventor with Ilsa if they did collaborate and Ron’s discovery was more than a straightforward application of well-known principles of chemistry. However, as discussed in Part II.C, Federal Circuit and Supreme Court precedents have created a basis for doubting whether courts would actually endorse this outcome.

2. The Role of the Claims

The problematic patent law conception of invention arises in part from the role of the claims in defining the invention. The claims, as noted above, are the part of the patent document that delimits the outer boundaries of the invention for which the applicant seeks protection. They are typically one sentence long and terse, if not cryptic, in their phrasing. Partly because the claims are drafted by patent lawyers or agents rather than the inventor, and partly due to the organic idiosyncrasies of language, the claims may or may not encompass all of the invention. In fact, they may encompass quite a bit more than the invention.[59] A claim may encompass subject matter that the inventor never conceived and never could conceive. There is no a priori reason why the claims will completely over­lap with the invention. An inventor who claims less than the invention may lose some of the benefit of the invention. There are several reasons why this might happen. One is poor drafting, although the consequences of poor drafting to the hapless patentee are mitigated by the doctrine of equivalents and the possibility of broadening reissue proceedings.[60] Another possible reason is that the inventor wishes to commercially exploit only a subset of an invention’s embodiments and to dedicate the remainder to the public.[61] In that case, patent coverage narrower than the invention as a whole poses no policy problems, because it is voluntary. In contrast, an inventor who claims more than the invention risks the overbroad claims being invalidated.[62] When the claims exceed what was enabled in the written description, they fail to honor the basic patent bargain of dis­closure of the invention in exchange for a legal monopoly over the claimed subject matter.[63] Patents with overbroad claims also fail to ensure that the inventor possessed the claimed subject matter, because there is no way to verify that the applicant did indeed invent what he cannot or will not describe. Indeed, such patents may suffer from a constitutional deficiency on the theory that Congress has no power to grant patents on subject matter the inventor did not invent.[64] Probably the most famous example of an inventor who tried to claim far more than he invented relates to the Morse telegraph.[65] Morse, recognized as the inventor of the electro-magnetic telegraph, obtained a reissue patent with eight claims. The first seven claimed various uncontroversial components and embodiments of the electric telegraph. The eighth claim, however, covered much broader subject matter:

I do not propose to limit myself to the specific machinery or parts of machinery described in the foregoing specifications and claims; the essence of my invention being the use of the motive power of the electric or galvanic current, which I call electro-magnetism, however developed for marking or printing intelligible characters, signs, or letters, at any distances, being a new application of that power of which I claim to be the first inventor or discoverer.[66]

The Supreme Court struck the claim as overbroad, because it covered subject matter that Morse had not described in his patent “and indeed had not invented.”[67]

Professor Morse has not discovered, that the electric or galvanic current will always print at a distance, no matter what may be the form of the machinery or mechanical contrivances through which it passes. You may use electro-magnetism as a motive power, and yet not produce the described effect, that is, print at a distance intelligible marks or signs. To produce that effect, it must be combined with, and passed through, and operate upon, certain complicated and delicate machinery, adjusted and arranged upon philosophical principles, and prepared by the highest mechanical skill. And it is the high praise of Professor Morse, that he has been able, by a new combination of known powers, of which electro-magnetism is one, to discover a method by which intelligible marks or signs may be printed at a distance. And for the method or process thus discovered, he is entitled to a patent. But he has not discovered that the electro-magnetic current, used as motive power, in any other method, and with any other combination, will do as well.[68]

If the claim had been upheld, Morse could prevent another inventor who had developed an entirely different and potentially better means of using electricity to transmit information for printing at a distance (for instance, by wireless radio frequency transmitter and receiver) from practicing his patent. The Court accordingly rejected the claim. The Morse case illustrates the interdependence of the Section 112 disclosure requirement and the concept of invention. The Court’s reasoning implied that the disclosure requirement does not only alert the public to the parameters of the protected intellectual property and teach the public how to make and use the invention, but tailors the scope of patent protection to what the inventor actually contributed to the technological field.[69] When the applicant claims more than he invented, the benefits of the patent to the public are attenuated and the possibility of a dysfunctional patent precedent arises. Patent prosecutors are typically more concerned about underclaiming than overclaiming. The patent examiner frequently whittles down an overly aggressive claim, and those exaggerated claims that survive prosecution will benefit from a strong presumption of validity, which gives the patentee useful leverage for negotiating with licensees and competitors, even if the claims are not strictly enforceable under a proper interpretation of the claim. At worst, the patentee can obtain a nar­row­ing reissue a few years down the road to ensure the claim’s enforce­ability.[70] An overly narrow claim, in contrast, risks dedicating the un­claimed subject matter to the public for all time. It is possible to seek a broadening reissue, but only within two years after the patent is granted.[71] As a result, many patent pro­secutors tend toward aggressively claiming more than the inventor actually con­ceived, so long as the overclaiming will not court rejection by the PTO examiner. The claims are important, but as the Federal Circuit has emphasized en banc, not all-important: “The claims . . . do not stand alone. Rather, they are part of ‘a fully integrated written instrument’” (i.e., the patent).[72] Yet, when not engaged in the narrow task of claim construction, courts have sometimes fetishized the claims as if they were the only legally material part of the patent. The importance of the claims in determining whether anticipation by prior art or patent infringe­ment has occurred has led many courts, including the Supreme Court, to use language that conflates the claims with the invention itself: the claims, they assert, “define the invention.”[73] The Federal Circuit has stated that a researcher must “conceive” of a “claim” (rather than an invention) to be an inventor,[74] and suggested that Section 116 of the Patent Act requires a coinventor to “make a contribution to the conception of the subject matter of a claim.”[75] The Patent Act says no such thing; it formerly referred to inventions and “subject matter” rather than claims, and now refers to these and to the “claimed invention” as well. Claims may or may not coincide with the full scope of the invention or even the “claimed invention.” Oskar Liivak has defined the invention as “the set of all of the embodiments that the inventor has invented. That is, the invention is the set of embodiments conceived by the inventor in enough detail so that each of those embodiments is capable of being reduced to practice.”[76] This definition, while incomplete, has the virtue of indicating how the invention is (or at least should be) more expansive than the claims, which are “written attempts to circumscribe varying subsets of the embodiments disclosed in the specification.”[77] Liivak’s point, that the claims (and indeed the “claimed in­ven­tion” of Section 100(j)) are a subset of the invention rather than the invention itself, is well taken. Courts have on a few occasions distinguished explicitly between the aspects of the invention protected by patent and the invention as a whole. For example, in a moment of clarity, a panel of the Federal Circuit observed that “the specification teaches an invention, whereas the claims define the right to exclude.”[78] The Supreme Court has also distinguished between the two, stating that claims are “construed . . . with a view to ascertaining the in­ven­tion.”[79] The necessity of distinguishing between the claims and the invention to which the claims pertain is also evident from the Patent Cooperation Treaty (PCT), to which the United States is a party.[80] The PCT allows inventors to file a single patent application that may result in patents in multiple countries. Under the “unity of invention” requirement, the patent application must relate to just one “invention” or a “group of inventions so linked as to form a single general inventive concept.”[81] Yet, in each state granting a patent resulting from the PCT application, the claims may differ to comply with local filing requirements.[82] A single invention could not receive the same patent protection in different countries using different claims unless the invention is something larger than the claims. The frequent judicial and scholarly statements that claims “define” the invention, then, are misleading.[83] The invention is more than the claims; the claims limit the invention but they do not alone define it. “[I]t is fundamental that claims are to be construed in the light of the specifications and both are to be read with a view to ascertaining the invention.”[84] The written description portion of the specification “‘is always highly relevant to the claim construction analysis. Usually, it is dispositive; it is the single best guide to the meaning of a disputed term.’”[85] If the claims must always be read in light of the specification, the claims cannot “define” the invention; they merely limit the extent to which the patent protects the invention. The Leahy-Smith America Invents Act, the relevant provisions of which take effect on March 16, 2013, suggests that Congress shares this perception of the distinction between the claims and invention. The Section 102 amendments substitute the term “invention” in the novelty provisions of the Patent Act with the term “claimed invention,”[86] which is in turn defined as “the subject matter defined by a claim in a patent or an application for a patent.”[87] From this, it appears that Congress considers the “invention” to be different from the “claimed invention.”[88] Because the patent protects only the claimed portion of an invention from infringement, it is possible to infringe on a “claim,” but not an “invention.” By extension,[89] prior art may anticipate specific claims in a patent application without rendering the invention unpatentable. Section 102 now more clearly suggests that novelty analysis consists in determining whether specific claims are anticipated by prior art, as opposed to analyzing whether the “invention” as a whole is anticipated by prior art, because each claim stands on its own. It follows that a person who infringes any one claim infringes the patent, even if the accused product or process does not infringe any other claim. For both anticipation and infringement purposes, then, claims do not “define” the invention in the conventional sense; they define the rough outlines of what part of the invention can overcome anticipatory prior art and accordingly what part of the invention the patent protects from infringement. This key distinction between the invention and the claims has unappreciated importance for determinations of inventorship, and especially of joint inventorship.

B. Inventorship and Joint Inventorship

1. The Evolved Doctrine

When a sole researcher conceives the entirety of an invention, she must satisfy all of the patentability requirements by herself. When the inventor fulfills these requirements, abstruse doctrinal or policy problems arise only in exceptional cases. But the image of the lone scientist confined to a laboratory for months, and ultimately emerging with a triumphant new invention has more romantic appeal than correspondence with reality. Although much government rhetoric exalts the independent inventor,[90] in modern industry, discoveries in applied science are usually a product of teamwork. Joint invention is now more common than sole invention.[91] In general, a joint invention “is the product of a collaboration between two or more persons working together to solve the problem addressed.”[92] Yet, the process of identifying whether a collaborator qualifies as a joint inventor was until very recently crude and uncertain, and even today the law remains under­developed. Before 1984, the Patent Act’s provisions on joint inventorship were entirely procedural; they offered no guidance to the judiciary on how joint inventorship should be understood. The Act merely stated: “When an invention is made by two or more persons jointly, they shall apply for patent jointly and each sign the application and make the required oath, except as otherwise provided in this title.”[93] It then went on to explain what procedural measures could be taken if a joint inventor refused to join in the application, as well as the consequences of a non-inventor being wrongly listed as an inventor (commonly known as “misjoinder”) or an inventor being erroneously omitted (“nonjoinder”).[94] For many years, courts struggled with the question of how to determine who was a joint inventor. Regarding the threshold of contribution required to qualify as a joint inventor, their formulated tests were frequently vague and sometimes contradictory, leading one district court judge to lament:

The exact parameters of what constitutes joint inventorship are quite difficult to define. It is one of the muddiest concepts in the muddy metaphysics of the patent law. . . . Perhaps one need not be able to point to a specific component as one’s sole idea, but one must be able to say that without his contribution to the final conception, it would have been less—less efficient, less simple, less economical, less something of benefit. . . . . . . . . . . . This situation does make it difficult to say . . . with real certainty, whether or not a given person “is” a joint inventor in a given case.[95]

Such guidance was indeed lean. “Less something of benefit” is rather overinclusive as a standard; it would reward routine, and indeed flatly obvious, contributions with coinventorship. Courts were plainly struggling with the doctrine. Another judge wrote:

The conception of the entire device may be due to one [inventor], but if the other makes suggestions of practical value, which assisted in working out the main idea and making it operative, or contributes an independent part of the entire invention, which is united with the parts produced by the other and creates the whole, he is a joint inventor, even though his contribution be of comparatively minor importance and merely the application of an old idea.[96]

These pronouncements were problematic not only because of their lack of guidance. The mere contribution of “an old idea” of “minor importance” is probably too low a standard, and has been at least called into doubt by Federal Circuit precedent.[97] If a joint inventor must contribute something nonobvious, then the nature of the contribution determines joint inventorship, not merely the effect of the contribution on the invention’s functioning. It would seem that a collaborator who contributed nothing more than information or designs well known in the art would not qualify as an inventor unless the application of that prior art to the problem at hand would have been nonobvious.[98] A decided improvement in the functioning of the invention is relevant to assessing the quality of the contribution but far from dispositive. While the historical standard of joint inventorship might have seemed low, in some ways it compensated for an aberration of patent law that discouraged team research. In case after case, federal courts denied joint inventors the right to file a single patent with multiple claims on the product of their collaboration, if all inventors did not contribute to each and every claim.[99] In other words, this “all claims rule” excluded from inventorship any collaborator who did not contribute to every claim of the patent. As a leading treatise of the time put it: “Only where the same single, unitary idea of means is the product of two or more minds, working pari passu, and in communication with each other, is the conception truly joint and the result a joint invention.”[100] One nineteenth century court even held that when two researchers invent distinct parts of the same new machine, they should obtain separate patents as sole inventors on each part of the machine rather than a single patent as coinventors on the machine as a whole.[101] In 1967, a federal trial court cleared up much of the doctrinal obscurity and tried to reverse the bias against collaborative research. Monsanto v. Kamp Co. arose from an interference proceeding before the U.S. Patent Office, predecessor to the PTO.[102] The invention in question was a specially coated plastic container for pharmaceuticals.[103] The plaintiffs challenged the designation of two defendants as joint inventors, which launched the court into an original discussion of the principles of joint invention:

A joint invention is the product of collaboration of the inventive endeavors of two or more persons working toward the same end and producing an invention by their aggregate efforts. To constitute a joint invention, it is necessary that each of the inventors work on the same subject matter and make some contribution to the inventive thought and to the final result. Each needs to perform but a part of the task if an invention emerges from all of the steps taken together. It is not necessary that the entire inventive concept should occur to each of the joint inventors, or that the two should physically work on the project together. One may take a step at one time, the other an approach at different times. One may do more of the experimental work while the other makes suggestions from time to time. The fact that each of the inventors plays a different role and that the contribution of one may not be as great as that of another, does not detract from the fact that the invention is joint, if each makes some original contribution, though partial, to the final solution of the problem.[104]

The district court decision in Kamp was (and is) cited with some frequency by other courts, including the Federal Circuit.[105] Although Kamp stood as a beacon of thoughtful interpretation, a district court decision could not unify patent doctrine nationwide. Eventually, Congress had to clarify the standards of joint invention. This took the form of the 1984 Patent Law Amendments Act (PLAA), which added the following language to what is now Section 116(a):

Inventors may apply for a patent jointly even though (1) they did not physically work together or at the same time, (2) each did not make the same type or amount of contribution, or (3) each did not make a contribution to the subject matter of every claim of the patent.[106]

This language has not been altered since and remains the main Patent Act provision defining the conditions for joint inventorship. The 1984 PLAA reversed the all claims rule, allowing a research team to apply for a single patent on a unitary inventive concept despite the fact that some team members may have contributed to only a single claim in a multiple claim patent. It also codified the trend advanced by the Kamp decision toward relaxing the standard of collaborative invention. It is now clear that direct interaction between or among collaborators is unnecessary. For example, consider the following scenario:

{2} Three scientists, Alpha, Beta, and Gamma, are involved in research calculated to lead to the discovery of a room-temperature superconducting material. Alpha is involved throughout the entire project, which takes a total of ten years before an invention has been conceived and is ready for patenting, but Alpha’s contribution, although of an inventive quality, is relatively minor. Quantitatively, Alpha’s contribution of effort, expertise, and creative thinking amounts to about 10% of all the effort, expertise, and creative thinking that went into the invention. Beta was involved from the beginning as well, but after nine years and six months, quit the project. Beta’s contribution was, however, over­whelming, accounting for 80% of the research effort, etc. During the last six months, Gamma became involved in the project and collaborated with Alpha to finalize the conception of the invention. Due to Gamma’s late arrival, Gamma contributed only 10% to the project. Beta and Gamma never directly communicated with one another; Gamma merely took up Beta’s almost-completed work.

In scenario {2}, a requirement of direct collaboration at the time the invention was completed would make Alpha and Gamma joint inventors, but not Beta, despite the fact that Beta individually contributed much more than the other two inventors combined. This result is not only unfair to Beta, it creates a powerful disincentive to collaborate in applied research, because, except in the rare situation that the research is certain to result in the successful development of a valuable invention in short order, a researcher’s contribution could be wasted if the researcher needs to leave the collaboration before a patentable result is produced. Many such reasons would be beyond the control of the collaborator, such as illness, financial distress during extended research, termination of employment with an employer sponsoring the research, and so forth. The PLAA establishes beyond peradventure that Beta is as much an inventor as his collaborators. Similarly, the PLAA suggests that in assessing inventorship in scenario {1b} above, if Ron rather than Ilsa discovered the utility of Ilsa’s compound, Ron would be a coinventor. Before the PLAA, courts could deny the importance of Ron’s inventive contribution. For example, in Eli Lilly & Co. v. Premo Pharma­ceutical Laboratories, Inc., two scientists had synthesized a compound (cephalexin) in “the hope that the resulting compound would be [an] effective oral antibiotic” for certain biochemical reasons.[107] A third scientist tested the compound and found that it was indeed an effective oral antibiotic for entirely different biochemical reasons than the first two scientists had guessed. The patent on the compound, however, listed only the first two scientists as inventors. When an accused infringer challenged the validity of the patent for nonjoinder, the Third Circuit held that the third scientist’s discovery of the invention’s “nonobvious trait” did not qualify him as a joint inventor of the compound. In attempting to justify this result, the court wrote:

it is without question that the named inventors, Drs. Morin and Jackson, were the only persons who performed the synthesis that created the patented product. In the words of § 116, cephalexin was ‘made by’ the two named inventors, not by the biochemist who first noted that the organic chemists’ predictions had been realized.[108]

After the PLAA, this result even less plausible than before. A compound whose utility or nonobvious traits are unknown has not yet been “conceived” as an invention under the patent law.[109] A collaborator who makes a significant, original contribution to conception, such as by discovering utility or the nonobvious trait, is a coinventor even if his contribution was different in type (testing of the compound versus synthesis of the compound) from that of his collaborators.[110] With regard to the level and kind of contribution required of a joint inventor, the PLAA added relatively little.[111] Instead, the newly organized Federal Circuit developed joint inventorship analysis in its jurisprudence—a jurisprudence that continues to dominate discussions of inventorship today.[112] Beginning in the early 1990s, the court decided a series of cases that delineated the basic substantive contours of joint inventorship.[113] For the most part, these did not radically depart from the main body of cases decided before the PLAA. Before the amendment, the concept of joint inventorship was rarely interpreted to require “give-and-take” between or among joint inventors. In a 1980 case, for example, a Florida district court found that when one researcher developed a prototype of the invention and passed the prototype to a second researcher, who perfected it, they were joint inventors despite the lack of any regular communication between them. “The ideas” of the first inventor “were presented daily” to the second “by the prototype, and the final result was a creation that exceeded the results of either inventor,” the court observed.[114] The court understood that, although collaboration always requires communication between the collaborators, communication can take nonverbal as well as verbal forms. There is no policy reason to prioritize verbal over nonverbal collaboration if both forms can accelerate technological development. The Federal Circuit continued this interpretation of joint inventorship law; so long as the inventors were consciously cooperating, one collaborator may simply read and build upon a report written by another, or profit from the suggestion of another at a meeting, to qualify them as joint inventors.[115] More importantly, the Federal Circuit unified and fleshed out the kind of contribution necessary to qualify a collaborator as a joint inventor. From Pannu v. Iolab,[116] three general conditions may be deduced. First, the alleged coinventor must contribute to the conception of the invention.[117] Second, his contribution to the invention must not be insignificant in quality, when measured against the dimension of the full invention.[118] Third, the contributor must do more than merely explain to the real inventors well-known concepts or the current state of the art.[119] The second and third conditions follow in a sense from the first. The alleged joint inventor’s contribution could not merely relate the state of the art to the collaborators, because providing well-known information “is not a contribution to conception.”[120] One does not invent merely by repeating public information or using ordinary skills; the contribution must go beyond well-known prior art. Similarly, non-technical contributions to the invention, such as suggest­ing a desirable goal of research (unless identifying that goal is technically difficult and not obvious) and providing management or financing, are not contributions to conception.[121] The Federal Circuit has repeatedly observed that Section 116 of the Patent Act “sets no explicit lower limit on the quantum or quality of inventive contribution required for a person to qualify as a joint inventor.”[122] However, the court has interpreted Section 116 to impose a second condition on the alleged joint inventor—that his contribution must be “not insignificant in quality, when that contribution is measured against the dimension of the full invention.”[123] Where to draw the line of significance on the scale between a critical contribution and a trivial one is a tricky and highly subjective exercise. Unfortunately, the case law interpreting significance remains notably thin. In one unpublished case, the U.S. Court of Appeals for the Fourth Circuit interpreted the significance factor to require that all joint inventors contribute to the specific aspect of the invention that made it patentable. In Levin v. Septodont, the developer of a new numbing cream to treat mouth ulcers, Dr. Kilday, was approached by a retired dentist, Dr. Levin, who had the idea of developing a numbing mouth rinse to reduce pain caused by routine dental procedures.[124] Working together, Kilday and Levin determined that the active ingredients in the rinse should be the antiseptic phenol and the anesthetic benzocaine.[125] Kilday determined the proportion of ingredients in the product.[126] However, Kilday also recognized that these ingredients could not be dissolved in an alcohol solution due to the burning sensation caused by alcohol. Kilday therefore approached Eastman Chemical to help develop a suitable suspension solution. After significant research, Eastman developed a new solution composed of a combination of polyethylene glycol and propylene glycol that solved the burning problem.[127] Eastman sought a patent on the resulting dental mouth rinse, but named only its employees, and not Kilday.[128] Septodont sought to correct the patent by joining Kilday as a coinventor. Levin resisted on the ground that the PTO examiner had opposed the patent application as obvious until a limitation was added specifically claiming the Eastman suspension solution. The court sided with Levin, interpreting joint inventorship doctrine to require contribution to the novel and nonobvious limitations of the claim—indeed, regardless of whether a specific collaborator’s contribution appears in the claims or indeed is the defining feature of the invention.[129] In other words, the Fourth Circuit would have interpreted an inventive contribution’s “significance” in light of its novelty and nonobviousness. Levin v. Septodont is not a precedential opinion,[130] but it illustrates a troubling misunderstanding of the concept of joint invention. The court effectively held that the collaborators who conceived the invention as a whole, determined its utility, and chose key ingredients and their proportions did not contribute significantly to conception. The problem arises from the Fourth Circuit’s dissection of the invention. The invention as a whole was novel and nonobvious—an anesthetic mouth rinse for routine dentistry—and it was conceived by Levin, Kilday, and Eastman scientists in collaboration. A collaborator’s contribution need not define the nonobvious aspect of the invention itself, so long as it satisfies the three Pannu conditions. No doubt the Eastman collaborators were inventors as well, given their contribution to an important and nonobvious aspect of the invention. Possibly their suspension solution was independently patentable, and in that case the Eastman scientists would stand alone as inventors of the solution. But the patent claimed not the solution itself but rather an “anesthetic and antiseptic mouth rinse” using a non-alcohol solution.[131] Simply put, the invention before the court would have been impossible without Levin and Kilday’s original contributions. In denying them status as coinventors, the Fourth Circuit not only deprived a valuable collaborator of the patent reward, but conferred that reward on Eastman for novel and nonobvious ideas that Eastman did not develop and probably could not have developed itself. It is probably fair to say that any contribution adding the novel, nonobvious, or useful element to the invention is ipso facto significant. To conclude from this, however, that any other contribution is insignificant is a non sequitur. The question of what kind of contribution does qualify as significant—a question the Federal Circuit has not had occasion to answer in great detail—remains open. Presumably, such factors as whether the collaborator: would weigh heavily in favor of finding an inventive contribution “not insignificant” if directed to any functional aspect of the invention.[132] The contributor’s skill set and abilities relative to other collaborators may have some bearing on whether these criteria are satisfied. The kind and level of expertise of the collaborator, his effort expended relative to other collaborators, and the importance and originality of his insight to the advantages of the invention, would all be relevant as well.[133] The consequences of designating too many or too few collaborators as inventors on a patent application may be trivial or (much more rarely) fatal, depending on the provable intent of the applicants. Misjoinder or nonjoinder of joint inventors is supposed to invalidate a patent if done intentionally, but an accused infringer cannot realistically be expected to produce clear and convincing evidence of the applicant’s state of mind in most cases. As a result, successfully proving misjoinder or nonjoinder usually results in a correction of inventors in order to save the patent from invalidity.[134]

2. The Distorting Effect of Claims Fixation

The joint inventorship doctrine described in Pannu v. Iolab is sound as far as it goes. It meshes with both the constitutional limitations on the congressional patent power and the language and policy purposes of the Patent Act. The problem is not in the general standard of joint inventorship, but in its implementation. The Federal Circuit has sometimes diverted the inquiry away from a fact-intensive examination of the realities of the inventive process into a rigidly formulaic test, with the regrettable result of facilitating mistaken or fraudulent nonjoinder. The judicial decisions carrying the greatest potential to warp inventorship standards probably are not substantive, but procedural. This relates back to the manner in which courts have confused the claims with the invention. The Constitution and Patent Act refer to inventors and their inventions, but as noted, courts have frequently written the invention out of patent law and substituted the claims, so that, for example, an inventor in the Federal Circuit’s jurisprudence does not conceive an invention, but rather a “claim,”[135] and a coinventor must contribute not to the conception of the invention, but to the conception of “the subject matter of a claim.”[136] The standard procedure developed by the Federal Circuit for evaluating joint inventorship was explained in Trovan, Ltd. v. Sokymat S.A. Irori.[137] There, the court equated inventorship analysis to infringement analysis.[138] Infringement analysis begins with a binding interpretation of the claims through a Markman hearing.[139] With the meaning of the claims established, the court then compares the accused product or process to the claims on a claim-by-claim basis; if the accused product or process literally or equivalently satisfies all the limitations of a claim, then infringement has occurred.[140] Under Trovan, the procedure to be followed in inventorship analysis also begins with an interpretation of the claims “to determine the subject matter encompassed thereby.” The next step is “to compare the alleged contributions of each asserted coinventor with the subject matter of the properly construed claim to then determine whether the correct inventors were named.”[141] As one patent lawyer has observed, inventorship is now determined “on a claim-by-claim basis”[142] as well, despite the fact that an inventor on any one claim is considered an inventor of the entire invention (and, therefore, all of the claims).[143] Trovan analysis, which has become the accepted procedure for evaluating disputes about inventorship, creates an aberration in the Federal Circuit’s otherwise mostly sound jurisprudence on joint inventorship. In undertaking inventorship analysis, courts now tend to fixate on whether an alleged coinventor can point to specific claims language directly attributable to that alleged coinventor’s contributions.[144] The result of fixating on claims is frequently an overly simplistic understanding of the inventive process. In Eli Lilly & Co. v. Aradigm Corp., the Federal Circuit justified its practice of ruling out some collaborators as joint inventors based on the “quality” of inventive contribution:

The line between actual contributions to conception and the remaining, more prosaic contributions to the inventive process that do not render the contributor a coinventor is sometimes a difficult one to draw. Contributions to realizing an invention may not amount to a contribution to conception . . . if they are too far removed from the real-world realization of an invention, or if they are focused solely on such realization.[145]

The first point, relating to highly abstract or indefinite contributions, is intended to weed out collaborators who merely point out the goal without contributing concrete ideas for achieving the goal. A collaborator who does nothing more than to propose that the research team develop a vaccine for HIV or a more efficient solar receptor is not an inventor even if the team succeeds in inventing the vaccine or receptor. Yet, it is quite possible to interpret the court’s language to deprive legitimate inventors of recognition for their contributions. Sometimes the most difficult and important part of the inventive process is understanding the problem to be solved. The contribution of a collaborator who merely identifies the problem may indeed be far removed from resolving it, but his work may be the foundation or even keystone of the invention. For example, suppose a research team working for a chemical developer is trying to synthesize an effective insecticide for termites. The goal is to discover a compound to whose toxicity termites find it difficult to adapt by rapid evolution. After years of fruitless and expensive labor on this project, one team member, Carlos, observes that exposure to certain nontoxic chemicals causes the termites to avoid tree wood and preferentially consume other forms of organic matter. After running some experiments, Carlos determines that it is possible that some such chemicals are likely to resist evolutionary adaptation, although it is not clear which chemical compounds would do the trick. After communicating his findings to the team, Carlosis transferred to a different research team. However, based on Carlos’s observations, the original team reorients its entire research program and, after a few more months, invents a nontoxic compound that effectively makes lumber termite-proof. Carlos’ contribution was undoubtedly “far removed from the real-world realiza­tion” of the resulting compound; he does not even know what the compound looks like or how it was produced. Yet, to say that Carlos was not a joint inventor of the compound disserves the goals of patent law. In the scenario, Carlos’ contribution was not a mere recitation of the prior art and was integral to and necessary for the invention of the compound. If he is not considered an inventor, he has no incentive to reveal his key insight to the other members of the research team, and the policy goal of Section 116 is subverted. Moreover, he is unfairly denied a reward for a breakthrough discovery that enabled the patented invention. The Eli Lilly court’s last point, regarding contributions solely to the “realization” of the invention, is supposed to weed out collaborators who merely performed technical steps in the reduction to practice of an invention already fully conceived.[146] The superficial appeal of this limitation is deceptive as well. The court could merely have been reciting the tautology that, if the invention was already complete, then no further contribution could be significant. If the condition is not tautological (or redundant of the other requirements of joint inventorship), then the collaborator in question must have made both a technically significant and nonobvious contribution to making or using a product invention (or performing a process invention). If the resulting idea could not be made or used (or performed) without a significant and nonobvious contribution from a third party, it has not been fully conceived yet. If the collaborator’s contribution converts an incipient concept into a mature, invented product or process, then he has satisfied the conditions of joint inventorship. Because knowledge of making and using an invention is part of the invention itself, a significant and nonobvious contribution “solely” to the “realization” (or reduction to practice) of an invention may indeed qualify the contributor as a joint inventor.[147] The necessity of the contribution for patentability strongly suggests that the invention had not yet been fully conceived. The court’s statement on this point relates once again back to its fixation on the claims; claims do not necessarily teach how to make and use the invention, so if the claims “define” the invention, then a collaborator who pains­takingly discovers how to make or use the invention properly is no inventor if the means of making or using the invention is not recited explicitly in a claim. While the ambiguities in the Federal Circuit’s jurisprudence on joint inventorship may seem technical, their effects are magnified by the all-or-none consequences of inventorship determinations. Section 116 specifies that collaborators are not precluded from designation as inventors by the fact that they did not contribute to every claim in the patent. In Ethicon Inc. v. U.S. Surgical Corp., the Federal Circuit interpreted this provision to give every inventor who contributed to any claim in the patent equal ownership in all claims, including claims to which that inventor never contributed.[148] Despite frequent criticism of the court’s opinion,[149] the result was all but dictated by the language of Section 116 and the concept of unity of invention that underlies patent law.[150] But this policy has the unfortunate consequence of giving the most important contributor to an invention no greater rights than the least important.[151] The probable result is to dampen the incentive to engage in team research, because adding new members to a research team—especially those not obligated to assign their rights to the same employer—could result in a loss of meaningful patent rights.[152] The current jurisprudence therefore gives patent applicants a strong incentive, and no realistic disincentive, to fraudulently omit joint inventors from the patent application in order to avoid sharing the patent bounty.

C. The Strong Presumption of Validity in Inventorship Analysis

The grant of a patent by the PTO creates a presumption that the patent is correct and valid. The 1952 Patent Act provides expressly that anyone challenging the validity of a patent bears the burden of persuasion that the patent is invalid.[153] This burden originates from early patent law cases in which the U.S. Supreme Court established a general rule of deference to informed decisions of the Patent Office, most notably in Morgan v. Daniels:

Upon principle and authority, . . . it must be laid down as a rule that where the question decided in the Patent Office is one between contesting parties as to priority of invention, the decision there made must be accepted as controlling upon that question of fact in any subsequent suit between the same parties, unless the contrary is established by testimony which in character and amount carries thorough conviction.[154]

The patent applicant actually benefits from a presumption in his favor beginning at the moment the application is filed. The Federal Circuit noted that the PTO bears the burden of challenging the patent applicant’s “presumptively correct” allegations regarding the utility of the claimed invention.[155] A private party challenging an issued patent before a court bears an even heavier burden. Morgan v. Daniels has long been interpreted by the Federal Circuit and its predecessor, the Court of Customs and Patent Appeals, to mean that anyone challenging the validity of a patent must satisfy an elevated burden of proof, by supplying clear and convincing evidence of invalidity.[156] The “clear and convincing” standard of proof is nowhere to be found in the Patent Act. Its origin is instead a long line of judicial decisions holding that a government agency such as the PTO is “presumed to do its job” correctly.[157] As with other aspects of patent validity, a private party challenging a patent on grounds of misjoinder or nonjoinder bears the clear and convincing burden of persuasion.[158] In enacting the 1952 Patent Act, Congress chose not to specify the standard of proof for parties alleging patent invalidity for reasons unknown. In the recent case Microsoft v. i4i, the Supreme Court inferred from Congress’s silence that the legislature intended to incorporate (part of) the judicially invented standard of proof into the statute.[159] Indeed, the Court asserted that it “must” presume—apparently as a canon of statutory construction—that Congress intended to continue the status quo unless the statute otherwise specifies.[160] It is doubtful that, whenever Congress fails to comment on a matter within the scope of a statute, Congress thereby automatically endorses all case law not expressly disclaimed by the statute.[161] However, the fact that Congress has overlooked numerous opportunities to change the highly visible standard of proof suggests that the clear and convincing standard generally aligns with congressional intent in this instance. Doug Lichtman and Mark Lemley have argued against the strong presumption of validity, except with respect to a narrow category of patent challenges, on policy grounds.[162] Although the courts have never articulated a fully developed policy basis for the strong presumption, Lichtman and Lemley infer the two most likely justifications as being deference to the PTO’s expertise and a desire to strengthen patent rights in order to encourage commercialization.[163] But Lichtman and Lemley point out that the PTO does not, probably with current funding cannot, and in any event lacks sufficient motivation to, invest the resources necessary to render an authoritative determination on all questions of patentability.[164] The absence of third-party information in most cases further limits the ability of PTO examiners to assess thoroughly the validity of the patent. As for the disincentive to invest in a patent of uncertain validity, Lichtman and Lemley observe that such uncertainty exists in any case and is only one (probably minor) factor affecting commercialization.[165] To the extent that the patent’s validity has been insufficiently verified by the PTO, strengthening the presumption of validity distorts patent markets arbitrarily. Regardless of whether this argument persuades, the justification for the strong presumption of validity fails utterly with respect to matters never considered by the PTO. The PTO relies heavily or entirely on the thoroughness, honesty, and candor of the patent applicant with regard to some patent matters. Even for those confident in the PTO’s expertise and rigor, a general presumption that the PTO does its job correctly does not justify a presumption of omniscience. Why should the mere submission of the patent application by self-interested applicants create a prima facie presumption of accuracy and truthfulness?[166] On this logic, parties have sometimes challenged the strong presumption of validity with respect to questions not examined by or raised before the PTO. In Eli Lilly & Co. v. Aradigm Corp., the plaintiff made the exceedingly narrow argument that a preponderance of evidence standard should be used “‘when there are two co-pending patent applications claiming the same subject matter’ in front of the [PTO], one of which issues as a patent allegedly omitting the inventor, and the other of which was filed by the allegedly omitted inventor.”[167] The plaintiff, in short, was not challenging the heightened burden of proof in general or even with respect to inventorship questions not considered by the PTO. Its argument was limited to the case in which an interference should have been declared but was not. The Federal Circuit rejected even that limited argument, pointing to the alleged coinventor’s “strong temptation” to misrepresent the extent of its involvement in the collaboration and the absence of any risk to its rights in its own patent application, which it could lose in an interference.[168] Neither of the Eli Lilly court’s contentions justifies the clear and convincing evidence standard. The patent applicant seeking to exclude joint inventors has as strong a temptation to misrepresent its role in the inventive process as the alleged coinventors. The same advantages are at stake for each of them—the ability to commercialize (or suppress) the claimed invention. As for the absence of risk, this argument is simply irrelevant. If the issued patent survives challenge by the alleged coinventor, then the patentee may enjoy its patent rights without sharing them with a coinventor. The alleged coinventor will have no right to obtain a junior patent on the same subject matter, so it risks its rights either way. Moreover, the costs of a full-scale patent litigation dwarf the costs of a review or reexamination, and the alleged coinventor faces as much litigation expense as the patentee. Notwithstanding the Eli Lilly decision, a panel of the Federal Circuit faced another, more aggressive challenge to the clear and convincing evidence standard in inventorship determinations in Vanderbilt University v. ICOS Corp.[169] Pointing to the PTO’s inability to consider the status of alleged coinventors not listed in the patent application, Vanderbilt proposed applying the normal burden of proof.[170] The court again rejected the argument, but this time noted that the plaintiff was “of course free to seek en banc reconsideration of our settled law on this issue.”[171] Although the Federal Circuit has always rejected these challenges to the standard of proof since its creation in 1982, courts frequently expressed doubt with regard to matters not before the PTO both before and after 1952. Morgan itself required deference only to questions “decided in the Patent Office,” and not to questions never considered by the agency.[172] Since then, numerous courts have observed that the rationale for the strong presumption of validity was weakened in such circumstances,[173] and in several cases have actually given reduced deference to the PTO.[174] In KSR International Co. v. Teleflex, Inc., the Supreme Court itself joined the chorus, holding that when a matter had not been raised before the PTO, “the rationale underlying the presumption—that the PTO, in its expertise, has approved the claim—seems much diminished.”[175] Nonetheless, the Supreme Court has recently enshrined a uniform clear and convincing evidence standard for all aspects of the issued patent. The Court in Microsoft Corp. v. i4i Ltd. declined to interpret prior cases to cast doubt on the clear and convincing evidence standard, despite seeming to recognize in the same pre-1952 cases many doubts about the logic of the clear and convincing evidence standard applied to matters unexamined by the PTO.[176] By the majority opinion’s logic, Congress should also be “presumed” to have codified these cases. Instead, the i4i Court adopted an expedient, holding that when a matter was not considered by the PTO, the PTO’s judgment “may lose significant force” and “the challenger’s burden to persuade the jury of its invalidity defense by clear and convincing evidence may be easier to sustain.”[177] Just how the challenged part of the patent “los[ing] significant force” differs meaningfully from lowering the burden of proof is hard to comprehend. Operationally, the Court has intimated that the trial judge can, or perhaps must, notify the jury that the facts at issue were never considered by the PTO while at the same time insisting that the jury apply the clear and convincing standard of proof.[178] But the jury never bases its judgment of patent validity on an assumption of the PTO’s infallibility. The facts are presented to the jury, and the court explains its interpretation of the law, including the burden and standard of proof. The jury may be expected always to decide the question of patent validity based on the facts presented to it, regardless of whether the PTO considered the same question earlier based on different (or no) facts. The jury is not called upon to judge the PTO’s thoroughness or accuracy. If the standard of proof does not vary, neither will the jury’s judgment. How could it? By holding that the 1952 Patent Act codified the clear and convincing evidence standard without equally codifying the case law doubting the uniform applicability of that standard, the Court has seriously undermined a key element of the patent system’s logic. The Court has at best created arbitrary doctrine favoring patent owners and at worst created an incentive for dishonest prosecution practices. The applicant controls the patent application process and decides which information to reveal to the PTO and which to withhold. The PTO, badly overburdened with applications[179] and staffed with examiners who are rarely expert in the specific technology at issue, does not always thoroughly examine such questions as novelty and nonobviousness, and never questions a superficially credible allegation of utility or inventorship. The clear and convincing evidence standard gives applicants a perverse incentive to obtain a patent fraudulently and enjoy special legal protection for their misdeeds. Recently, the Federal Circuit managed to compound the errors of the i4i jurisprudence under the bizarre circumstances of Sciele Pharma v. Lupin Ltd.[180] The patent applicant in that case had received nonobviousness rejections on some of its claims by the PTO examiner. In response, the applicant canceled the claims, effectively creating a public record that the applicant and examiner agreed that the claims were invalid. Nevertheless, by a PTO error, the patent as ultimately issued reinstated the canceled claims.[181] The patentee, instead of seeking a certificate of correction, reexamination, or reissue to correct the error, ignored it. When the patentee attempted to assert these claims in an infringement litigation, the accused infringer attacked the claims as invalid. The claims, the defendant argued, should not benefit from the strong presumption of validity, because they were clearly invalid by public acknowledgement of both the patent applicant and PTO. Nonetheless, the Federal Circuit insisted that admittedly invalid claims should be presumed valid with the elevated standard of proof under i4i.[182] One wonders whether the next doctrinal step will be for courts to uphold as valid claims that the patentee openly admits are invalid during the litigation itself. The consequences for joint inventors alleging unlawful nonjoinder to the patent will be especially unfor­tunate. The flaws in using an enhanced burden of proof for unexamined patent matters are multiplied in the inventorship context by two key differences between the PTO’s inventorship “determination” and any other possible challenge to the patent. First, a charge of nonjoinder or misjoinder does not usually invalidate the patent. The normal remedy is merely correction by judicial amendment of the patent.[183] The consequences of being forced to share patent rights with a coinventor are rarely as dire as the complete invalidation of the patent or its operative claims, and so justify less extraordinary protections for the patentee. Second, as noted, the PTO virtually never questions the applicant’s assertions of inventorship when examining the patent. It relies on the oath of the inventors, which must identify each inventor of the invention for which a patent is sought.[184] Patent applicants seeking to exclude a coinventor for self-serving purposes may have little compunction about misrepresenting inventorship. Because the PTO is well known not to examine inventorship, the chances of discovery are slim unless the patent is challenged in court under the enhanced burden of proof. Even if the patent is challenged, as noted, the result will not be any kind of sanction but mere correction of inventorship. There is no minimally convincing justification for applying the clear and convincing evidence standard to questions of inventorship. What is perhaps most ironic about the strong presumption of validity of the PTO’s fictitious inventorship “determination” is that the Federal Circuit has consistently treated inventorship as a question of law rather than a question of fact. Specifically, it has called the “overall inventorship determination” a question of law, which is “premised on underlying questions of fact.”[185] The court has also observed that the “determination of whether a person is a joint inventor is fact specific, and no bright-line standard will suffice in every case.”[186] These pronouncements suggest some confusion as to the distinction between questions of law and fact. Questions of law necessarily reference the facts that must be present to satisfy the test of legality. Whether these facts are present in any given case is, by definition, not a question of law. The question of fact in the case of joint inventorship is the alleged coinventor’s role in conceiving the invention. The manner of and degree to which a contribution enables or improves the conception of the invention, or the process of conception, are also questions of fact. Whether the facts satisfy the legal standard of significance requires an assessment of how those facts match up to the legal standard—traditionally, a matter within the province of the factfinder, not the court.[187] The Federal Circuit is mistaken, then, in concluding that “whether a person is a joint inventor” is a question of law. The standard of law is indeed a bright line: either the alleged coinventor made a significant contribution to conception of the invention, or she did not. There is nothing in between. The pertinent question of law is typically what kind of contribution to conception is sufficiently “significant” to qualify the contributor as a joint inventor. What the court means in saying that there is no bright line, rather, is that it feels unable to articulate with great precision what is the standard of “significance” as a rule for all cases. The question of law is what the standard of significance should be, not whether any specific alleged coinventor’s contribution is significant. In any case, given that the Federal Circuit has adopted the position that inventorship is a question of law, it follows that the appellate court should give no deference to the PTO’s decision on inventorship and should thus review challenges to inventorship under the normal preponderance of evidence standard. Appellate courts defer to lower tribunals on questions of fact only, while reviewing determinations of law de novo.[188]

II. Relinking Inventorship Analysis to Technology Policy

The previous part revealed how the dysfunction in joint inventorship law can be traced to two key doctrines. The first is the treatment of claims as if they defined the invention rather than merely limited the scope of its patent protection. The second is the strong presumption of validity for all aspects of the patent, including inventorship, in the absence of any cogent policy justification for such a presumption. Combined, these two positions threaten to warp the logic of inventorship and to undermine the incentives for collaborative research. In Part III.A, the alignment of modern inventorship doctrine with the legislative intent underlying the adoption of the 1984 amendments to Section 116 of the Patent Act will be evaluated. Part III.B follows with an analysis of the policy consequences of the current inventorship doctrine and a proposal for how to correct it to conform to legislative intent and the public policy favoring team research.

A. Legislative History of the 1984 Patent Law Amendments Act

Congress faces relatively few constitutional limitations on its ability to grant and enforce patent rights. The journals of the 1787 constitutional convention record no debate on the intellectual property clause, which appeared in the earliest drafts of the Constitution and was unanimously approved and passed. The clause empowers Congress “To promote the Progress of . . . useful Arts, by securing for limited Times to . . . Inventors the exclusive Right to their . . . Discoveries.”[189] It has long been accepted that the clause means approximately what it says; it authorizes Congress to give to inventors an “exclusive” right to practice their inventions[190] for the purpose of encouraging the development of new applied technology (the “useful Arts”). The Constitution thus seems to mandate that patent legislation be interpreted in the manner most likely to advance technology. In pursuing this goal, patents must be granted to the true inventors of the invention patented; the Constitution nowhere authorizes the grant of exclusive rights to persons other than inventors.[191] Both the first Congress and the Supreme Court accordingly interpreted the clause to require the patent reward to be directed to the true inventor of the invention for which a patent is sought.[192] The legislative history of the 1984 PLAA tells much about its policy purposes, and these align well with the Constitution’s mandate. The PLAA began as a pair of bills introduced in the Senate and House of Representatives in 1983 “to make certain clarifications with respect to joint inventors.” The bills in each case merely provided that each inventor “need not have made a contribution to each claim contained in the application.”[193] These early drafts were intended to reverse the judicial “all claims rule,” requiring every collaborator to have contributed to every claim in a patent in order to qualify as an inventor on that patent.[194] The all claims rule forced research teams to struggle to parse out each collaborator’s contribution to each claim and to file multiple patents on closely related inventions, with a resulting risk that some patents would become prior art and preempt others by the same collaborators. The 1983 bills were popular with the patent bar and inventing industries, but these also viewed the bills as an opportunity to codify certain other key pre-Federal Circuit judicial decisions, such as Monsanto v. Kamp.[195] The patent bar lobbied Congress through the American Intellectual Property Law Association (AIPLA) and other bar and industry organizations. On March 15, 1984, AIPLA submitted a proposed amendment to the bills specifically intended to open the door to joint inventorship in the kinds of collaboration not involving direct contact between joint inventors or an equal contribution by each. The amendment’s goals were described in testimony by AIPLA’s president:

Researchers in an organization sometimes work on one aspect of an invention, while others may work on a different aspect. Personnel are continually added to the research team, while others may leave the team. Concepts and development are often generated through brainstorming and cannot accurately be attributed to a particular inventor or inventors. The criteria for joint inventorship, as the amendments to Section 116 would state such criteria, have been judicially recognized.[196]

The main judicial recognition to which the AIPLA president referred was Monsanto v. Kamp. His testimony made clear that paragraphs (1) and (2) of the new Section 116 were adopting the criteria cited by the Kamp district court.[197] The AIPLA report to the House Subcommittee on Courts, Civil Liberties, and the Administration of Justice repeated these concerns:

A research project in today’s environment may include many inventions and some inventions may have contributions which are made by some individuals who were not involved in other aspects of the invention. It is appropriate to recognize the contribution of each individual even though the individual may not have been involved in, or may not have contributed to, all aspects of the invention.[198]

High-ranking PTO officials supported the AIPLA amendments with testimony that paralleled or duplicated the AIPLA rationales for amending Section 116.[199] In his written statement to Congress, the Commissioner of Patents echoed AIPLA’s point about the difficulty of pinpointing the source of inventive contributions during collaborative research:

Scientists or researchers in an organization often work on a particular aspect or embodiment of the invention, or on only a portion of the invention, while others work on different aspects, embodiments or portions. Scientists are continually added to a research team, while other scientists leave the team. Concepts and development plans generated through brainstorming cannot always be accurately attributed. . . . H.R. 4527 would eliminate the need for making these sometimes chancy, complex and time-consuming determinations by specifying that joint inventors need not have contributed jointly to each claim in an application. As we understand the provision, inventors would also be regarded as joint inventors whether or not they physically worked together at the same place or at the same time in developing the invention. Further, joint inventorship would not require that each inventor make the same type or amount of contribution to the invention or that each make a contribution to the subject matter of each claim of the patent.[200]

The Commissioner was mistaken in asserting that the amendment would “eliminate” the need to make a determination of inventive contribution, because, after the amendment, the applicants would still have to determine whether each collaborator listed as an inventor had made the threshold contribution to conception. The amendment would, however, create a more uniform standard and lower threshold for qualifying as a joint inventor, which would reduce the risk of nonjoinder and misjoinder. Identical statements were made by the Commissioner of Patents before the Senate Judiciary Committee.[201] Industry representatives also testified overwhelmingly in favor of the amendments.[202] Representatives of industries in which collaborative research is extremely common if not critical to commercial success, such as industrial chemicals and pharmaceuticals, supported the bill on these grounds. The American Chemical Society’s letter of support asserted:

The proposed modification of Section 116 of 35 U.S.C. is appropriate and just, for it recognizes that much research that results in an invention is conducted on a team basis. Team members may each contribute to a significant stage of the research, but seldom does each team member contribute to each stage. The ACS supports this modification for it removes the inequity of depriving an individual of the status of joint inventor when that person was a significant contributor to an invention.[203]

In response to these suggestions, the sponsor of House Bill 4527 (Rep. Kastenmeier) let that version die in committee and introduced a new bill, H.R. 6286, in November 1984 containing AIPLA’s language verbatim.[204] In doing so, Kastenmeier also mentioned the new bill’s expected benefits to “universities and corporations which rely on team research”[205]—a purpose also reflected in the official analysis of the bill.[206] The Senate bill was similarly amended, with a parallel explanation of the purpose of the bill described as recognizing “the realities of modern team research.”[207] The report of the post-conference bill made identical references to these “realities” and made explicit that paragraphs (i) and (ii) adopted “the rationale of decisions such as Monsanto v. Kamp.”[208] The President’s statement on signing the bill into law, too, emphasized its utility for removing technical obstacles to team research.[209] With strong industry support and minimal opposition to the bills, Congress enacted the amendment to Section 116 in the precise words proposed by AIPLA.[210] The legislative history indicates that the primary purpose of the PLAA was to codify a flexible standard of joint invention in order to encourage team research and, ultimately, efficient technological development. Congress evidently perceived no coherent policy purpose in forcing large research companies and university science and engineering laboratories to conform to arbitrary rules about which researcher may participate on a team without precluding a valid patent or at least raising the costs of patenting. Ideally, the sole determinant of research collaboration should be the most efficient use of research personnel and resources. Interfering with the research efficiency hardly advances the patent law’s purpose of fostering technological development. With its exclusive concern the promotion of collaborative research, the PLAA amendment to Section 116 left the substantive definition of conception untouched. The amendment does confirm that different collaborators may make different types and amounts of contributions without losing their status as coinventors, and it does not limit such contributions to any specific subclass of inventive input. Beyond that, the PLAA had nothing to say about the standard of inventorship. Yet, the Federal Circuit soon began interpreting inventorship in a manner that privileges certain kinds of inventive contributions while dismissing other, equally valuable contributions. In Ethicon v. U.S. Surgical Corp.,[211] the court’s analysis focused on whether the putative inventor’s contribution appears in the claims. The court qualified this by stating elsewhere in the case that coinventorship depends not on a contribution to the claim itself or specific language in the claim, but rather on whether the collaborator “contributed to the invention” defined and limited by a claim.[212] Nonetheless, the court in Trovan reinforced the doctrinal fixation on claims by laying out a formal procedure for determining inventorship that seemed to rigidly ignore every inventive contribution not reflected in specific claim language.[213] As discussed, Ethicon and Trovan set forth a procedure for determining inventorship. However, the Federal Circuit has never explicitly held that an inventor’s contribution must appear expressis verbis in the claims. In both Ethicon and Trovan, the inventor had allegedly contributed specific, identifiable limitations or elements to a combination invention. In Trovan, the alleged contribution was the addition of a wire support feature to an electronic transponder. In Ethicon, the alleged contribution was the addition of a safety feature (a “detaining means”) to a surgical trocar. In both cases, the contribution was a specific physical feature integrated into the invention. Such features are almost invariably set forth explicitly in the claims, and the Federal Circuit accordingly focused on whether the alleged coinventor’s contribution was encompassed by the claims. The contribution to the claims in such cases served as a satisfactory proxy for an analysis of whether the inventor contributed to the invention itself. In neither case did the Federal Circuit repudiate the importance of contributing to the invention when that contribution was not recited explicitly in the claims. Indeed, the court has recognized that a coinventor must contribute to the conception of the “invention” rather than the claims.[214] The picture that emerges from these cases is one of a jurisprudence mired in confusion about the critical distinction between the claims and the invention. The most problematic aspect of Trovan analysis, which has become the normal procedure for determining inventorship, is its elision of the distinction between claims and invention through the inapposite analogy of inventorship to infringement analysis. Infringement analysis is not a good parallel to inventorship analysis for several reasons. Most importantly, the two analyses serve entirely different policy purposes. Infringement analysis focuses tightly on the claims because the Patent Act’s disclosure requirements limit the scope of the patent. In the infringement context, focusing on the claims makes good sense. The claims are meant to limit the inventor’s enforcement powers. By definition, infringement means that the accused product or process has trespassed on one or more of the specific claims. It follows that infringement analysis should revolve around the claims. However, even in the infringement context, the claims are not the only or even last word on the scope of patent protection. Courts have generally recognized that, as important as claims are to determining whether patent infringement has occurred, they do not exist in a vacuum, but rather should be interpreted in the context of the invention as a whole, as described in the specification and prosecution history.[215] Moreover, using the doctrine of equivalents, courts have expanded patent protection to encompass more of the invention than the inventor technically claimed.[216] Why should the patent owner have the right to enforce the patent against a product or process not explicitly claimed in that patent? The answer can only be that the doctrine is designed to protect the invention when the claims do not fully capture that invention. Such a theory is possible only if the invention is something greater than the claims. The theory underlying the doctrine of equivalents is precisely that the invention and the claims are different but overlapping concepts. The need for claims relates especially to their public notice function. Claims serve the purpose of putting the public on notice of what practices the patent forbids, while the patent document as a whole also teaches the public how to practice the invention. An infringement determination is the precondition to enforcing the patent, and so focusing on the claims protects accused infringers against being subjected to damages and penalties that were unforeseeable based on the claim language. Limiting infringement actions to the claims requires the inventor, as the person with the most control over the phrasing of the patent application, to define as clearly as possible the scope of the invention so that competitors will be given fair notice of the subject matter protected by the patent.[217] Accordingly, infringement (at least, literal infringement) means intrusion on the claims by definition. If the accused infringer’s activities did not fall within the scope of the claims as reasonably interpreted, and did not equivalently infringe, the accused infringer need fear no sanctions. In contrast, an inventorship determination is designed neither to protect the public’s expectations about which technologies they are free to use nor to maintain the patent’s value against threats by competitors. It is instead intended to encourage innovation and, in the context of Section 116 of the Patent Act, to preserve the incentives for collaboration between persons having complementary skill sets or whose research work otherwise synergizes well. Accordingly, Section 116 of the Patent Act nowhere says or implies that the inventive contribution must appear in the claims portion of the specification. Getting inventorship right encourages inventive activity by rewarding those persons who make a significant contribution to the invention and serves equity by ensuring that no collaborator free-rides off another collaborator’s important contribution. For these purposes, the appearance of a collaborator’s contribution in the claims may indicate that the collaborator is a bona fide coinventor if the contribution is significant, goes to conception, and does more than recite the state of the art. Nothing in the pre-1984 case law or the PLAA suggests the adoption of a change in the substantive standard of inventorship to exclude any collaborator whose contributions do not appear in the claims. The policy rationale for not limiting inventorship analysis to the claims is further explored in the next section. Unfortunately, the clarity of the jurisprudence on joint inventorship has been further degraded by imprecise language in American Board of Education v. American Bioscience, Inc.,[218] where the Federal Circuit suggested that a collaborator could not be a joint inventor of a class of chemical compounds unless it “conceived” one “of the claimed compounds.”[219] These dicta do not actually reflect the trend of Federal Circuit jurisprudence on the whole. The court stated in Ethicon that each joint inventor “needs to perform only a part of the task which produces the invention,”[220] as provided in Section 116 of the Patent Act. In Fina Oil, the court was even clearer: “One need not alone conceive of the entire invention, for this would obviate the concept of joint inventorship.”[221] Imagine the probability of all joint inventors simultaneously forming identical pictures in their individual minds of the entire operative invention.

B. Reconstructing Inventorship

It will be helpful here to summarize the main points of the argument so far. Claims limit the protectable aspect of the invention; they do not “define” the invention, as courts frequently assert. Inventorship analysis cannot be analogized to infringement analysis, as the Federal Circuit has erroneously suggested in Ethicon, Trovan, and a long line of subsequent cases. It is unnecessary for a collaborator on a research team to have contributed specific language to a claim in order to qualify as a joint inventor, so long as he or she satisfies the three Pannu conditions. These observations lead to the conclusion that a collaborator who makes a significant contribution to the conception of an invention may qualify as a coinventor without the need for Trovan analysis. More specifically, a collaborator who makes a sufficient contribution to the utility or nonobvious aspect of the invention, or to the method of making or using it, may qualify as an inventor even if his contribution does not appear in the claims. Without such contributions, there might have been no invention at all, and therefore no patent with claims to enforce. Claims limit the patentee’s enforcement powers; they do not, however, suffice to qualify the invention for patent protection. That protection is contingent on the invention’s novelty, nonobviousness, utility, and adequate disclosure. Unless the coinventors collectively satisfy all of these requirements, no patent may issue in the first place. A uniform requirement to conduct Trovan inventorship analysis, in con­fusing the claims with the invention, undermines this dynamic. If the invention were nothing more than the claims, the analogy to infringement analysis would make sense. But the invention is different, sometimes radically different, from the claims. The resulting merging of the claims with the invention disenfranchises any coinventor whose contribution the patent applicant did not expressly describe in the claims as a limitation, no matter how important that contribution might have been. This section will discuss how a correction of the judicial understanding of “invention” can point the way toward a correction of determinations of joint inventorship. With such a correction, the incentives toward collaborative research and disincentives toward strategic nonjoinder of inventors can be restored.

1. Equity and Strategic Nonjoinder

Dan Burk and Mark Lemley have explained some practical consequences of overemphasizing the claims and ignoring the invention as a whole. As they point out, patent applicants can draft claims narrowly in order to get around objections by the examiner and to obtain a patent, and then later interpret the claims broadly in light of the specification or through the doctrine of equivalents:

The shift in focus from the invention to the claim language allows both sides to game the process. It permits—and indeed even encourages—over claiming by patentees, particularly patentees drafting or interpreting claims years after the invention itself. If the focus is on the language of my claims, not the product that I actually built or described, I can interpret the language creatively to claim, in retrospect, to own inventions that I didn’t have in mind when I wrote the patent claims.[222]

The result is to undermine the disclosure and public notice function of Section 112 of the Patent Act. The patent bar is well aware of these consequences and capitalizes on them as a matter of course. Although some patent lawyers avoid overclaiming to strengthen the patent against potential challenges, most seek to draft claims broadly by default.[223] With regard to inventorship determinations, the focus on the claims would allow legally sophisticated researchers to exclude naïve collaborators from inventorship by drafting claims in a manner that underplays or ignores the latter’s crucial inventive contributions, without which the patent might have been unobtainable. Consider the following scenario:

{3} Three scientists, Delta, Epsilon, and Zeta, jointly begin a research project for the discovery of a flexible, bulletproof fabric. Through extensive testing, Delta and Epsilon rule out hundreds of fiber candidates and suggest several molecular and material characteristics that a bulletproof fabric would have, with clues to how one could fabricate such fibers. With this guidance in hand, Zeta successfully creates the material.

In this case, all three collaborators qualify as joint inventors on the resulting patent. The contributions of each are ex hypothesi more than mere recitations of well-known principles and are plainly significant to the invention’s novelty, nonobviousness, and utility. The Federal Circuit has accurately observed that “a person is not precluded from being a joint inventor simply because his or her contribution to a collaborative effort is experimental,”[224] so the fact that Delta and Epsilon were not involved in the final steps in inventing and crafting the patented material does not deny them joint inventor status.[225] Yet, suppose Zeta, who is the only inventor in possession of a complete mental picture of the fabric, were to apply for a patent without crediting Delta and Epsilon as coinventors. Zeta could rely on Trovan and its progeny to support his self-identification as the sole inventor of the material. Focusing on the claims, a court would compare the inventive contributions of Delta and Epsilon to the claim language. Nothing in the claims would necessarily point to the contributions of Delta and Epsilon, especially if Zeta drafted the claims strategically. The claims would certainly not mention anything about the fiber candidates that Delta and Epsilon had so painstakingly ruled out; claims describe what the applicant regards as the invention, not what the applicant regards as irrelevant or nonfunctional. Yet, much of the inventive process is typically “negative discoveries” that winnow the paths toward the inventive goal to the smallest number possible.[226] Nor would the claims necessarily refer to the material’s bulletproof characteristics; claims (especially in the chemical, metallurgical, and pharmaceutical arts) commonly describe an invention without reference to its utility or purpose. This information would most likely appear in the written description portion of the patent. The reason for its omission from the claims makes sense in the context of patent law, where a patent grants exclusionary rights to the patented invention used for any purpose, regardless of whether it was a purpose contemplated by the inventor. Patent applicants accordingly omit reference to the utility and other details of the invention in the claims to avoid narrowing the scope of the claims. The motivation of seeking a patent that covers unforeseen uses of, or methods of making, a patented product is logically unconnected to, and has no effect on, the determination of inventorship of that product. The utility of an invention, and the knowledge of how to make and properly use it, are essential aspects of the invention, even if they do not necessarily limit its protections. As the only collaborator with knowledge of the material’s molecular structure and fibers, and perhaps the only one with knowledge of how to fabricate the material, then, Zeta may rely on a great deal of Federal Circuit precedent to arrive at precisely the wrong conclusion—that he is the sole inventor.[227] The effort and creativity that Zeta put into the invention may be dwarfed by those of his collaborators, but the logic of Trovan analysis, strictly applied, awards him the entire patent right anyway. This problem does not arise as readily in patents covering mechanical devices or processes, because a coinventor will usually have contributed an identifiable physical component to a device or a step to a process. For example, in Ethicon v. U.S. Surgical Corp., the alleged coinventor of surgical trocar could point to a specific safety means limitation, incorporated into the relevant claim, that he contributed to the trocar.[228] Moreover, mechanical device patents frequently specify their utility in the claims themselves,[229] so a collaborator’s contribution to utility would survive Trovan analysis. But no great imagination is required to envision hypothetical cases posing the same problem in the mechanical engineering field. For example, a novel device invented by Iago may be obvious over the prior art until Charles discovers a nonobvious use for the device.[230] Or a novel manufacturing process invented by Ilsa may be nonfunctioning when applied to certain materials, but if Calvin discovers that the process functions well when applied to a nonobvious choice of materials, Calvin is a coinventor. The inventive contributions of Charles and Calvin rendered their respective inventions patentable, but those contributions need not appear in the claims. The arbitrary exclusion of some inventors caused by overreliance on the claims creates special problems in materials, chemical, and pharmaceutical development. As noted, patents in these fields typically claim the compound itself with no mention of its utility. Discussion of utility is relegated to the written description portion of the patent for such inventions.[231] Claims that omit any mention of the invention’s utility, or the method of making and using it, highlight the merely partial role played by the claims as a definition of the invention. They also prevent a collaborator who discovers a compound’s utility from qualifying as an inventor under a strict Trovan analysis. Pharmaceutical inventions illustrate this point with special aptness. In developing a new pharmaceutical invention, the scientific problem is not to create a novel compound in a vacuum. A pharmaceutical compound, once created, does not “do” anything; it has no utility outside of its target biological organism. In order for a patent covering such a compound to issue, the applicant must credibly allege that the compound has some biochemical effect on a target organism. In other words, a pharmaceutical compound is not a functioning invention until introduced into its target organism. The claims, to the extent that they merely recite some molecular structure without mention of a therapeutic utility, limit the patentable scope of the invention but cannot “define” it in any meaningful sense. The incomplete role of the claims in “defining” pharmaceutical inventions reveals its significance clearly in any description of the inventive process. The process of conceiving and synthesizing the claimed compound is often a relatively small part of drug design. In order to complete conception of the compound through trial-and-error testing, the inventors may need to perform all of the following steps: (1) understand the biological mechanism or agent causing the disorder; (2) select or develop a target chemical molecule (“candidate”) potentially capable of treating the disorder; (3) develop an efficient method for producing the candidate on a mass scale; (4) develop or select a series of systems for screening the candidate against the biological mechanism, or an environment that simulates the mechanism; and (5) modify the candidate as needed to increase its potency; reduce toxicity, side effects, or contraindications; reduce the molecular half-life; and determine proper dosage and delivery method.[232] As a result, pharmaceutical inventions usually result from team research involving biologists, biochemists, medicinal chemists, and pharmacologists performing different tasks. Inventors may have contributed to any of these steps. In some cases, the biological mechanism underlying the disorder to be treated may be well understood in the art, but the kind of molecule that will remedy the biological disorder will be the subject of uncertainty. In other cases, the biological mechanism will be mysterious or misunderstood, and the conception of the invention may require unique insights into the biological source of the disorder. The latter class of inventions is by far the most challenging to develop and perfect; all pharmaceutical inventions require the selection, testing, and possible modification of a bioactive compound, but not all require extensive original research into human or other animal biochemical functioning. Yet, because the claims will typically recite only the chemical structure of the molecule, an inflexible Trovan-type analysis would arbitrarily privilege the second and fifth steps over the others, despite the fact that in any given case the former may require the least effort, expertise, creativity, or other inventive contribution.[233]

2. Restoring Incentives for Collaborative Research

Claim fixation results in an inventorship analysis obviously inequitable to the excluded inventors, but it also poses a threat to collaborative research. Often, private companies seek to collaborate with university scientists and engineers to develop innovative products. For example, drug developers frequently collaborate with university biologists and chemists to discover diagnostics and treatments for persistent human diseases and disorders. University researchers are typically involved in exploring human or other animal biology or biochemistry, while the drug companies have expertise in medicinal chemistry and pharmacology. In such collaboration, Trovan analysis seems to suggest that the university researchers could never qualify as inventors if they did not contribute some specific molecular arrangement to the claimed compound. The patent incentive to collaboration with drug companies disappears. An interesting illustration of the flaws in Trovan analysis occurred in a previously mentioned case, Board of Education ex rel. Florida State University v. American Bioscience.[234] There, FSU scientists and their post-doctoral fellow, Dr. Tao, had been experimenting with variants of paclitaxel, a naturally-occurring anticancer compound. At the FSU lab, Tao learned a “semi-synthetic” process for producing paclitaxel analogs. During this time, FSU scientists undertook a research project to develop nitro-taxols as chemo­thera­peutic radio­sensitizing taxanes. Within a relatively short time, the FSU scientists discovered that a specific compound they had synthesized, PNIP, showed promise as a radiosensitizing agent. One of the FSU scientists (Dr. Nadizadeh) had developed a “secret” method for synthesizing PNIP, of which Tao was informed. At some point during this research, one of the FSU scientists spoke at a conference regarding the synthesis of paclitaxel. Scientists at a pharmaceutical developer, VivoRx, had attended and decided to start researching radiosensitizing agents using a variant of paclitaxel, known generically as docetaxel, as the parent structure. Soon after having begun research, VivoRx hired Tao to assist in the project, and with his help in synthesizing compounds (allegedly using Nadizadeh’s secret method of making nitro-taxols), VivoRx obtained patents on several docetaxel analogs. In evaluating whether any FSU scientists qualified as joint inventors of the patented compounds, the Federal Circuit discounted FSU’s experimental work: “general knowledge regarding the anticipated biological properties of groups of complex chemical compounds is insufficient to confer inventorship status with respect to specifically claimed compounds.”[235] The court then observed that “invention does require conception, and there is no evidence that FSU’s inventors conceived any of the claimed compounds.”[236] With regard to the use of FSU’s methods to make the compounds, the court concluded that this was irrelevant, because the patent did not claim the method of making the compounds.[237] Specifically, FSU alleged that Tao had used Nadizadeh’s secret method to synthesize the patented compounds. To this, the court replied:

despite the fact that Nadizadeh may have developed a method of making PNIP and other taxol derivatives, the record in the present case indicates that he did not conceive the claimed compounds; only ABI’s inventors were in possession of both the structure of the claimed compounds and an operative method of making those compounds.[238]

Although the court acknowledged that conception of an invention requires disclosure of the method of making it, it held that the inventor of the method could not be a coinventor of the compounds made by his method unless he “conceive[d] of the claimed compounds” himself.[239] The American Bioscience court’s focus on the “claimed compounds” and discounting of the (unclaimed) method of making them is symptomatic of claims fixation. Whether FSU scientists conceived of an entire claimed compound is not the pertinent question; the patented compounds were conceived by at least some of the inventors. The FSU scientists need only have contributed to the conception of the invention in some way; conception of the entire compound by any specific inventor is entirely irrelevant, as long as the research team collectively conceived the claimed compounds. A proper analysis of inventorship would have acknowledged that the contribution of necessary background biological knowledge, useful suggestions for functional groups of the claimed compounds, and techniques for making the compounds, could all qualify the contributors as coinventors if the contributors had collaborated with the VivoRx team and satisfied the three Pannu conditions of coinventorship. Under this analysis, it is possible that some of the FSU scientists would not have qualified as joint inventors. It is also possible that some would have qualified. The court’s repeated distortion of the law of inventorship in the case did considerable mischief on the facts before it, where FSU scientists who may have contributed substantially to the VivoRx inventions were excluded from consideration as joint inventors. In effect, the Federal Circuit sanctioned what may have been the misappropriation of a university’s valuable research, gathered over the course of many years, by a private company that promptly commercialized it for its own benefit and the preemption of any benefit by the university scientists. More troubling still, the decision has had a predictably perverse effect on the development of joint inventorship doctrine. For example, it misled the district court in Vanderbilt University v. ICOS Corp. into concluding that failure of any alleged collaborator to independently conceive of the entire claimed invention negated any possibility of joint inventorship.[240] Very recently, the Federal Circuit has retreated from the American Bioscience reasoning and partially rectified the wayward doctrine. In Falana v. Kent State University, the plaintiff, a university researcher, developed an original method of making certain compounds for use in liquid crystal displays.[241] After the plaintiff (Dr. Falana) left the university’s employment, his supervisor used Falana’s method to synthesize a compound that he patented without naming Falana as a coinventor. The university read American Bioscience to dictate that, because it had not sought to patent Falana’s method, and Falana had never conceived the patented compounds, he was not a coinventor of any compound resulting from the use of his method. The Federal Circuit disagreed with this understanding of American Bioscience.[242] Actually, the university’s reading of the case was quite reasonable. The American Bioscience opinion was plainly dismissive of a researcher who had contributed a method of making the patented compound without conceiving of the compound itself. To get past American Bioscience without reversing itself, the court had to somehow reconcile its own holding with the fact that Dr. Falana, like Dr. Nadizadeh in American Bioscience, had never made or seen the university’s patented compounds. To do this, the court resorted to the time-honored tactic of distinguishing the cases on the facts. In American Bioscience, Nadizadeh never actually worked with the patentee’s inventive team directly; he worked with Tao, who in turn worked with the patentee to develop the compound. In Falana, the eponymous plaintiff worked directly with the university research team, although he left before the compound was discovered. This distinction is, of course, irrelevant; the Patent Act does not require that all joint inventors work together, merely that they are exposed to each other’s work in the course of some cooperative enterprise. The Falana court also suggested that, in American Bioscience, Nadizadeh had simply “taught skills or general methods” that merely facilitated the later invention “without more.”[243] But this too is irrelevant; the operative question is whether the skills and methods taught by the collaborator (a) contributed more than well-known principles, and (b) made a significant contribution (c) to conception. As the court itself recognized in Falana, “Where the method requires more than the exercise of ordinary skill . . . the discovery of that method is as much a contribution to the compound as the discovery of the compound itself.”[244] Problematic as the distinction between the cases may be, it is salutary that the Federal Circuit is beginning to perceive a more accurate concept of invention. The misfortune is that, in not reversing the mistakes of American Bioscience unambiguously, the court leaves standing contradictory approaches to the role of claims in determining inventorship. Also problematical is the potential of claims-fixation to lead to an oxymoronic “inventorless invention.” A slight extension of scenario {3} demonstrates how this paradox might arise:

{4} Eta, Theta, and Iota are collaborating to discover a cure for lung cancer. Through diligent and unusually insightful research, Eta discovers a hitherto unknown biochemical mechanism that makes cancer cells susceptible to certain kinds of polypeptides. Theta invests several months into creating a series of nonobvious screening systems for testing compounds against the discovered vulnerability in cancer cells. Together, Eta and Theta propose to Iota, a medicinal chemist, the characteristics that such a compound should have. Following these instructions, Iota pulls a well-known drug compound off his shelf, modifies it into a series of novel derivative compounds using well known techniques, and runs them through Theta’s screening systems. One of the derivative compounds turns out to be a wonder drug that cures lung cancer with a very high probability in humans. Together, they seek a patent on the compound only.

Under the facts of scenario {4}, neither Eta nor Theta ever pictured in their minds the finished compound. Nor did they contribute any specific physical structure to the claimed molecule. They lacked the requisite knowledge of medicinal chemistry to do so. Neither Eta nor Theta would be coinventors under a strict Trovan analysis or the American Bioscience approach, because none of their contributions appear in the claims. Iota synthesized the patented molecule, and was the first and only collaborator to conceive it in his mind, but nothing in his contribution exceeded the state of the art of medicinal chemistry. His work was the least inventive of the three, and arguably did not rise to a contribution to conception. If the claims merely recite the molecular structure of the claimed compound, the result will be that an invention of undoubted novelty, nonobviousness, and utility, might have no inventor at all.[245] Under an expansive interpretation of Trovan, the cure for cancer described in scenario {4} could well be doctrinally unpatentable.[246] The result is diametrically opposite to any patent policy of encouraging collaborative research and the equitable sharing of its benefits.

Conclusion

The argument of this Article—that the concept of invention and inventorship has been erroneously construed as “defined” by the claims—has relied on an argument sufficiently complex to proceed in stages. First, I showed that the invention cannot be defined by the claims, because the claims do not necessarily express all the prerequisites for a patentable invention. These omitted elements include most typically utility, a method for making the invention, and a method for using the invention. Without these, no novel and nonobvious product or process has yet attained the status of an “invention” ready for patenting. Therefore, a research collaborator who contributes one or more of these aspects to an invention qualifies as a co-inventor even if his contribution does not appear in the claims. Moreover, treating the claims as defining the invention contradicts both the unity of invention rule and the rationale underlying the doctrine of equivalents. The Trovan sequence of analysis relied upon by the Federal Circuit consequently cannot be the only valid method for determining inventorship. Second, I explained how the consequences of fixating on the claims undermined Congress’ clearly expressed intentions in amending Section 116 of the Patent Act to encourage collaborative technological research. Third, I discussed the aggravating effects of two erroneous and contradictory doctrines adopted by courts: (a) the strong presumption of validity accorded to all aspects of an issued patent, regardless of whether the PTO examined the aspect at issue, and (b) the treatment of inventorship as a question of law. Finally, I suggested how inventorship doctrine could be rectified, most generally by treating the claims as defining the outward limits of patent protection for the invention rather defining than the invention itself. Procedurally, courts should hew to Falana analysis and resort to Trovan analysis only when the alleged co-inventor’s sole contribution appears expressis verbis in the claims. Until the Supreme Court or Federal Circuit decisively clarifies the distinction between the invention and the claims, and the role of joint inventors in the former, the law of inventorship will continue to impose unnecessary costs on collaborative research and to skew inventorship determinations in favor of legally sophisticated but unprincipled researchers to the disadvantage of their less urbane collaborators. The problems resulting from claim fixation are further aggravated by adherence to an unjustifiably strong presumption that the persons named as inventors in the patent are the true and only inventors. There is no statutory basis for a strong presumption of validity on matters unexamined by the PTO, and no cogent rationale for judicially imposing a “clear and convincing evidence” standard. Because all coinventors have equal and nonexclusive ownership rights in the patent by default,[247] the stakes for nonjoinder are potentially very high. Absent an assignment agreement, all inventors own rights to all claims in their patent, regardless of whether they contributed to those claims in any way. The Federal Circuit’s decision in Ethicon giving inventors ownership rights to claims to which they did not contribute was perhaps inevitable given the unity of invention principle and the phrasing of Sections 116 and 262 of the Patent Act, but it had the unfortunate consequence of increasing the risks of collaborative research outside of an employment contract. Collaborations between private companies, or between research universities and private industry, stand to suffer unnecessarily. Any adjustment of the ownership rights in a patent with multiple inventors would best be accomplished legislatively, but many of the current doctrine’s ill effects can be remedied by rectifying the wayward judicial conception of the invention. The overwrought fixation on claims and strong presumption of validity in inventorship determinations appear nowhere in the Patent Act. They are progeny of the judiciary and correctible by it. The judicial attitude toward overemphasizing claims should mirror that of Bill Cosby’s father to the young Bill: “I brought you into this world, and I can take you out.”[248]

---