Intellectual property regimes frequently employ boundary screens. Boundary screens protect the different balances of competition and protection that Congress has struck in the different regimes by preventing the goods whose protectability should be determined by one regime from infiltrating into and receiving protection under another regime.

Prior scholarship on boundary screens offers in-depth analyses of the functionality screens in nonpatent intellectual property that avoid upsetting patent law’s competition–protection balance for functional innovation. This Article turns the table, asking a previously unasked question about how patent’s authorship screen—that is, its boundary screen that prevents infiltration by the authorial innovation that is the proper domain of copyright—does and should work. Shortcomings in patent’s authorship screen upset copyright’s competition–protection balance, allowing patents to function as abnormally thick backdoor copyrights, just as shortcomings in copyright’s functionality screen allow copyrights to function as abnormally long and easy-to-obtain backdoor patents.

In addition to its normative assessment of the authorship screen’s importance as a barrier to backdoor copyrights and its descriptive analysis of the statutorily diffuse set of patent doctrines that collectively enforce the authorship screen, this Article presents case studies focusing on architectural innovation, an unstudied zone of overlap on the copyright–patent boundary that illustrates the authorship screen in action.

Introduction

Intellectual property is not a homogeneous body of law. There are several distinct intellectual property regimes—copyright, trademark, utility patent, and design patent, most prominently—and Congress has struck a different balance of competition and protection within each regime. Different regimes have different procedural and substantive requirements for acquiring rights and different rules for determining the scope of rights. Given this diversity, each intellectual property regime must have boundary screens. Boundary screens prevent costly infiltration: they keep the goods that should be protected by a first regime from infiltrating a second regime and upsetting the competition–protection balance for those goods that Congress struck in the first regime.

When the topic of intellectual property boundary screens arises in the law reviews and courts, the usual question is how copyright, trademark, and design patent should deploy functionality screens, that is, the boundary screens that prevent the functional innovation that should be governed by utility patent law (or just patent law, for short) from receiving nonpatent intellectual property protection.1 The idea–expression dichotomy of Baker v Selden2 and the “useful articles” doctrine3 implement copyright’s functionality screen, trademark employs the eponymous functionality doctrine,4 and design patent screens out functional innovation by protecting only the ornamental features of manufactures.5 All of these doctrines prevent nonpatent intellectual property regimes from sanctioning “backdoor patents” on functional innovation and upsetting the competition–protection balance for functional innovation established in patent law.6 Recognizing that backdoor patents would generate social costs, scholars have investigated the nature of the functionality that should lead to exclusion from a nonpatent regime,7 compiled lists of doctrinal strategies for excluding functional innovation from nonpatent regimes,8 and offered normative frameworks for evaluating which of these strategies are appropriate along which boundaries.9

Despite this voluminous scholarship on boundary screens, the inverse question has received no attention at all: How does and should patent law screen out the goods that, if they are to be protected, ought to be protected as nonpatent intellectual property?10 The oversight is puzzling because this question raises similar concerns about social costs that its oft-studied mirror image raises. Contrary to conventional wisdom, patent protection for the goods that should be protected by the nonpatent regimes is not costless. It upsets the competition–protection balances of the nonpatent regimes by overriding nonpatent law’s limitations on scope, mirroring how nonpatent protection for the goods that should be protected by the patent regime upsets the competition–protection balances of the patent regime by overriding patent law’s limitations on duration and protectability. The principal difference between the two questions is a flip in identity of the infiltrating good and the infiltrated regime.

To begin the enterprise of understanding and evaluating patent’s boundary screens, this Article examines patent’s boundary with copyright.11 Copyright’s subject matter is works of authorship,12 so this Article coins the phrase authorship screen to describe patent’s mirror image of copyright’s functionality screen, and it employs the term authorial innovation to describe the goods for which copyright should have the authority to establish intellectual property’s competition–protection balance.13 Patent’s authorship screen (hopefully) prevents patents from protecting authorial innovation and serving as backdoor copyrights, just as copyright’s functionality screen prevents copyrights from protecting functional innovation and serving as backdoor patents. The argument unfolds in three steps: it articulates a normative framework for identifying the to-date-overlooked costs of patents that function as backdoor copyrights, provides a descriptive analysis of the statutorily diffuse patent doctrine that constitutes the authorship screen, and details case studies of how the authorship screen does and should work in an unstudied zone of overlap on the copyright–patent boundary—namely, architectural innovation.

Patents that serve as backdoor copyrights are problematic because they override the competition–protection balance that Congress has established for authorial innovation via the copyright regime. Roughly speaking, copyright protection is long in duration, thin in scope, and easy to obtain, whereas patent protection is short, thick, and difficult to obtain. The well-recognized concern of copyrights serving as backdoor patents is that protecting functional innovation with copyright makes protection long and easy to obtain when it should be short and difficult to obtain.14 Inversely, the to-date-unrecognized concern of patents serving as backdoor copyrights is that protection for authorial innovation becomes thick when it should be thin. More specifically, sanctioning patent protection for authorial innovation overrides copyright’s requirement to demonstrate copying as part of infringement,15 its idea–expression dichotomy as applied in Nichols v Universal Pictures Corp,16 and its fair use doctrine.17

Recognizing the policy that underpins the authorship screen leads to a more comprehensive cost-benefit accounting of patent protection than is prevalent today. Conventional patent policy is strictly inward looking: it tallies costs and benefits only for the functional innovation that it is supposed to govern, ignoring the costs generated by protecting authorial innovation and tilting copyright’s competition–protection balance further toward protection.18 At the margin, these additional costs may tip the scales and suggest that patent protection ought not to extend as far as it should under the conventional inward-looking policy analysis.19

As a doctrinal matter, there is no authorship screen in black-letter patent law that announces itself as such. Patent treatises and hornbooks do not mention it. Yet there is clearly law, somewhere, that, somehow, keeps most authorial innovation from infiltrating the patent regime. (Contemporary patent law is not that broken.) In gross, the authorship screen resides in patent law’s insistence that a good must “do” something innovative to be protected. In its various rhetorical guises, it requires that innovation be “useful,”20 “produce some new effect, or . . . produce an old effect in a new way,21 or have changed “functions” or “method[s] of performing them.”22

The initial challenge, of course, for describing the authorship screen is to identify what properties count as functional and what events constitute doing something in the patent sense. Given the doctrinal requirement for utility, effects, or functions, one might expect patent law to cough up a coherent, positive definition of functionality. The assumption baked into the functionality screens of the nonpatent intellectual property regimes—that functionality has a coherent, positive definition—reinforces this expectation.23 This Article inverts this conventional wisdom, arguing that the best way to capture what patent law’s authorship screen does is to recognize a positive definition of the authorial innovation that it screens out. More specifically, this Article posits that authorial innovation is the sum of informative innovation that represents informational content to a human audience and aesthetic innovation that allows a human audience to enjoy pleasurable, form-centered experiences. Although copyright does not expressly recognize a distinction between informative and aesthetic works of authorship, patent’s definition of unprotectable authorial innovation is incomprehensible without it.

Even with a rough definition of authorial innovation in hand, the descriptive task of documenting patent law’s authorship screen is a bit of an adventure. The authorship screen is divided into two parts on two intersecting dimensions.24 First, as foreshadowed above, patent law has two distinct doctrinal screens. An informative authorship screen enforced by the printed matter doctrine screens out informative innovation, and an aesthetic authorship screen enforced by a combination of the utility and nonobviousness doctrines screens aesthetic innovation. Second, within each of these screens, two different statutory provisions perform two different roles. The first provides a coarse screen. It does the easy work of excluding authorial goods, that is, goods that lack any functional features at all, such as loose-leaf textbooks and sculptures. The facet of the printed matter doctrine that is lodged in the doctrine of patent eligibility of 35 USC § 101 does this work for the informative authorship screen, whereas the utility doctrine, also grounded in § 101, does it for the aesthetic authorship screen.25 The second provides a finer screen. It excludes functional goods that embody authorial innovation—that is, goods with functional properties that are not the locus of the innovation, such as hammers with new sculptural handles and machines with new labels.26 The facet of the printed matter doctrine that is lodged in the doctrines of novelty and nonobviousness of §§ 102 and 103, respectively, does this work for the informative authorship screen, whereas the nonobviousness doctrine alone today does it for the aesthetic authorship screen.27

The principle of nonobviousness that this descriptive account tasks with administering the finer of the aesthetic authorship screens—termed here the functionality mandate of nonobviousness—is worth highlighting. The functionality mandate dates back to Thomas Jefferson’s administration of the Patent Act of 179028 and the express text of the Patent Act of 1793.29 Yet, today, it has devolved into an advisory—perhaps even obsolete according to some accounts—rule of thumb in contemporary codifications of patent law.30 Without a functionality mandate in nonobviousness, patent law lacks an effective aesthetic authorship screen, so what we currently call an advisory rule of thumb may in fact be more binding, given that patent law must not generate backdoor copyrights.

Patent law’s authorship screens limit patent protection in a wide array of different technologies. The informative innovation screen rebuffs not only applicants who attempt to patent literary storylines,31 but also, for example, applicants in the pharmaceutical industry who attempt to use the content of written or oral instructions about how to take a drug to differentiate a claimed invention from the prior art.32 Similarly, the aesthetic innovation screen prevents the visually compelling effect of consumer and industrial goods—ranging from footwear and furniture to graphical user interfaces on the web—from supporting applicants’ arguments that their innovations embody patentable departures from the prior art,33 or at least it does so when an innovative functional feature, such as bioluminescence or the refraction of light, is not generating the visually compelling effect.34

Despite the breadth of the authorship screen’s general applicability, this Article examines tightly focused case studies, sacrificing some breadth in order to achieve the depth needed for a fruitful examination of the authorship screen in action. The case studies focus on patents on a particular type of architectural innovation: dispositions of space, or the arrangement of spaces that constitutes a building’s programmatic layout, as are commonly represented in floor plans and their three-dimensional equivalents.35 Dispositions of space inseparably embody both functional and aesthetic qualities, and they exist in an awkward zone of copyright–patent overlap. They therefore offer excellent terrain for observing and evaluating the authorship screen.

More specifically, the case studies offer three lessons concerning patent law’s authorship screen. First, they illustrate that the authorship screen is permissive. Unlike copyright’s strict functionality screen, the authorship screen grants protection when functionality and authorship are inseparably fused. This permissiveness sanctions patents that can be leveraged into market power based on consumers’ aesthetic preferences, even when the screen is rigorously enforced. Second, the case studies show that the authorship screen systematically mislabels look-like innovation—that is, innovation in which the only functional advance is giving one thing a shape that makes it look like something else—as functional when it should be authorial. Look-like innovation falls into an unplanned gap in between the aesthetic and informative authorship screens, and patents on look-like innovation are backdoor copyrights that upset copyright’s competition–protection balance. Third, the case studies show that the US Patent and Trademark Office (PTO) seems to issue patents on what is plainly authorial innovation in dispositions of space on an unusually frequent basis.36 The problem here is not poorly calibrated doctrine but rather PTO errors in the application of the doctrine. One possible explanation for this high frequency of erroneously issued patents is that the authorship screen is too costly to rigorously administer when the claimed technology is highly multifunctional, as dispositions of space are. As the number of different things that an innovation “does” in the patent sense multiplies, the complexity of the argument (and examiner’s workload) required to demonstrate a failure to satisfy the functionality mandate increases at something in between a geometric and an exponential rate.37 At some point, the examiner’s burden grows too large, the authorship screen tears open, and authorial innovation pours into the patent regime.

Part I identifies the overlooked costs of patents protecting authorial innovation and serving as backdoor copyrights. Part II defines authorial innovation and identifies the doctrine that enforces the authorship screen. Part III examines the authorship screen in action, using patents on architectural dispositions of space as case studies.

  • 1. See generally, for example, Christopher Buccafusco and Mark A. Lemley, Screening Functionality, 103 Va L Rev (forthcoming 2017), archived at http://‌perma.cc/
    67KF‌-BWQ8 (discussing the imperfect functionality screens for copyright, trade dress, and design patent regimes); Mark P. McKenna and Christopher Jon Sprigman, What’s In, and What’s Out: How IP’s Boundary Rules Shape Innovation, 30 Harv J L & Tech 431 (2017) (discussing the formation of functionality screens in response to the utility patent’s historically preferred place in intellectual property jurisprudence, as well as their inconsistent operation in maintaining these boundaries). See also generally Mark P. McKenna and Katherine J. Strandburg, Progress and Competition in Design, 17 Stan Tech L Rev 1 (2013) (discussing the supremacy of utility patent law); Mark P. McKenna, (Dys)Functionality, 48 Houston L Rev 823 (2011) (discussing mechanical and aesthetic functionality screens); Viva R. Moffat, The Copyright/Patent Boundary, 48 U Richmond L Rev 611 (2014) (discussing the role of copyright’s “useful articles” doctrine that excludes “useful things” from the realm of copyright); Viva R. Moffat, Mutant Copyrights and Backdoor Patents: The Problem of Overlapping Intellectual Property Protection, 19 Berkeley Tech L J 1473 (2004) (discussing the expansion of intellectual property protections, the more recent overlap between these areas of protection, and the resulting gaps in the doctrine); J.H. Reichman, Charting the Collapse of the Patent–Copyright Dichotomy: Premises for a Restructured International Intellectual Property System, 13 Cardozo Arts & Enter L J 475 (1995) (discussing the lines between copyright and patent law for international intellectual property law, as well as hybrid intellectual property regimes that blur these lines); Pamela Samuelson, Strategies for Discerning the Proper Boundaries of Copyright and Patent Protections, 92 Notre Dame L Rev 1493 (2017) (presenting various strategies from the literature for distinguishing between patent and copyright law). The reach of copyright protection into computer software raises a thorny boundary-screen issue, but this issue is sui generis due to Congress’s decision to extend copyright into an inherently functional subject matter. See note 174. Some scholarship also addresses exclusionary screens at the boundaries of other non-utility patent regimes. See generally, for example, Laura A. Heymann, The Trademark/Copyright Divide, 60 SMU L Rev 55 (2007); Richard W. Pogue, Borderland—Where Copyright and Design Patent Meet, 52 Mich L Rev 33 (1953).
  • 2. 101 US 99 (1879).
  • 3. See 17 USC § 101. See also Richard P. Sybert and Lindsay J. Hulley, Copyright Protection for “Useful Articles, 54 J Copyright Society USA 419, 420–25 (2007) (discussing the genesis of the “useful articles” doctrine and its more modern applications).
  • 4. See TrafFix Devices, Inc v Marketing Displays, Inc, 532 US 23, 29 (2001) (noting the “well-established rule that trade dress protection may not be claimed for product features that are functional”).
  • 5. See 35 USC § 171 (“Whoever invents any new, original and ornamental design for an article of manufacture may obtain a patent therefor, subject to the conditions and requirements of this title.”).
  • 6. Moffat, 19 Berkeley Tech L J at 1523–25 (cited in note 1) (describing the “backdoor patent” and the Supreme Court’s failure to address this issue).
  • 7. McKenna and Sprigman, 30 Harv J L & Tech at 516–40 (cited in note 1).
  • 8. Samuelson, 92 Notre Dame L Rev at 1517–35 (cited in note 1).
  • 9. Buccafusco and Lemley, 103 Va L Rev at *48–65 (cited in note 1).
  • 10. But see Shubha Ghosh, Patenting Games: Baker v. Selden Revisited, 11 Vand J Enter & Tech L 871, 894–95 (2009) (arguing that patents should not encompass some expressive gaming innovation); Pamela Samuelson, Benson Revisited: The Case against Patent Protection for Algorithms and Other Computer Program-Related Inventions, 39 Emory L J 1025, 1037 n 36 (1990) (arguing that the “printed matter” doctrine prevents patents from upsetting copyright’s balance of protection and nonprotection for texts). Professors Mark McKenna and Christopher Sprigman examine the reach of patent law at its boundary with nonpatent regimes, but they do this in order to determine how the nonpatent regimes should enforce their functionality screens. McKenna and Sprigman, 30 Harv J L & Tech at 516–39 (cited in note 1).
  • 11. This Article also implicitly addresses the boundary screen that prevents the ornamental features of manufactures that are supposed to be protected by design patent from obtaining protection under the utility patent regime because copyright also governs these features.
  • 12. 17 USC § 102(a).
  • 13. This Article employs “innovation” in a regime-neutral sense. Functional innovation refers to nonobvious, patentable subject matter, and authorial innovation refers to original, copyrightable subject matter.
  • 14. See Robert P. Merges, Peter S. Menell, and Mark A. Lemley, Intellectual Property in the New Technological Age 491 (Aspen 6th ed 2012); Buccafusco and Lemley, 103 Va L Rev at *6–8 (cited in note 1); Moffat, 48 U Richmond L Rev at 612 (cited in note 1); Reichman, 13 Cardozo Arts & Enter L J at 482 (cited in note 1); Samuelson, 92 Notre Dame L Rev at 1516–17 (cited in note 1).
  • 15. See Arnstein v Porter, 154 F2d 464, 468–69 (2d Cir 1946).
  • 16. 45 F2d 119, 121–22 (2d Cir 1930). The idea–expression dichotomy is usually conceptualized as a single limitation on copyrightable subject matter. However, the dichotomy’s two leading cases—Nichols and Baker—have different normative goals and announce different doctrinal rules. Understanding the difference between these two strains of copyright’s idea–expression dichotomy is crucial for understanding patent law’s authorship screen. See notes 57–60, 63–64 and accompanying text.
  • 17. See 17 USC § 107.
  • 18. See notes 78–82 and accompanying text.
  • 19. Look-like innovation provides a concrete example of when the outward-looking policy analysis of the authorship screen may tip the scales against patentability. See Part III.B.2.
  • 20. 17 USC § 101.
  • 21. Woodcock v Parker, 30 F Cases 491, 492 (CC D Mass 1813).
  • 22. William C. Robinson, 1 Treatise on the Law of Patents for Useful Inventions § 238 at 322 (Little, Brown 1890).
  • 23. McKenna and Sprigman, 30 Harv J L & Tech at 493 (cited in note 1) (arguing that the functionality screens in nonpatent intellectual property work only if there is “a reasonably clear and stable sense of what belongs to utility patent law”).
  • 24. For an overview of the resulting two-by-two matrix, see Part II.D.
  • 25. See 35 USC § 101 (requiring an invention to be a “new and useful process, machine, manufacture, or composition of matter” to be patentable).
  • 26. Patent doctrine often denies that a claim’s point of novelty is relevant to the determination of patent validity, but it regularly pays attention to it behind the curtain. See Kevin Emerson Collins, Getting into the “Spirit” of Innovative Things: Looking to Complementary and Substitute Properties to Shape Patent Protection for Improvements, 26 Berkeley Tech L J 1217, 1238–42 (2011) (arguing for a “spirit” of innovation approach in patent improvement cases that pays attention to the innovative “properties” of improvement inventions, an approach that is conceptually similar to the point-of-novelty approach); Mark A. Lemley, Point of Novelty, 105 Nw U L Rev 1253, 1266–74 (2011) (arguing that the “point of novelty” approach should be explicitly adopted by courts so as to better promote innovation). The finer authorship screen tracks the features that make claimed goods innovative, so it is another example of how identifying a claim’s point(s) of novelty is essential for assessing patent validity.
  • 27. See 35 USC §§ 102–03 (requiring that an invention be both novel and nonobvious in light of the prior art). Given that the authorship screen is spread out across the distinct statutory provisions for utility and patent eligibility (35 USC § 101), novelty (35 USC § 102), and nonobviousness (35 USC § 103), identifying the authorship screen requires breaking through the rigid statutory silos that structure most patent discourse. The isolation of these silos helps to explain the black-letter silence on the authorship screen in treatises and hornbooks.
  • 28. 1 Stat 109.
  • 29. 1 Stat 318. For a discussion of how the functionality mandate emerged, see notes 130–34 and accompanying text.
  • 30. For a discussion of the functionality mandate’s diminishment, see notes 142–46 and accompanying text.
  • 31. See Ex parte Knight, 2014 WL 6984090, *16 (PTAB).
  • 32. See King Pharmaceuticals, Inc v Eon Labs, Inc, 616 F3d 1267, 1277–80 (Fed Cir 2010).
  • 33. The aesthetic authorship screen in utility patent law is the reason why recent fraught discussions over the patentability of product designs and graphical user interfaces have focused on design patents rather than utility patents. See McKenna and Strandburg, 17 Stan Tech L Rev at 38–44 (cited in note 1). See also generally Michael Risch, Functionality and Graphical User Interface Design Patents, 17 Stan Tech L Rev 53 (2013).
  • 34. See notes 236–42 and accompanying text (illustrating the irrelevance of “reverse merger” in the patentability requirements).
  • 35. No previous legal or architectural scholarship considers the patentability of architectural dispositions of space, so this Article opens up a whole new area of copyright-patent overlap to scholarly inquiry. The case studies are enabled by a unique patent dataset that I collected through a citation-snowballing process. For a description of the collection methodology, see note 171.
  • 36. Some scholars label this error a “false positive” because an intellectual property right that should be invalidated is upheld. Joseph Scott Miller, Error Costs & IP Law, 2014 U Ill L Rev 175, 181. Other scholars call it a “false negative” because a boundary screen should invalidate an intellectual property right but does not. Buccafusco and Lemley, 103 Va L Rev at *51–52 (cited in note 1).
  • 37. See note 293 and accompanying text.