The Yale Law Journal


Why Modularity Does Not (and Should Not) Explain Intellectual Property

11 Oct 2007

A virtual cottage industry of intellectual property (IP) models has sprung up in recent years. To the mix, Henry Smith adds modularity, in which there are intense interactions within, and few interactions between, components. But before displacing more traditional explanations, the theory must address six fundamental challenges: (1) explain why modularity should be IP’s defining feature; (2) account for the roles played by IP statutes and doctrine; (3) specify clear boundaries for innately imprecise patents and copyrights; (4) justify discerned differences between patent and copyright; (5) rationalize the enhanced injunctive relief suggested by the model; and (6) prove the game is worth a new empirical candle.

First, Smith does not demonstrate that modularity should be IP’s defining characteristic. IP historically has balanced many important goals: providing incentives for innovation, encouraging dissemination and disclosure, and balancing rights with limits that preserve competition and expression. In recent years, the regime has been the subject of vigorous debates in multiple arenas, including the Supreme Court and Congress, on issues such as patentability requirements and appropriate relief for patent and copyright infringement.

In this context, it is hard to see the value of modularity’s benefit in “making decisions more manageable” by displacing the need “to foresee how . . . actions [within modules] might affect the insides of other modules.” Until now, this has not been viewed as a problem in IP. And in addition to not showing how modularity offers a solution to IP’s pressing issues, Smith does not reveal the deficiencies with the regime’s centuries-old justifications. Nor would the model lead to useful results, as it threatens—as discussed below in the context of remedies—adverse effects for industry and innovation. Finally, Smith’s use of modularity in property (in which it hides existing uses) does not directly carry over to IP (in which it also must create new uses).

Second, in viewing the IP universe through modularity-colored glasses, Smith neglects the important roles played by IP statutes and doctrine. He underscores exclusion’s indirect role in allowing an inventor to pursue multiple uses. But such a role does not take account of other, more obvious explanations. History tells us that the Framers intended that the right to exclude would provide incentives that would “promote the Progress of Science and useful Arts.” Economists tell us that exclusion allows inventors to raise prices, thereby recouping their costs and encouraging innovation. Even moral rights theorists rely on exclusion’s support in pockets of copyright such as the Visual Artists Rights Act and derivative rights. Modularity’s role in explaining exclusivity is tangential at best.

Smith’s failure to consider traditional explanations also appears in his rationalization of the difference between patent and copyright. Smith posits information costs as the reason for patent interdependence and copyright independence. But he does not consider the obvious differences between the IP statutes. Because an inventor can receive an improvement patent, blocking patents are common. Because only a copyright owner can create derivative works, blocking copyrights are not.

Another example of modularity’s overemphasis is Smith’s primary example of an IP governance regime: the fair-use defense in copyright law. Smith views fair use as a needed refinement to an exclusionary regime that is “inadequate to accommodate multiple uses cost-effectively.” But the defense makes more sense as a counterweight in a system that treats commonplace activity—such as photocopying a single page in a book—as infringement.

Moreover, fair use does not call for the “evaluation of conflicting uses” at the core of Smith’s governance strategy. Unlike nuisance, which considers both parties’ uses in determining liability, copyright evaluates the adequacy of only the alleged infringer’s uses in a setting in which a prima facie case of infringement has already been made. This distinction matters: staring down the barrel of a cease-and-desist letter, even parties with viable fair-use defenses often will stop engaging in the activity. And on the subject of overzealous enforcement, most observers likely would not agree that the Digital Millennium Copyright Act (DMCA) represents a “detailed governance regime of fine-tuned balancing.”

The first two hurdles thus question the justification for a theory of modularity. But even if there were a need for the new framework, there would still be challenges with the model itself.

The third involves the requirement of boundaries. Modularity’s overriding prerequisite is a set of boundaries that allows activity in other modules to be disregarded. This condition could be satisfied in real property when a person stands in the middle of a clearly marked parcel of land. But boundaries are not as lucid in IP, despite Smith’s assertions of “easy to describe” inventions, “delineation costs [that] are not fundamentally different from those . . . in property,” and information coordination “[a]cross [IP] boundaries” that “only requires monitoring a simple on/off signal.”

As intangible creations, and in contrast to “the exclusivity of real and personal property[, which is] easy to grasp,” inventions lack real-space equivalents and boundaries. Relatedly, patent claim construction is notoriously difficult. The inventive step forming the basis for a patent often cannot be expressed in words, which helps to explain the Federal Circuit’s reversal of roughly one-third of district court determinations on claim construction. Though modestly clearer, copyrights also present challenges in tracing the elusive idea-expression dichotomy, as Smith recognizes in acknowledging the difficulty of “defining a protected literary character or style.” In short, a theory centered on boundaries in a discipline in which they are inherently difficult to discern confronts considerable challenges.

Fourth, many of the differences between patent and copyright that Smith situates in the modularity paradigm are not sufficiently supported. The “[m]ost basic[]” involves the contrast between patent law’s exclusion and copyright law’s “various use rights.” But are the two regimes really so different? Patents give their owners “the right to exclude others from making, using, offering for sale, or selling the invention.” Copyrights give “exclusive rights” to reproduce the work, prepare derivative works, distribute copies, and publicly perform and display the work. Each, in short, is based on a right to prevent others from engaging in particular activities.

Another difficulty appears in the syllogism that (1) “when the stakes are high, resource conflicts call for governance rules”; (2) “the stakes are . . . higher in patent law”; but (3) patent law is characterized by a “more exclusionary” regime. Smith explains the syllogism’s failure by pointing to patents’ “high measurement cost[s].” But the premises themselves may be to blame: (1) “high stakes,” in raising the importance of innovation incentives, could call for exclusion rather than governance rules; and (2) Smith illustrates patents’ higher stakes only through tangential evidence (at least some of which he concedes “provides little guidance”) and in disregard of his recognition that <“inventions come in many different varieties.”

An additional unproved assumption involves patents’ multiple uses, which form the basis for stronger exclusion, as compared to copyright’s fewer uses. But again, “inventions come in many different varieties,” and patents in fact have starkly disparate uses—or even non-use—in various settings. Nor are copyrights susceptible to a blanket conclusion of few uses. Finally, Smith does not address the argument that it is copyrighted works that have higher information costs because of their (1) subjective content and (2) more frequent encounters among a large number of potential infringers.

Fifth, even if there were a need for modularity, the model Smith presents would threaten undesirable consequences. Just as the Federal Circuit, on the Supreme Court’s orders, modified its rigid ordering of injunctions in cases of patent infringement, Smith aims to rectify the “overlooked benefits of exclusion” by proposing more frequent use of injunctive relief. In particular, he asserts that “judges should be more likely to find irreparability of harm” where owners’ “use of rival inputs to information is difficult to second-guess.” Smith does not explain how to resolve such a nebulous inquiry. But even if he did, a return to more frequent injunctions would exacerbate a real problem.

Some of the most intense and dangerous interactions today occur not within one particular patent (as modularity would suggest) but in industries such as semiconductors or computer software in which a single product contains many patents. In these cases, the owner of a patent on one of the components could seek an injunction that would “force the downstream producer to pull its product from the market.” This danger has been emphasized in the context of non-practicing entities (often called trolls), who, because they do not make a product themselves, are not subject to reciprocal infringement lawsuits and are tempted to seek sums greatly in excess of the patent’s value. Smith does not explain how his enthusiasm for injunctions addresses these problems, makes sense in light of reasonable patent trends moving in the other direction, or is consistent with his preference for a damages remedy for inadvertent infringers.

Sixth, it is not clear that the game is worth the empirical candle that Smith is illuminating. For a regime premised largely on utilitarianism, IP is devoid of clear answers. George Priest famously stated that analysis of the patent system is “one of the least productive lines of inquiry in all of economic thought.” But uncertainties about the ideal breadth and duration of rights do not mean that a set of new empirical questions would lead to more definitive answers for more important questions.

In fact, new empirical studies would be even more difficult in comparing the costs of exclusion (in others’ foregone access and use) with its benefits (in owners’ coordination and development of uses). How are we to compare foregone costs and coordinated benefits? No common denominator is offered, nor is one immediately apparent. Moreover, in viewing IP’s primary benefits as a type of prospect theory (by which the owner coordinates development after creation), the formula downplays the role of invention itself.

In conclusion, substantial obstacles remain before IP is stripped of two centuries of law and policy and robed in a new paradigm of modularity.

Michael A. Carrier is a Professor of Law at Rutgers University School of Law -Camden.

Preferred Citation: Michael A. Carrier, Why Modularity Does Not (and Should Not) Explain Intellectual Property, 117 Yale L.J. Pocket Part 95 (2007),