The question raised in Holden as to the potential for overlapping regimes of protection also arises with regard to patent law and PVP, although in the case of overlapping utility patents, limitations will arise out of horizontal conflicts between the two federal statutes, rather than due to vertical conflicts between a federal statute and a state statute. The permissibility of overlapping patent and PVPA protection has been addressed by the US Supreme Court in the much ballyhooed decision Pioneer Hi-Bred Int'l. Inc. v J.E.M. Ag Supply,6 where the defendant, accused of saving seed in violation of a patent-based 'seed-wrap' licence, challenged the propriety of utility patent protection for plants. The applicability of utility patents to plant varieties was upheld by the high court in an astonishingly badly reasoned opinion authored by Justice Thomas. Although the analytical sins of the opinion are legion, I shall focus only upon the two most directly germane to the question of GURT deployment.
The first of these deals with overlapping IP regimes. The J.E.M. opinion both assumes and asserts that IP regimes with overlapping subject matter are quite routine, even unremarkable. To support this proposition, Justice Thomas relies upon citations to the Court's previous opinions in Kewanee Oil v Bicron and in Mazer v Stein. But these decisions are at best irrelevant and may indeed point in the opposite direction from Justice Thomas's claim. Neither case in fact dealt with overlapping IP regimes. In Kewanee Oil, the Court held that state trade secrecy law is not pre-empted by the federal IP system; but this result was based on the assumption that the two forms of intellectual property protection (IPP) are mutually exclusive: to obtain patent protection, one must disclose the invention in the published patent document, thus forgoing trade secrecy.
Similarly, Mazer v Stein7 considered the division between what has traditionally been the dividing line between patent and copyright - the utilitarian and the aesthetic - but far from allowing the two to overlap, the opinion keeps them distinctly compartmentalized. The opinion in Mazer is famous for establishing the rule that functional aspects of copyrightable works are not covered by copyright. Only aesthetic portions of the work, to the extent that they are physically or conceptually separable from the work's functional aspects, can be protected by copyright. In order to keep the two IP regimes separate, should the utilitarian and esthetic portions of the work prove inseparable or inextricably co-mingled, copyright protection becomes unavailable.
The inapposite nature of these decisions to Justice Thomas's reasoning is not only ironic, but altogether prophetic with regard to overlapping patent protection for trans-genic plants. The rule established in Mazer has proven to be exceptionally important for software, which is one of the very rare artefacts that can be simultaneously subjected to both the patent and copyright regimes. Copyright has no purchase on software to the extent that it is functional - only the literal code and non-functional aspects of its structure can be protected by copyright. This in turn allows copyright's fair use doctrine to provide an effective reverse engineering privilege for software. Courts have held that the creation of intermediate or temporary copies, made in the course of accessing uncopyrighted and unprotectable aspects of the program, to be fair use.
To the extent that patent law may be available to protect a computer program's functional aspects, it may serve to override or negate this reverse engineering privilege. Patent law includes no fair use provision, nor indeed much else in the way of user privileges or exemptions. This has been a source of concern to commentators reviewing the state of the software industry, as software innovation depends upon interoperability and patents may serve to block the development of interoperable products. The introduction of a fair use doctrine, or its equivalent, into patent law has been suggested to alleviate this problem (Burk, 2000; O'Rourke, 2000; Cohen and Lemley, 2001).
The negation of exemptions and privileges by overlapping patent protection in software is precisely the same problem elided by the Court in J.E.M. And this leads us to the second key analytical flaw in the opinion. Much of the holding in J.E.M. rested upon the questionable conclusion that the patent and PVPA statutes are in some fashion compatible, so that no inference might be drawn that Congress did not intend the two statutes to overlap.8 But in fact this conclusion of compatibility is blatantly wrong in a number of respects, particularly in that the patent protection is entirely incompatible with the research and farmer's exemptions in the PVPA. The overlay of patent law onto PVPA subject matter negates these exemptions in the same fashion that patent protection overrides the reverse engineering exemption for software. Whatever public benefit or constituent balance that Congress intended for the exemptions is thus lost, much as it would be with the overlay of GURTs upon plant varieties.
This problem is not new to the Supreme Court jurisprudence; indeed, overlapping IPP has posed a problem in a variety of situations where one form of IP threatens to disrupt the substantive limits or policy balance of another. Instead of looking to the inapposite Kewanee Oil and Mazer v Stein decisions, the Court in J.E.M. might better have looked to the Supreme Court precedent in the area of trademark law, where the court has repeatedly expressed its concern that overlapping patent and trademark protection not be permitted to disrupt federal patent policy. While the term of patent protection has been carefully limited by Congress, trademark or trade dress protection can last as long as the owner continues to use it in commerce - theoretically forever. The Court has repeatedly held that one form of IP cannot be used to subvert the limits of another.
Thus, in Kellogg Co. v National Biscuit Co.,9 the Court held that consumer recognition of a distinctively shaped 'shredded wheat' biscuit could not be used under the law of trademark to extend a patent owner's rights beyond the term of the patent on the biscuit. The Court reaffirmed that principle with regard to distinctively shaped traffic sign assemblies in Traffix Devices v Marketing Displays10 - a case decided just before J.E.M., but curiously not mentioned in the opinion. Neither is the principle limited to patent law; in Dastar Corp. v Twentieth Century Fox,11 the Court held that documentary films for which copyright had lapsed could not be protected under the Lanham Act's false designation of origin provision. The reasoning in J.E.M. sits uncomfortably amid these decisions, offering little indication as to whether the confluence of patent and PVPA protection is more acceptable than the confluence of patent and trademark, or copyright and trademark.
And indeed, when combined with cases holding that there is no conflict between the overlap of the PVPA and trade secrecy, this holding essentially reads the PVPA out of existence: if a plant developer has the election between robust patent protection for 20 years, or perpetual protection under trade secrecy, it is difficult to see when he would ever opt for PVPA protection that is shorter than trade secrecy and weaker than patent. Some commentators have suggested that the PVPA is indeed unnecessary and redundant (Janis and Kesan, 2002). But the statute does exist, and until the Congress repeals it, the statute demands that the IP system should be envisioned so as to give it effect.
Was this article helpful?