On Friday the Federal Circuit issued its long-awaited en banc decision in In re Bilski (No. 2007-1130 , Serial No. 08/833,892), clarifying the legal test for patentable subject matter under 35 U.S.C. § 101. The nine of twelve judges who signed the majority opinion announced a "machine-or-transformation" test for patentable subject matter. Although urged to do so by Judge Mayer in dissent, they declined to overrule State Street-the 1998 decision that has produced a flood of software and business-method patents. The rule they announced, however, replaces the "useful, concrete and tangible result" test of State Street.
Under the new rule, inventions involving "fundamental principles"-laws of nature, natural phenomena, or abstract ideas-are patentable only if they use the fundamental principles in a specific machine or in a process that transforms a specific article into another state or thing. In order to pass that test, the machine or process of transformation must involve more than "insignificant extra-solution activity," i.e., structure or steps beyond applying a formula, algorithm or other fundamental principle.
In announcing the new test, the en banc majority took pains to discuss and reject several alternatives, including some that the Federal Circuit had previously endorsed in panel and even en banc decisions. One was the "useful, concrete and tangible result" test of State Street and an earlier en banc decision (In re Alappat). Another was the so-called Freeman-Walter-Abele test, which excludes patent claims that pre-empt (cover) all practical uses of an algorithm. The Bilski majority repudiated each of these tests as "inadequate" in themselves and "no longer valid." It also rejected the "mental steps" test, which held a process claim unpatentable if the process can be performed entirely in one's mind.
While rejecting the alternative tests as legal norms, the majority explained how the results of each-at least the correct results-might be reached by applying the new machine-or-transformation test. A purely mental process, for example, might fail that test for lack of a tangible machine or transformed article in the claim's most general form. In that sense the "mental steps" test may be consistent with the theoretical basis of the machine-or-transformation test, "namely the prevention of pre-emption of fundamental principles."
The majority also rejected tests proposed by various amici curiae. One was a test based on an attempt to define "technology" or "technological arts." The majority saw such a test as having no basis in the statute and creating uncertainty as technology advances. It also rejected proposed categorical exclusions of software and business-method inventions, preferring to assess those special categories of inventions case by case under the machine-or-transformation test. Yet the majority noted how its own test might also exclude patentability for business methods:
"Purported transformations or manipulations simply of public or private legal obligations or relationships, business risks, or other such abstractions cannot meet the test because they are not physical objects or substances, and they are not representative of physical objects or substances."
Applying its test to an invention that claimed risk-hedging transactions for commodities investing, but did not mention any machine or article, the majority found the claims nonstatutory.
Besides its new subject matter test, the majority opinion in Bilski had several notable features. First, it began by carefully parsing the language of relevant Supreme Court precedent. That in itself was somewhat novel. In some of the very decisions that it abandoned, the Federal Circuit had analyzed its own precedent in great detail, giving short shrift to decisions of our highest court.
In recent appeals, the Supreme Court has reversed the Federal Circuit's formalistic rulings at least six times. Perhaps that poor "batting average" is finally having an impact. Whatever the reason, the Bilski majority repeatedly emphasized the Supreme Court's guidance. It carefully developed an understanding of what it saw as the Supreme Court's preferred test, thereby acknowledging its own subordinate position in our judicial system, despite its presumed patent expertise.
The second notable point of Bilski was the majority's view of the subject-matter inquiry. It ruled that subject matter is or can be a threshold test, to be applied before the other substantive requirements of patents-namely, novelty, utility, nonobviousness and proper disclosure. While refusing to endorse threshold treatment in every case, the Bilski majority held threshold treatment proper and acceptable.
That point could affect patent litigation. If there is no dispute over the interpretation of claims, the district court could decide a subject-matter challenge to a patent by summary judgment. If the court found no patentable subject matter, it could dispose of the case without the considerable expense and delay of full-blown patent litigation. It could avoid as many as two trips to the Federal Circuit-one for definitive claim construction and a second for review of the district court's decision on infringement and remedies.
The third notable point of Bilski was its treatment of patent claims' breadth. Several earlier Federal-Circuit decisions had deemphasized claim breadth. They had viewed breadth as the sole concern of Section 112, which deals with proper disclosure, rather than Section 101, which governs patentable subject matter. Judge Newman, in her dissent in Bilski, forcefully argued for continuing this same approach.
Yet limiting breadth to Section 112 is a bit like arguing that two plus two cannot be four because one and three are. That fact that one statutory section implicates breadth does not mean another does not also.
Both Sections 112 and 101 necessarily involve a patent's breadth. Disclosure implicates breadth because the claims must be narrow enough to receive support from the disclosed specification. But the claims' breadth also implicates proper subject matter because the claims define the invention.
Before one can determine whether an invention constitutes statutory subject matter, one must know what the invention is. The breadth of a patent's claims provide the answer. Under the new test, breadth also reveals whether the invention may be a fundamental principle divorced from any machine or transformation, or a principle appropriately limited to a specified machine or article transformation. The breadth of a patent's claims is therefore inextricably tied to both the nature of the invention and its status as patentable subject matter vel non.
The Bilski majority applied these principles to the famous eighth claim in Reilly v. Morse, which involved Morse's telegraph machine. That claim covered any use of electromagnetism to transmit intelligible information at a distance, and the focus of the subject-matter inquiry was quite properly its breadth. Unlimited to any particular machine, the claim would have monopolized a whole unexplored field of technology by claiming a scientific principle. Today it would cover not just telegraphy, but telephone, radio, television, satellite communication, and the Internet-none of which Morse invented. Avoiding just that sort of gross overbreadth, which can arise from claiming a fundamental principle, is one of the main statutory and constitutional purposes of Section 101.
The Bilski majority explicitly reaffirmed that analyzing a claim's breadth or scope is part of assessing patentable subject matter under the machine-or-transformation test. As the majority put it, "the use of a specific machine or transformation of an article must impose meaningful limits on the claim's scope to impart patent-eligibility" (emphasis added). In addition, the limits must be real and practical: "the involvement of the machine or transformation in the claimed process must not merely be insignificant extra-solution activity."
The final notable aspect of the majority opinion was its humility and judicial restraint. In the past, the Federal Circuit has been all too ready to adopt its own legal tests and formalistic rules-an approach that often proved troublesome later. So the en banc majority abandoned it. Having assiduously extracted the "machine-or-transformation" test from the Supreme Court's precedent, it recited that test repeatedly and resolved to stick to it. While noting a probable need for future modification, it promised to leave that to the high court.
Greater wisdom underlay this restraint and humility. For the first time, an en banc majority of the Federal Circuit appeared to realize how difficult it is to get patentable subject matter right. As the cutting edges of biology and computer science enter the realm of pure information, a patent law designed for the mechanical and chemical inventions of eighteenth and nineteenth centuries might fail to protect and encourage real technological advance. At the same time, patents on business transactions, fundamental principles and what people do in their heads-let alone law, literature, religion and government-might create harmful monopolies, impede commerce and industry and retard, not promote, technological progress. No formal rule can ever resolve the tension between these two truths. Nor can it reduce the difficulty of charting a safe path between two harmful extremes.
The Bilski majority recognized the tension, acknowledging the nature and the difficulty of the subject-matter inquiry. Within the contours of the four key terms in Section 101 (process, machine, manufacture and composition of matter) and the Federal Circuit's own new test, it recognized the essential nature of the inquiry as part of the federal common-law process. Like other federal common law, the test for patentable subject matter ought to develop case by case. The guidance of a higher court with greater experience in resolving a wide range of difficult common-law issues should be helpful in that development.
Besides the nine-judge majority opinion, there were four others, three nominally dissenting and one concurring. Judges Newman and Rader each wrote separate dissents, rejecting the majority's machine-or-transformation rule as too narrow and too begrudging of incentives for innovation in the information age. Judge Mayer wrote separately to reject the majority's rule as too unclear and uncertain; he would have outlawed business-method patents entirely and overruled State Street and its progeny-a step that the majority failed (at least explicitly) to take. Judges Dyk and Linn wrote a joint concurrence solely to support the majority's opinion and refute Judge Newman's lengthy historical analysis of English and American patent law.
Although Judge Mayer styled his separate opinion a "dissent," he would have disallowed the Bilski patent as one claiming a business method. So his opinion is properly classified as a concurrence in result, albeit with vastly different rationale. Accordingly, the no-patent result in the case, as distinguished from the machine-or-transformation rule, attracted the votes of ten judges, with only Judges Newman and Rader viewing the invention at bar as patentable subject matter. In that limited sense, Bilski was a 10-2 decision.
An appeal in the case, if made, seems unlikely to succeed. The alleged invention was a pure business method, claimed in the most abstract and sweeping terms. The Supreme Court is unlikely to reverse the decision to exclude it as unpatentable subject matter, for such a decision would contravene nearly unanimous scholarly opinion opposing that sort of patent, which Judge Meyer summarized. Rather than take certiorari merely to adjust the Federal Circuit's verbal formula (an approached abjured in Warner-Jenkinson), the high court will probably wait to see how the new verbal formula works in practice before intervening.
