- The Court suggested that the terms in the Patent Act should bear ordinary lay meanings of the term "process" rather than how they have been traditionally understood in the field of patent law. Justice Stevens commented that this is absurd because if one were to take this literally, anything that constitutes a "series of steps" would be patentable so long as they were novel, non-obvious and sufficiently particularized.
- Justice Stevens comments that Justice Kennedy's opinion uses language that seems inconsistent with the centuries-old reliance on the machine-or-transformation criteria as a clue to patentability (not a test). In particular, this is evident when the Court suggests that the test may operate differently when addressed to new technology.
- Justice Stevens states that it is not clear how the Court arrives at the decision that the Bilski patent is an abstract idea and thus not patentable. Justice Stevens cites that the court states that this determination is "clear" from the cases of Benson, Flook and Diehr, but Justice Stevens is clearly unsatisfied with such limited and underdeveloped reasoning. He is especially concerned because the Court appears to have artificially limited the Bilski patent claims to that just of 'hedging' and discounted the application's discussion of what type of data to use and how to analyze that data. "Why the Court does this is never made clear", writes Justice Stevens, "One might think that the Court's analysis means that any process that utilizes an abstract idea is itself an unpatentable, abstract idea. But we have never suggested any such rule, which would undermine a host of patentable processes." Although the majority opinion did recognize that the Bilski application was phrased broadly, this itself does not mean it is should therefore be constituted as unpatentable. Claim specification, wrote Justice Stevens, was determined by Section 112, not 101, and to invoke this reasoning at the 101 stage would itself be calling into question the Court's prior decisions. Concluding, he stated:
"The Court, in sum, never provides a satisfying account of what constitutes an unpatentable abstract idea. Indeed, the Court does not even explain if it is using the machine-or-transformation criteria. The Court essentially asserts its conclusion that petitioners' application claims an abstract idea. This mode of analysis (or lack thereof) may have led to the correct outcome in this case, but it also means that the Court's musings on the issue stand for very little."These three broad criticisms and one final damning paragraph out of the way, Justice Stevens sped away into the main thrust of his argument.
"What is particularly incredible about the court's stated method of interpreting Section 101 is that the Court deviates from its own professed commitment to the "ordinary, contemporary, common meaning" approach by accepting a role for the "atextual" machine-or-transformation "clue" [as well as accepting] that we have "foreclose[d] a purely literal reading of Section 101" by holding that claims that are close to "laws of nature, natural phenomena, and abstract ideas" do not count as processes under Section 101. . . It is strange to thing that the very same term must be interpreted literally on some occasions, and in the light of its historical usage on others."
"In fact, the Court's understanding of Section 101 is even more remarkable because its willingness toe xclude general principles from the provision's reach is in tension with its apparent willingness to include steps for conducting business. The history of patent law contains strong norms against patenting these tow categories of subject matter. Both norms were persumably incorporated by Congress into teh Patent Act in 1952."