Overview Edit

The mental steps doctrine, which had a rather inauspicious beginning under that label in Halliburton Oil Well Cement Co. v. Walker,[1] experienced a series of novel elaborations in the hands of the Patent Office as the foundation of its attack on program-related process claims. A quotation from Prater I provides an appropriate introduction to the continuing attempts by the Patent Office to bootstrap a variety of patentability disqualifications onto the "mental steps" bar:

The examiner's position before the board, as set forth in his Answer, includes a rejection of the method claims for failure to comply with 35 U.S.C. §§101, 102 and 112. A sequence of initial steps of reasoning adopted by the examiner in this rejection, is: (1) that if the invention is to fall within a statutory class under Section 101 it must be as a “process”; (2) that the claims are readable upon a mental process; and (3) claims to a mental process are unpatentable (in support of which proposition the examiner quotes portions of In re Abrams, 188 F.2d 165, 38 C.C.P.A. 945 (1951)). The examiner then proceeds with development of the rejection in two alternative forms:

In the first form of the rejection under 35 U.S.C. §§ 101 and 102, the examiner argues, on the one hand, that if the novel part of a claim is readable on subject matter which is unpatentable because it is outside the statutory classes of patentable subject matter, then the claim as a whole is unpatentable under 35 U.S.C. § 101; and, on the other hand, if the claim reads on the physical process of a person marking paper during calculation by hand, then the claim is unpatentable under 35 U.S.C. § 102.

In the second form of the rejection, under 35 U.S.C. §§ 101 and 112, the examiner argues that if the claim covers subject matter outside the statutory classes (as well as subject matter within the statute) then the claim fails to particularly point out and distinctly claim the invention, as required by 35 U.S.C. § 112.[2]

The thrust of the earlier, non-program cases was that a process which specifically claimed steps disclosed as performed mentally and particularly if requiring interpretive procedures was nonstatutory subject matter. On the other hand, these early cases established that "[a] method is not per se unpatentable because its practice requires that the operator thereof must think"[3]— an observation perhaps too obvious to mention.

Program-related processes do not, in fact, require human thought, interpretive or otherwise, in performing those steps associated with the computer. The real focus of the "mental steps" doctrine in this area has been upon process claims which were readable in some fashion on mental implementation (or its equivalent pencil and paper practice). Early hopes that the "mental steps" objection could be avoided by the disclosure of fully automatic machines were not fulfilled, as the Government's Prater argument revealed.[4] Of course, claims that definitely read on nonstatutory subject matter would necessarily fail. Therefore, attempts to rescue programs from rejection on this ground followed two basic approaches: (1) denying that the mental processes in question are or ought to be relevant to the subject matter issue and (2) asserting that claims to a program-related process do not read on mental activity at all.

The line of reasoning on the irrelevance of mental processes is as follows: (a) infringement is a tort; (b) torts require "acts" and harm to the patentee; (c) mental practice, which is neither an "act" nor harmful, cannot infringe; and (d) mental practice should therefore not be an issue, since "monopolization" by patent necessarily entails the power to bring infringement actions."' In a similar vein, it has been argued that claims reading on mental practice in the abstract should nevertheless stand if no person would, in fact, practice the invention mentally."' The Prater invention provides an obvious example of such a claim, since the claimed process involved selecting one subset of ten equations from among 184,756 possible subsets.[5]

A more adventuresome task was to argue that the claimed thought is not, or ought not to be considered, nonstatutory subject matter. The U.S. Supreme Court held as early as 1863 that "ideas" were not patentable.[6] More recently, it was asserted that, by virtue of their alleged relationship to thought, program-related process patents raised constitutional questions under the first, ninth, and tenth amendments. The response of the pro-patent community was a series of efforts to distinguish patentable mental activity from that which is nonpatentable. One variant was to suggest coverage for all processes, mental or otherwise, as long as they are commercial or industrial in nature.

The following rumination illustrates this line of reasoning:

Why is pure mathematics non-statutory? I have found no authority but I think the reason is that historically, mathematics has not been a "useful art" in the constitutional sense because, historically, there has been no immediate correlation between the discovery of a mathematical relationship and the implementation of that relationship in the physical, tangible world of machines.[7]

The second principal response to the "mental steps" objection was to deny that the claim read on a mental process at all. The strategy was to stress the differences between thought and program processes; specifically, by pointing out that information is physically represented in computers and that machine processes do not, in fact, duplicate human thought processes.

Patent proponents, attempting to widen the conceptual gap between mental and program processes, also stressed the similarities between software and hardware implementations. The following dicta from Ex parte Egan[8] was frequently cited on that point:

We agree that the process under consideration is properly analogous to a method of operating a computer, since the charts employed are quite analogous to a preconstructed computer. The method operations in operating a computer are distinct from the method of computation itself. It is perfectly possible to have a patentable process in which apparatus is used in a particular way to get a useful result.[9]

Reliance on this passage had its own problems, however. Not only were "programs" not mentioned, but Egan distinguished the use of the charts from the method of their construction (which may well have been unpatentable as "mathematical computations"). Thus, the quoted dicta did not address the patentability of a “program" conceived as a method of configuring a "preconstructed computer." In any case, the primary employment of analogies between hardware and software were in defense of apparatus, not method, claims.

The particular uses of the "mental steps" doctrine discussed in this article do not begin to exhaust its flexibility as a springboard for assertions that program processes were nonstatutory. The doctrine evolved from a relatively narrow, albeit challenged, preoccupation with human participation in claimed processes, to at least moral support for assertions that claims to totally automatic process steps threaten our “intellectual patrimony” and sought monopolies on “scientific truths.”


  1. 146 F.2d 817, 64 U.S.P.Q. (BNA) 278 (9th Cir. 1944)(full-text).
  2. 415 F.2d 1378, 1381, 159 U.S.P.Q. (BNA) 583, 586-87 (C.C.P.A. 1968)(full-text)(emphasis in original).
  3. Ex parte Bond, 135 U.S.P.Q. (BNA) 160, 162 (Pat. Off. Bd. App. 1961).
  4. In re Prater (Prater II), 415 F.2d at 1399-1400, 162 U.S.P.Q. at 546-47.
  5. 415 F.2d at 1395 n.13, 162 U.S.P.Q. at 543 n.13.
  6. Burr v. Duryea, 68 U.S. (1 Wall.) 531, 570 (1863)(full-text).
  7. Popper, Current Status of Patent Protection for Programmable Processes, 7 Pat. L. Ann. 37, 42 43 (1969) (emphasis in original).
  8. 129 U.S.P.Q. (BNA) 23 (Pat. Off. Bd. App. 1960). The opinion upheld the patentability of a method for measuring the depth of well bores that required making entries on a series of specially constructed charts.
  9. Id. at 26.
Community content is available under CC-BY-SA unless otherwise noted.