The Mind as Computer Metaphor: Benson and the Mistaken Application of Mental Steps to Software (Part 1)

Go to: Part 1, Part 2, Part 3, Part 4, Part 5

(This series of posts is based on an upcoming paperfor the AIPLA Spring 2016 meeting.)

By Robert R. Sachs1

Of the three recognized judicial exceptions to Section 101—laws of nature, natural phenomena, and abstract ideas—none has proved more resistant to reasoned judicial analysis than the last. From its inception in Gottschalk v. Benson2 to the Supreme Court’s explicit refusal to define the term in Alice Corporation v. CLS Bank International,3 the boundaries of this exception have remained elusive. Rather than principled analysis drawn on well-developed theories in philosophy of language, linguistics, cognitive psychology and other disciplines, the courts have gone “hunting for abstractions”4 to slowly but steadily sweep a variety of different and unrelated constructs into this “murky morass.”5 Thus, abstract ideas are said to encompass fundamental economic practices, methods of organizing human activities, mathematical algorithms, ideas themselves, and finally mental steps. No one seems to notice that these four erstwhile subcategories have little, if anything, in common. Courts now routinely invoke this last subcategory to invalidate claims for software inventions that “can be performed in the human mind, or by a human using a pen and paper.” CyberSource Corp. v. Retail Decisions, Inc.6 The emphasis on can be is intentional and important: it reflects the fundamental shift in the patent eligibility jurisprudence from considering whether the claimed invention was intended in fact to performed mentally (the “factual form” of mental steps) to a hypothetical embodiment of whether it could be (the “fictional form” of mental steps).

It is this shift in framing that has led to the vigorous use of the mental steps doctrine in the field of software patents. Between the June 2014 Alice decision and March 29, 2016, there have been 175 federal court decisions invalidating patents under Section 101, and 24% of those decisions relied upon the “mental steps” doctrine. The eighty-two patents thus invalidated were not limited to suspect categories such as “business methods,” but included electronic design automation,7 computer and database security,8 information retrieval,9 microbiology,10 user interfaces for interactive television,11 telecommunications,12 and digital image management.13

How did the mental steps doctrine come to have such sweeping breadth? The answer lies at the intersection of the popularity of the “mind as computer” metaphor and aggressive advocacy. This paper will examine the development of the mental steps doctrine, focusing in particular on its transformation in Benson from the relatively narrow factual form of the doctrine to an unbounded fictional form of the doctrine. I. The Original Doctrine: Inventions Requiring Mental Steps in Fact

Historically, the “mental steps” doctrine was used only in very narrow circumstances, where an invention as conceived by the inventor and as claimed necessarily required steps performed in the human mind. This “factual form” of the mental steps doctrine arose in cases involving inventions that occurred before the use of computers in business and industrial applications. The patent disclosures thus described the invention in terms of mathematical or other procedures that could only be performed mentally by “head and hand,” or human judgments, at best guided by mathematical or other considerations. That is, there was no disclosure of any way to perform the mathematical operations except by mental operations, at best assisted using existing mechanical devices, such as adding machines. The earliest Court of Customs and Patent Appeals case that expressly applied the mental steps doctrine to method claims is In re Heritage (1945).14 The claims were directed to a method of “producing a porous coated fiber board” including coating samples with varying amounts of material, testing them for sound reduction coefficient, selecting a specimen within the range of coefficients, and using the amount of material applied to the selected specimen as the criterion for further coatings.15 There was no disclosure of any apparatus or machine used to make the selection. Rather, all of these steps had to be carried out by the artisan using his own judgment. The court held that “owing to the fact that claims 1 and 2 are essentially directed to a purely mental process of making a selection of the amount of coating material to be used in coating a porous fiber board in accordance with a predetermined system they do not define patentable subject matter.”16

The C.C.P.A. applied In re Heritage in a line of cases considering mental steps before the advent of claims for computer implemented inventions.17 By the time of In re Prater (1951),18 programmed digital computers were in common use in commercial and scientific settings. The claims before the court included a method of performing spectral analysis on gas mixtures to identify an unknown component with minimum error. Claim 9 recited:

In mass spectrographic analysis where, from a given sample of material there is generated a spectrum function having peaks therein spaced along a mass scale with respect to which the relationship between concentration, contribution factor of each of the m constituents of the mixture and the magnitude of each of the n peaks in said spectrum is represented by a set of m linear algebraic equations and where n is an integer greater than m, the method of selecting for analysis a set of m peaks least susceptible to error in concentration determination which comprises

dividing each said contributing factor for each peak by a normalizing function,

successively generating a determinant function for each said set of peaks,

successively generating output indications of the magnitudes of said determinant functions, and

selecting therefrom the determinant function of greatest magnitude for identification of said peaks least susceptible to error.19

The court distinguished its prior decisions of In re Abrams and In re Yuan, in which there was no physical apparatus to perform the claimed processes.20 That is, the court viewed the mental steps embodiment as a necessary one that required human performance and judgment--and not merely a sufficient one. The Prater court stated that “Although in view of our decision here we find it unnecessary to analyze and/or review in depth the so-called "mental steps" doctrine, it would appear that the disclosure of apparatus for performing the process wholly without human intervention merely shows that the disclosed process does not fall within the so-called "mental steps" exclusion.”21 In support of this narrow interpretation of the mental steps doctrine, the court cited an early treatise on the patent eligibility of software, “Disclosure of apparatus for performing the process without human intervention may make out a prima facie case that the disclosed process is not mental and is, therefore, statutory.”22 Ultimately, the court did invalidate the method claim, but under Section 112, not Section 101, because claim did read on “a mental process augmented by pencil and paper markings,” and “thus interpreted, reads on subject matter for which appellants do not seek coverage, and therefore tacitly admit to be beyond that which "applicant regards as his invention.”23 As to the apparatus claims, the court held that “We do not perceive of any "mental steps" issue in regard to apparatus claim 10. It is quite clear that claim 10, in typical means-plus-function language as expressly permitted by the third paragraph of 35 USC 112, does not encompass the human being as the "means" or any part thereof.”

The question of whether a method claim that recited computer elements was statutory was intentionally left open in Prater. The Court of Customs and Patent Appeals addressed that specific question a few months later In re Bernhart (1951).24 The claims included method and apparatus claims for automatically making a two-dimensional portrayal of a three-dimensional object from any desired angle and distance and on any desired plane of projection. The patent disclosed the use of a general purpose computer and the specific equations used to generate the projections. The method claim recited steps of programming the computer to perform equations and generate the plots.25 The apparatus claims were set forth in means-for form, including “electronic digital computer means programmed” to compute a set of signals values based on a specified equation. Affirming the eligibility of the apparatus claims first, the court stated broadly:

Moreover, all machines function according to laws of physics which can be mathematically set forth if known. We cannot deny patents on machines merely because their novelty may be explained in terms of such laws if we are to obey the mandate of Congress that a machine is subject matter for a patent. We should not penalize the inventor who makes his invention by discovering new and unobvious mathematical relationships which he then utilizes in a machine, as against the inventor who makes the same machine by trial and error and does not disclose the laws by which it operates.26

This is a profound insight into not just the legal jurisprudence of patent eligibility, but into the nature of the creativity and innovation. The inventor who discovers and discloses the law governing the operation of a machine for new purposes makes a greater contribution to the store of human knowledge than one who merely discovers the machine by trial and error. The former teaches principles that can be further developed upon and extended, often to new and unanticipated domains; the latter merely contrives a new artifact, leaving it to others to determine—or perhaps not—the governing relationships. The latter is at best a highly skilled craftsman; the former is a true engineer.27

The court also upheld the method claims:

In the case now before us, the disclosure shows only machinery for carrying out the portrayal process. In fact it is the chief object of the invention to eliminate the drudgery involved in a draftsman's making the desired portrayals. Accordingly, a statutory process is here disclosed. Looking then to method claim 13, we find that it in no way covers any mental steps but requires both a "digital computer" and a "planar plotting apparatus" to carry it out. To find that the claimed process could be done mentally would require us to hold that a human mind is a digital computer or its equivalent, and that a draftsman is a planar plotting apparatus or its equivalent. On the facts of this case we are unwilling so to hold. We conclude that the method defined by claim 13 is statutory, and its patentability must be judged in light of the prior art.28

Here the court expressly denounces the fictional approach to mental steps, since that would require holding that the human mind was the equivalent of a digital computer, and there were no facts in the record (nor could there have been) in support of such a finding. The court affirmed this approach a year later in In re Mahoney (1952), where it upheld method claims for synchronizing a receiver with an incoming bitstream based on comparisons of bit values.29 The court held that the words "bit" and "bit stream," as used in the claims and understood in the art, make “mental performance of the claimed process impossible.”30 The court emphasized “it would be absurd to say that the claims reasonably read on a mentally implemented process. We are aware of no way in which the human mind can operate on such signals.”31

This line of reasoning was further extended by the court in In re Musgrave, in an opinion by Judge Rich.32 The claims there dealt with methods of correcting seismographic signals for timing errors caused by variations in the soil that the signals pass through before being detected at a seismograph.33 The court again rejected the fictional form of the mental steps argument that the claims were ineligible “merely because some or all the steps therein can also be carried out in or with the aid of the human mind or because it may be necessary for one performing the processes to think.”34 Instead of asking whether mental steps were involved, the court proposed the technological arts test.35 Judge Baldwin, the author of Prater, concurred in the result, but expressed grave concern that the latter was an unnecessary change in the law that had “foreseeable problems” in “interpreting the meaning of "technological arts.””36 Nonetheless, Judge Baldwin agreed that “cases before this court have made it clear that there is now only a very narrow scope to this "fearful" mental steps doctrine,” noting that “in reality very little remains of the "mental steps" doctrine.”37

The Court of Customs and Patent Appeals maintained this view of mental steps in In re Foster,38 In re McIlroy,39 and most significantly, in its decision in In re Benson.40 In In re Benson, the USPTO rejected the claims as being directed to mental steps. The court disagreed, and upheld the claims as statutory.41 The court was nonplussed by the argument that a human could perform the claim “manually although in actual practice it seems improbable anyone would ever do that,” because of “speed measured in milli- or even micro-seconds being essential in the practical utilization of such a process.”42 The court’s acknowledgement of the importance of speed of operation in practical embodiments is another example of how the Rich court was sensitive to the realities of actual technology practice—in contrast with the present view of the Federal Circuit that speed of computation is irrelevant.43 The court went so far as to state that even mental performance was irrelevant, because such mental steps were “only to the extent necessary to assure that [the operator] is doing what the claim tells him to do. In no case is the exercise of judgment required or even the making of a decision as between alternatives.”44

Thus, until the Supreme Court’s Benson decision, no court had invalidated claims computer-implemented inventions using the fictional form of the mental steps doctrine, that the claims “could be” performed by a human. Indeed, this approach had been repeatedly rejected by Judge Rich and others on the Court of Customs and Patent Appeals. Instead, the court applied the factual form of the doctrine: claims to a computer-implemented invention were ineligible only if mental implementation was necessary. A disclosure of a programmed computer was sufficient structure to demonstrate the mental implementation was not necessary. To extend the mental steps doctrine into the fictional form, the Court in Benson needed support for the assumption that the operations of a computer are “the same procedures which a human being would perform” by “head and hand.” As will be shown next, this view was set forth by the Solicitor General in its brief to the Court.


1 My thanks to Adam Lewin for his significant contributions and suggestions to this article. 2 409 U.S. 63 (1972). 3 573 U.S. ___, 134 S.Ct. 2347 (2014). 4 CLS Bank Int'l v. Alice Corp. Pty, Ltd., 717 F.3d 1269, 1298 (Rader, J., dissent) (“A court cannot go hunting for abstractions by ignoring the concrete, palpable, tangible limitations of the invention the patentee actually claims.”). 5 MySpace, Inc. v. GraphOn Corp., 672 F.3d 1250, 1259 (Fed. Cir. 2012). Apparently, not much has changed in forty-two years; see In re Musgrave, 431 F.2d 882, 890 (C.C.P.A. 1970) (“whatever law there may be on the subject [of “mental steps”] cannot be attributed to Congress. It is purely a question of case law.That law we, like others, have found to be something of a morass.”). 6 654 F.3d 1366, 1373 (Fed. Cir. 2011) (emphasis added). 7 Synopsys Inc. v. Mentor Graphics Corp., 78 F. Supp. 3d 958 (N.D. Cal., 2015). 8 Kinglite Holdings Inc. v. Micro-Star Int'l Co. Ltd. Case No. CV-14-03009 JVS (PJWx), 2015 U.S. Dist. LEXIS 145121 (C.D. Cal. Oct. 16, 2015); Protegrity USA, Inc. v. Netskope, Inc., Case No. 15-cv-02515-YGR 2015, U.S. Dist. LEXIS 142633 (N.D. Cal., Oct. 19, 2015); FairWarning IP, LLC v. Iatric Sys., Inc., Case,8:14-cv-2685-T-23MAP, 2015 U.S. Dist. LEXIS 81999 (M.D. Fla. June 24, 2015)

9 Collarity, Inc. v. Google Inc., C.A. No. 11-1103-MPT, 2015 U.S. Dist. LEXIS 159031 (D.Del., Nov. 25, 2015). 10 Univ. of Utah Res. Foundation v. Ambry Genetics Corp., 774 F.3d 755 (CAFC Dec. 17, 2014). 11 Netflix, Inc. v. Rovi Corp., No. 11-cv-6591 PJH, 2015 U.S. Dist. LEXIS 92766 (N.D. Cal., July 15, 2015). 12 Stanacard v. Rubard, LLC, 12 Civ. 5176 (CM (MHD), 2015 U.S. Dist. LEXIS 157345 (S.D.N.Y., Nov. 18, 2015); Comcast IP Holdings I, LLC v. Sprint Commc'ns Co. L.P., No. CV 12-205-RGA, 2014 WL 3542055 (D. Del. July 16, 2014); Telenit Techs., LLC v. Alteva, Inc., Civ. No. 2:14-CV-369 2015, U.S. Dist. LEXIS 125991 (E.D. Tex. Sept. 21, 2015). 13 Wolf v. Capstone Photography, 2-13-cv-09573, 2014 U.S. Dist. LEXIS 156527 (C.D. Cal., Oct. 28, 2014); In re: TLI Comm'ns LLC Patent Litig., 87 F. Supp. 3d 773 (E.D. Va. 2015). 14 150 F.2d 554 (C.C.P.A. 1945). 15 Id. at 1172. 16 Id. at 1174. 17 In re Abrams, 188 F.2d 165 (C.C.P.A 1951) (claims ineligible as mental steps where no No specific apparatus or machine was disclosed for performing the steps, thus requiring human performance); In re Yuan, 188 F.2d 377 (C.C.P.A 1951) (relying on Abrams, claims ineligible as mental steps where no specific structure disclosed). 18 415 F.2d 1393 (C.C.A.P. 1969). 19 Id. at 1407. 20 Id. at 1402, 1407

21 Id. 22 Id., citing Kayton, Patent Protectability of Software: Background and Current Law, in The Law of Software 1968 Proceedings B-25 (1968). 23 Id. 24 417 F.2d 1395 (C.C.P.A. 1969). 25 13. A plotting method comprising: (a) a first step of programming the computer to compute the position of planar Cartesian coordinate axes in the given plane relative to the given set of object points, (b) a second step of programming the computer to compute and produce an output defining in sequence the coordinates of the projection of each given point on the plane with reference to the Cartesian coordinate axes, and (c) the step of applying the computer output to the input of a planar plotting apparatus adapted to provide on a plane a succession of straight-line segments that connect between sequential points having positions corresponding to the coordinates computed by the second step. 26 Id. at 1399 (emphasis added). 27 See Dasgupta, Technology and Creativity 12-14 (1996). 28 417 F.2d at 1400 (emphasis added). 29 421 F.2d 742 (C.C.P.A. 1970). 30 Id. at 745. 31 Id. 32 431 F.2d 882 (C.C.P.A. 1970). 33 Id. at 884. 34 Id. at 892. 35 Id. 36 Id. at 895 (Baldwin, J., concurring). 37 Id. 894 (Baldwin, J. concurring; “this court should concern itself only with realities and let the law professors worry about academic problems.”). 38 438 F.2d 1011 (C.C.P.A. 1971). 39 442 F.2d 1397 (C.C.P.A. 1971). 40 441 F.2d 682 (C.C.P.A. 1971). 41 Id. at 688. 42 Id. 43 See Bancorp Servs., LLC v. Sun Life Assurance Co. of Can., 687 F.3d 1266, 1278 (Fed. Cir. 2012) (“[T]he fact that the required calculations could be performed more efficiently via a computer does not materially alter the patent eligibility of the claimed subject matter.”); OIP Techs., Inc. v., Inc., 788 F.3d 1359, 1363 (Fed. Cir. 2015) (“relying on a computer to perform routine tasks more quickly or more accurately isinsufficient to render a claim patent eligible.”). 44 Id. (emphasis added). The acknowledgement that claims would still be eligible even if mental steps were required, so long as the person simply performed the claimed steps without the exercise of judgment provides a basis for the patent eligibility of games (e.g., card games, board games) using standard components (e.g., playing cards, balls, etc.), as well as some business methods.

*The perspectives expressed in the Bilski Blog, as well as in various sources cited therein from time to time, are those of the respective authors and do not necessarily represent the views of Fenwick & West LLP or its clients.