Software corresponding structure

In American patent law, claim elements drafted as means for performing a function – e.g., “means for adding two numbers” – read on the structure corresponding to that function disclosed in the specification “and equivalents thereof.”  35 USC § 112 para. 6.  “Structure disclosed in the specification is ‘corresponding’ structure only if the specification or prosecution history clearly links or associates that structure to the function recited in the claim.”  Medtronic v. Advanced Cardiovascular Systems, Inc., 248 F.3d 1303, 1311 (Fed. Cir. 2001).  Nonetheless, the hunt for corresponding structure in the specification can often be difficult.  The Federal Circuit long ago said that if there is no corresponding structure for a means plus function element, the claim is indefinite.  In re Donaldson Co., 16 F.3d 1189, 1195 (Fed. Cir. 1994) (en banc) (“[I]f one employs means-plus-function language in a claim, one must set forth in the specification an adequate disclosure showing what is meant by that language.  If an applicant fails to set forth an adequate disclosure, the applicant has in effect failed to particularly point out and distinctly claim the invention . . . .”).

It has been quite rare for claims to be found indefinite for lack of corresponding structure.  For example, in Atmel Corp. v. Information Storage Devices, Inc., 198 F.3d 1374 (Fed. Cir. 1999), the means plus function element was “high voltage generating means disposed on said semiconductor circuit for generating a high voltage from a lower voltage power supply.”  The panel majority found that the title of an article cited in the patent, “On-Chip High Voltage Generation in NMOS Integrated Circuits Using an Improved Voltage Multiplier Technique,” might be deemed to supply sufficient corresponding structure, referring as it did to a voltage multiplier.  The panel majority consequently reversed the summary judgment of indefiniteness and remanded, presumably for trial.

Only in the disfavored technology of software has Donaldson’s “no corresponding structure” indefiniteness found its home.  “For computer-implemented means-plus-function claims where the disclosed structure is a computer programmed to implement an algorithm, the disclosed structure is not the general purpose computer, but rather the special purpose computer programmed to perform the disclosed algorithm.  Thus the patent must disclose, at least to the satisfaction of one of ordinary skill in the art, enough of an algorithm to provide the necessary structure under § 112 P. 6.”  Finisar Corp. v. DirectTV Group, Inc., 523 F.3d 1323 (Fed. Cir. 2008) (invalidating claims at issue); Aristocrat Technologies Australia Pty Limited v. International Game Technology, 521 F.3d 1328 (Fed. Cir. 2008) (same); Net MoneyIn, Inc. v. Verisign, Inc., 545 F.3d 1359 (Fed. Cir. 2008) (same); Blackboard, Inc. v. Desire2Learn, Inc., 574 F.3d 1371 (Fed. Cir. 2009).  Adequate flowcharts or pseudocode in the specification would surely have avoided these invalidations.  However, the patent attorneys who prosecuted the patents in suit in those cases had apparently not foreseen that an “algorithm” would someday be required of them.

More recently, in Typhoon Touch Techs. v. Dell, Inc., No. 10-1589 (Fed. Cir. Nov. 4, 2011), the panel gave us a taste of how little extra disclosure would have saved the patents in Finisar, Aristocrat, NetMoneyIn, and Blackboard. The claim term at issue was “means for cross-referencing responses to said inquiries with possible responses from one of said libraries.”  The panel found sufficient corresponding structure in the following:  “the memory of the portable computer stores a data collection application and has locations for storing data entered manually by touching the touch sensitive screen”; “the CPU of the portable computer executes the application and processes the manually entered data pursuant to the application”; “Cross-referencing entails the matching of entered responses with a library of possible responses, and, if a match is encountered, displaying the fact of the match, otherwise alerting the user, or displaying information stored in memory fields associated with that library entry.”  To my mind only the last of these provides any “algorithm” support in the sense of the Aristocrat line of cases; it provides an algorithm consisting of just two steps:  matching and displaying.  Typhoon Touch can thus plausibly be read to hold the following:  all that is needed to escape indefiniteness for a software means plus function element is to recite in your specification at least two steps that would be carried out in accomplishing the function of the claim element, and clearly link those steps to the function.

Before I became a patent attorney, I used to write software for a living.  Based on that life experience, I believe that in the real world of software development, it is relatively rare to say what software does by specifying the algorithm(s) it executes.  Software is normally specified by saying what it accomplishes from the user’s perspective.  For example, when I submitted software that I was in the process of writing to a design review, I would put up on the screen slides of what the user would see when running the program, not flowcharts or other descriptions of what my code would do internally.

In the cases that imposed the algorithm requirement for software means plus function elements, the judges appeared to be concerned to avoid “pure functional claiming” (as stated in Aristocrat).  They were afraid that without the proper kind of corresponding structure in the specification, the means plus function element might read on anything that accomplishes the recited function.  However, the recitation of further functional requirements in the specification, for example screen displays or data formats, normally limits the permissible structures to accomplish the function of the means plus function element.  Such recitations thus prevent the feared reading of the claim element on everything that accomplishes the recited function.

For these reasons, the holdings of cases like Finisar and Aristocrat seemed to me undesirable as a matter of policy when they were handed down:  they judge descriptions of software in patents by a standard that requires them to be different from descriptions of software in the real world.  It is thus good that Typhoon Touch suggests that the algorithm component of the description of a software means plus function element in a patent can be small and brief.  Of course, as with any other Federal Circuit case, with Typhoon Touch it will be necessary to see what subsequent panels make of it before drawing any definitive conclusions.

Flavio M. Rose, Irell & Manella LLP