In the past week, two Federal Circuit decisions – Lighting Ballast v. Philips Electronics, and Williamson v. Citrix – have taken up the issue of the sufficiency of functional claim language, and the scope of claims interpreted according to the “means-plus-function” claim form under 35 USC 112(f).
This topic has been simmering for some time, and academics and reform advocates have been clamoring for the courts to pick up this issue. The Federal Circuit now appears ready to entertain these arguments, and a new front has opened in the “patent reform” wars. This topic bears much discussion, as it is likely to be at least as contentious and important as Alice.
The federal courts have been busy, of late, issuing decisions under 35 USC 101 in view of the Alice decision. Their recent decisions present a 61% invalidation rate – which is actually significantly lower than similar metrics from last year – while the PTAB has maintained an incredible 100% invalidation rate of cases reviewed under 35 USC 101. It remains clear, as I wrote last year, that no one in the patent community knows the extent to which the trajectory of the Alice decision will affect the scope of patent law.
Additional uncertainty abounds, both within the USPTO and examining corps and within the patent community, as to the manner in which Alice is to be applied. Much of this uncertainty stems from the indefiniteness of the original Alice decision, which explicitly avoided providing clear definitions of the advocated test, but only examples of analytic factors that the Court hoped to be instructive. The Court thus turned the key question of patent-eligibility under 35 USC 101 into a “smell test,” of the “I know abstractness I see it” variety.
Given that unprecedented level of activity over key questions of patent validity, it is shocking that the Federal Circuit has spontaneously opened a new front in the “patent reform” wars. Two cases in the past week have undertaken the topic of whether certain types of functional claim language are to be interpreted as “means-plus-function” claims under 35 USC 112(f).
In Williamson v. Citrix, the Federal Circuit reviewed the claims of U.S. Patent No. 6,155,840, which recited a method of “conducting distributed learning” that involved “a server comprising… a module for providing a first graphical display” that included some visual controls and elements. Surprisingly, this case turned on the issue of 35 USC 112, rather than patent-eligibility under 35 USC 101. The opinion focused on whether this claim limitation invoked the means-plus-function format of 35 USC 112(f), despite the absence of the term “means.” Finding in the affirmative, the court reined in its previous precedent that such an absence raised a “strong presumption” that 35 USC 112(f) does not apply. Instead, the Federal Circuit held that 35 USC 112(f) applies unless “the words of the claim are understood by persons of ordinary skill in the art to have a sufficiently definite meaning as the name for structure.” Even if the term “means” is absent, 35 USC 112(f) can apply if “the claim term fails to ‘recite sufficiently definite structure’ or else recites ‘function without reciting sufficient structure for performing that function.'”
Notably, the Williamson opinion features fifty instances of the term “structure,” and yet declines to provide any more specific definition of what “structure” constitutes. Some notable patent reform advocates have been quick to cite Williamson as requiring a physical structure. Yet, both the text of 35 USC 112(f) and the Williamson decision describe the requisite supporting structure for means-plus-function claims as “the recital of structure, material, or acts… corresponding to the claimed function and equivalents thereof.” The issue therefore appears to be the level of detail – i.e., a “means” for achieving a result can be supported by a recitation of particular “acts” that indicate how the result is achieved – rather than a rigid requirement of physicality. Notably, the Federal Circuit’s analysis hinged on this point:
While portions of the claim do describe certain inputs and outputs at a very high level (e.g., communications between the presenter and audience member computer systems), the claim does not describe how the “distributed learning control module” interacts with other components in the distributed learning control server in a way that might inform the structural character of the limitation-in-question.
This key phrase – how the module interacts with other components – is certainly interesting: this is the very definition of software architecture. Software architecture is routinely abstracted from physical interactions – where the data resides, where the modules live, etc. – and strictly focuses on the functional relationships of elements: the communications protocol, the interaction sequence, the logical structure of the data. The implication that means-plus-function interpretation can be avoided by disclosing and claiming the specific logical architecture and functional interactions is distinctly at odds with how opponents of software patents would like 35 USC 112(f) to be used.
The second case, Lighting Ballast v. Philips Electronics, is at the other end of the spectrum, involving the presentation of the term “voltage source means” in the context of a lighting device without any attempt to recite structural elements that could be used. Here, the Federal Circuit reached the exact opposite decision, finding that the term “voltage source means” was a fully sufficient recitation of structure to stand on its own, and did not invoke 35 USC 112p6 means-plus-function construction. The Federal Circuit agreed with the district court that even if the specification limited the scope of the term “voltage source” to the class of devices known as rectifiers, the term still provided sufficient structure in the field of electronics.
This conclusion is baffling for several reasons. First, consider the vast array of devices that act as “voltage sources”:
- A voltaic cell, i.e., a traditional battery
- A liquid-fuel generator
- A wall outlet connected to a power utility line
- A solar panel
- A thermoelectric generator (a Sterling or Peltier engine)
- A lightning rod
- A hydroelectric dam or water wheel
- A fusion reactor
- A potato
Similarly, even if we limit “voltage sources” to “rectifiers,” the field of electrical engineering has identified a large class of devices as “rectifiers.” The introductory chapter of Sedra’s Microelectronic Circuits lists no fewer than seven distinct types of rectifiers – using completely different structures and techniques to provide different kinds of rectification – all of which are united only by the fact that they achieve the functionality of “rectification.”
Surprisingly, even the Lighting Ballast decision repeatedly describes the ostensibly definite “structure” as any device that achieves a particular functionality:
- “The district court cited testimony from an expert for Lighting Ballast, both of whom testified that one of skill in the art would understand the claimed ‘voltage source means’ to correspond to a rectifier which converts alternating current to direct current, or other structure capable of supplying useable voltage to the device.”
- “The district court determined that ‘while the “voltage source means” term does not denote a specific structure, it is nevertheless understood by persons of skill in the lighting ballast design art to connote a class of structures, namely a rectifier, or structure to rectify the AC power line into a DC voltage for the DC input terminals.‘”
- “The district court further noted that ‘it is clear to one skilled in the art that to provide a DC voltage when the source is a power line, which provides an AC voltage, a structure to rectify the line is required.’ … We defer to these factual findings.”
In other words, both the district court and the Federal Circuit found the terms “voltage source means” to be so structurally specific as to avoid 112(f) interpretation, despite the presence of the term “means” – even though both courts could only describe such “structure” as any device that exhibits the specified functionality.
These cases are logically irreconcilable. 112(f) cannot be held to include claims that do not recite “means” when not enough detail is provided – and also exclude claims that do recite “means” when the term covers an entirely undefined range of devices that exhibit a particular functionality.
Sadly, the only explanation for this inconsistency is that 112(f) is to be applied very expansively in the area of software, yet extremely conservatively for other technologies. The courts’ disrespect of software, as a misunderstood and openly denigrated area of technology 1, marches on.
Functional claim language is a big, big issue – perhaps bigger than Alice – and is no easier to resolve than four decades of pontification (with no end in sight!) over the solitary term “useful” in section 101. Much has been written about functional claiming, and much of that writing is dead-wrong – technically, legally, and historically. Expect this post, which barely reveals the tip of the iceberg in the emerging dysfunctionality, to be the first of many to highlight these deficiencies.
It is astounding that the Federal Circuit has chosen, sua sponte, to open this issue to debate. It is astounding that the Federal Circuit would do so in the midst of the throes of Alice and its progeny. And it is astounding that the Federal Circuit’s first efforts in this area, mere days apart, are unresolvably inconsistent. Taken together, these indicators bode poorly for the future stability of patent law.
- Note the following comment from the Alice oral argument transcript:
JUSTICE KENNEDY: Well, let me put it this way. If you describe that to a second year college class in engineering and said here’s - here’s my idea, now you go home and you program over this weekend, my guess is - my guess is that that would be fairly easy to program.
It is unimaginable that a judge or justice would base the patent-eligibility of a claim element in any other area of technology on speculation over whether “a student could make it over the weekend.”
My earlier post on the disrespect of software by the courts focused on the courts’ peculiar tendency to be dismissive about the merit of software inventions. Drugs and circuits and engines – those things are deemed to be mysterious boxes of technology that are the product of complex, hard sciences, while the prevalence and ease-of-use of software make it seem effortless – more artistic than technical. This attitude reflects a woeful misunderstanding of the depth and complexity and intensity and difficulty of computer science – one that will only be resolved over decades, as our current crop of judges is replaced with technologists with actual experience in the software industry. ↩