A few years ago, law professor Mark Lemley wrote a paper arguing that one key way to deal with the problem of software patents was to get the courts to go back to actually applying existing law that effectively forbade the use of “functional claiming.” In short, “functional claiming” means using a patent to claim ownership over an entire function — i.e., traveling through time, buying something with one click — rather than a specific invention on how to do the function. Existing patent law (namely 35 USC 112(f)) says that patents can have functional claiming, but with very specific limitations:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
This is legal speak, but in English, it’s supposed to mean that you can put a functional claim in your patent, but it can only cover what’s described in the actual patent specification. In short, you can claim the function, but you then need to lay out the explicit cases where it applies, by saying what it’s a “means” for. So, “a means for traveling through time” would be a functional claim, but it would only be allowed if you then explained the explicit case in which that applied, such as “using this particular machine built out of wood and string and in order to prevent catastrophe” or whatever.
And here’s where patent lawyers got sneaky, and the courts let them get away with it. They basically started putting language into patents to avoid invoking 112(f). Specifically, the courts became so focused on the magic word of “means” that rather than saying “a means for x” patent lawyers would just write “a computer programmed to” — but would then make the same functional claim: i.e., “a computer programmed to let people travel through time.” Because that patent didn’t use the term “means” the courts argued that it wasn’t functional claiming, and thus it didn’t need the more limited specification. This is — for fairly obvious reasons — a big problem. Especially in the software patent world.
Thus, by using sneaky language, software patents (especially) were able to do very broad functional claiming, which is where many of the patent disputes come in today. And the courts completely let them get away with this.
Well, until yesterday (mostly).
In what has the potential to be an important ruling, the appeals court for the Federal Circuit (CAFC) (perhaps sick of always being wrong on patent cases) went the other way and rejected such bogus functional claiming in the Williamson v. Citrix case concerning some old patents from At Home — but only takes one step towards fixing the problem where it could have gone much further.
Here, CAFC explicitly rejected its own earlier precedents, starting with the Lighting World case, in which it allowed that the absence of the word “means” was enough to get around the limitations of 112(f). The court no longer accepts that as a “strong presumption” that 112(f) is not implicated:
Our consideration of this case has led us to conclude that such a heightened burden is unjustified and that we should abandon characterizing as “strong” the presumption that a limitation lacking the word “means” is not subject to § 112, para. 6. That characterization is unwarranted, is uncertain in meaning and application, and has the inappropriate practical effect of placing a thumb on what should otherwise be a balanced analytical scale. It has shifted the balance struck by Congress in passing § 112, para. 6 and has resulted in a proliferation of functional claiming untethered to § 112, para. 6 and free of the strictures set forth in the statute. Henceforth, we will apply the presumption as we have done prior to Lighting World, without requiring any heightened evidentiary showing and expressly overrule the characterization of that presumption as “strong.” We also overrule the strict requirement of “a showing that the limitation essentially is devoid of anything that can be construed as structure.”
Unfortunately, the court doesn’t go as far as it could have gone. It still grants a presumption that the lack of the word “means” suggests that the limitations of 112(f) don’t apply, and thus you wouldn’t have to specify the “means.” All it does is lower the standard to “rebut” that argument, making it easier for someone to argue, instead, that even though you didn’t use “means” you really were trying to do the same thing in order to get a functional claim (while avoiding the limitations).
That’s useful, but still will allow many functional claims that don’t use “means” to pretend they’re not functional claims. As noted in a partial dissent by Judge Reyna, this ruling still leaves in a big problem in the way “means” is interpreted. That is, while this ruling rejects this attempt to get around the avoidance of the use of “means” to avoid the limitations in 112(f), it still makes it pretty easy for others to do so, by still focusing on this idea that “means” is a magic word that suddenly brings 112(f) into play. Instead, as Citrix argued in the case (and as others have suggested), it would be much better to not look for magic words, but rather to just recognize when someone is making a functional claim and rejecting it if there is not a truly limited specification to go with it.
Still, it’s nice to see CAFC rejecting its own bad standards that have helped make broad patents more powerful. This is a step in the right direction, and one that hopefully will go further in future rulings.