Yesterday, I went through the history of the ContentGuard v. Apple and ContentGuard v. Google cases. I talked briefly about the Markman process; today, I’d like to go into more detail on what Markman is, how it works, how Markman affected the ContentGuard cases, and why it’s such an important issue in patent litigation in general.
Patent litigators will frequently spend a lot of time on Markman because it often decides who wins or who loses. After you get a Markman order, you know what arguments you’ll be able to make at trial. It can be so important that many cases settle right after Markman, with one side or the other determining that going on with the case is unwise.
The ideal is that the Markman hearing and claim construction order determines the meaning a term would have had to a skilled person when the patent was invented. The reality is that both sides argue for meanings of claim terms that provide them with an advantage, whether it’s making it easier to argue that the claim isn’t infringed or harder to argue that the patent is invalid. And winning the judge over to your view on even one term can be the difference between infringement and non-infringement.
Let’s take a look at the Markman process, particularly with respect to the ContentGuard cases, to illustrate why it’s so important.
Markman v. Westview
Ironically, Markman v. Westview Instruments didn’t actually have anything that looked like the procedure we refer to as a Markman hearing. In Markman (the case), the problem was that the claim term wasn’t construed by the judge, but instead the jury was allowed to come to their own conclusions about what the language of the claim meant to a skilled engineer. The Markman case determined that that was the wrong approach. The meaning of the words in a patent claim are for judges to interpret and construe, not for juries. (I’ll use the terms construe and construction essentially interchangeably with define and definition; construe is patent lawyer-ese for defining a word or phrase.)
This led to the development of the Markman hearing, in which a judge hears evidence and arguments about the meaning of claim terms, and eventually issues a Markman order (also referred to as a claim construction order) which provides a definition of the meaning of disputed terms in the patent.
In order to determine what the claims mean, courts look at intrinsic evidence and extrinsic evidence. Intrinsic evidence is the evidence that’s part of the patent—the words of the claims, the specification of the patent, and the history of the back and forth with the Patent Office during prosecution. Extrinsic evidence is helpful evidence that’s not part of the patent—dictionary definitions, academic articles, statements made by the inventors that are unrelated to the patent but relate to the technology, test results, and any other evidence outside the patent. Extrinsic evidence only comes into play if the intrinsic evidence isn’t sufficient to provide a clear definition.
The evidence is used, along with a number of rules for interpreting patent terms developed over the years, to come to some sort of definition of what the claim terms mean. This definition is what’s set down in the claim construction order, and it’s the definition the jury will be instructed to use.
ContentGuard: Who Decided What Terms Were Defined?
ContentGuard, like most patent cases, had a Markman hearing. The Eastern District of Texas, where the trials were held, has Local Patent Rules which set forth a very well-defined Markman process.
- The plaintiff and defendant exchange a list of terms that they think the judge should construe.
- The plaintiff and defendant exchange proposed definitions for each identified term on the list, as well as identifying any extrinsic evidence they might rely on at the hearing to support their proposed definition.
- The parties negotiate to see if they can agree on any definitions or if they can agree that certain terms just don’t need to be construed because they aren’t important enough.
- The parties provide a joint list of terms to the court, including the terms, any evidence they’re going to rely on.
- The parties then file formal briefs with the court supporting their arguments, and then have an opportunity to file a response to the opposing side’s brief.
- Armed with their briefs, the lawyers for each side show up in Marshall, Texas, and try to convince the judge that their definition better follows the evidence and rules.
In other words, the answer to who decides is that the lawyers on each side decide, based on what they think will help their case. This whole process takes months from start to finish and the rest of the case keeps going the entire time (including the expensive discovery process.)
ContentGuard: What Was Defined?
In the interests of space (and readability), I’m not going to look at each of the 54 disputed terms that were construed in the ContentGuard cases (the judge’s claim construction order is 144 pages just to contain all the terms and his reasoning behind each definition.) Instead, I’ll focus on the one that is a primary focus in ContentGuard’s appeal—“usage rights.”
“Usage rights”, in the ContentGuard patents, refer to some set of rights that define how the protected content is handled. However, ContentGuard and the defendants disagreed about exactly what that meant. Their proposals are shown below.
One key dispute is whether usage rights are statements in a language, or simply an indication. The other key dispute is whether a usage right must be permanently attached to a digital work.
The significance of these disputes is pretty clear. An “indication” could be just about anything; “statements in a language” is much narrower, requiring a more formal structure and even implicitly requiring some kind of DRM language. Similarly, “permanent attachment” implies a close connection and co-location between the usage right and the protected content; leaving that out allows the indication to simply refer to the content, not necessarily being attached to it. So a permanently attached usage right might have to be incorporated into the file storing a particular copy of Sting’s Songs from the Labyrinth, while a non-attached usage right might be on a server somewhere indicating that all copies of Peter Gabriel’s “In Your Eyes” should be treated in a particular way.
The court handled the dispute over “language” first. As I said above, first we look at the words in the claims. In this case, that wasn’t helpful, so the court moved on to the specification of the patents. The specifications described “usage rights” as “a language for defining the manner in which a digital work may be used or distributed, as well as any conditions on which use or distribution is premised.” The specification also talked about usage rights being in a language, or specified via a ‘rights code.’ Judge Gilstrap, faced with these various statements as to how a usage right could be defined, determined that it didn’t explicitly require a language.
Next, the court turned to whether the usage right had to be “permanently attached.” Again, the words in the claims didn’t answer the question, so the specification had to be consulted. The specification, among many other statements, said that a “key feature of the present invention is that usage rights are permanently ‘attached’ to the digital work” and that it was “fundamental to the present invention that the usage rights are treated as part of the digital work.” When a court sees words like “fundamental” and “key feature,” those tend to be signals that the patent applicant was describing what their invention really was. Given this, some offhand references in the specification about merely associating the usage rights were determined not to be relevant. In addition, the court examined ContentGuard’s statements to the Patent Office during prosecution, when it said that a prior art reference which talked about storing the rights within the digital file “attached” the rights. Given this consistent statement, Judge Gilstrap determined that the usage right had to be attached to the content; however, he rejected the idea that it had to be a permanent attachment as unsupported by the patent.
Based on that, the Court came to a middle definition. Judge Gilstrap determined that usage rights are “indications that are attached, or treated as attached, to [content] and that indicate the manner in which the [content] may be used or distributed as well as any conditions on which use or distribution is premised.”
A PTAB Post-Script
After the term “usage rights” had been construed in court, ContentGuard filed a motion to amend their claims in one of the PTAB challenges to their patents. As part of that motion, they told the Patent Office that the construction Judge Gilstrap had issued was a correct construction. (Yes, that’s the same construction ContentGuard is now saying is wrong and should be overturned on appeal.)
You can’t do that. In fact, there’s a long line of cases about how you can’t do that, going back (at least) to Southwall. You can’t construe claims “one way in order to obtain their allowance and in a different way against accused infringers.” But that’s exactly what ContentGuard is trying to do here.
ContentGuard: Why Does It Matter?
ContentGuard is appealing the construction of “usage rights.” ContentGuard argues that Judge Gilstrap’s construction of “usage rights” was wrong, and that they deserve a new trial as a result. Apple and Google argue that the construction was correct, and that the jury’s verdict of non-infringement should remain untouched.
ContentGuard, in challenging the construction, points to the same evidence from the specification that they used in front of Judge Gilstrap. ContentGuard also complains that, if their patent is interpreted to require attachment, then a product they built would be excluded. But that doesn’t matter—your product doesn’t describe what the words in your patent mean. Essentially, ContentGuard is saying Judge Gilstrap got it wrong, and that the specification is clear that attachment isn’t required. Apple and Google, on the other hand, in addition to the arguments that Judge Gilstrap adopted in requiring “usage rights” to be attached, or treated as attached, also point out that ContentGuard is making inconsistent arguments between the Patent Office and the courts.
If ContentGuard is right, and the construction was wrong, then Apple and Google still might win their cases. They have other arguments for non-infringement. But if ContentGuard is wrong, and the construction holds up (as it should)? Then ContentGuard loses their infringement case. That’s what I mean when I say Markman can determine the outcome of a case.
But even if they lose on infringement, ContentGuard might still be able to keep their patents and sue other people and other products down the road. Tomorrow, I’ll talk about the issue Google and Apple are appealing—whether ContentGuard’s patents are valid.