By Jason Rantanen
Teva recently filed its merits brief in Teva v. Sandoz (previous PatentlyO discussion here and here). It’s main argument on the issue of claim construction is that claim construction involves making findings of fact and Rule 52(a)(6) of the Federal Rules of Civil Procedure states that “[f]indings of fact, whether based on oral or other evidence, must not be set aside unless clearly erroneous.” In support of the first part of this argument, the brief points to determinations about the perspective offered by a person of ordinary skill in the art and the use of extrinsic evidence to resolve ambiguities; these, the brief argues, are inherently factual determinations for which deference should be given to the district court.
A key issue will be whether these types of judicial determinations are actually factual findings, or whether they are something else. Certainly reviewing courts make determinations about some types of historical evidence all the time – one need only look at Justice Breyer’s discussion of legislative history and Congressional intent in Aereo for recent example – without invoking deference. These determinations aren’t limited to “purposivists”: Justice Scalia’s originalist approach to interpreting legal statutes, for example, involves looking to how the text would have been understood at the time it was issued, something that can involve historical linguistic research. Figuring out whether the “facts” involved in claim construction are akin to something like legislative history or the historical determinations involved in an originalist approach, or instead are the type of “facts” encompassed by Rule 52 is central to the invocation of that rule. Much of Teva’s brief seeks to tackle this challenge, arguing that the facts involved in claim construction are exactly the kinds of facts that are encompassed by Rule 52.
Of these arguments, I found the most compelling to be the point that even as determinations such as the knowledge of a person of ordinary skill in the art are factual and reviewed with deference in one context (such as obviousness and enablement), they are reviewed without deference in claim construction. It will be interesting to see how Sandoz responds.
One important tension that the brief fails to resolve, however, is the key question of how deference on factfinding plays into decisions about claim construction. If the ultimate determination of the meaning of claim terms remains one of law, does that basically invite an obviousness-like determination wherein certain subsidiary issues – such as the knowledge possessed by a PHOSITA and content of extrinsic evidence – are reviewed with deference, but the reviewing court makes the final determination, weighing each of the various pieces of information, without deference? Or is it something else?
The brief does seem to indirectly offer a vision of how deference to factual findings should work, but it strikes me as leaving no place for the other side of the mongrel practice: the legal component. The vision offered by the brief appears in Part IV., on pages 53-54, where Teva argues that the Federal Circuit erred by not granting deference to the district court on the “understanding of Figure 1, of SEC technology,
and of the prosecution history.” In other words, Teva’s view of deference manifests as the lens through which the interpreter of the legal doctrine looks at not just the extrinsic evidence (the SEC technology), but also the patent document itself and the prosecution history. If this is the correct approach to deference – that the reviewing court must give deference to the district court’s interpretation of the patent document and prosecution history – it is hard for me to see anything left that isn’t entitled to deference. (Maybe the ordinary meaning of the claims themselves? But that’s “the ordinary meaning to a person of skill in the art,” so that doesn’t work.). This “lens” approach to deference seems to go too far, in my mind, and is fundamentally at odds with the way other factual findings and deference work in patent law doctrines such as nonobviousness and enablement.
The brief also challenges the various rationales for reviewing factual issues in claim construction de novo: “Markman requires it” argument, the interpretation of legal documents is a question of law argument, and the uniformity argument. It also argues that allowing de novo review of facts on appeal produces poorer decisions and is costly to the patent system, citing to the reversal rate data. (Although this latter point seems to depend heavily on how deference is actually implemented – it would seem that the de novo review of certain subsidiary elements of claim construction, alone, would have only a very marginal effect on decision quality and litigation cost).
You can download a copy of the brief here: Teva’s Opening Brief