Professor Tun-Jen Chiang is an Associate Professor of Law at the George Mason University School of Law. Below, he responds to Camilla Hrdy and Ben Picozzi’s commentary on The Interpretation-Construction Distinction in Patent Law. As with the previous two posts in this series, the real name policy applies to comments on this post.
Thanks to Jason and Dennis for inviting me to post, and to Jason for the extremely flattering introduction. I should make clear at the outset that I am only speaking for myself, not my coauthor, on this blog, though of course the article represents our joint views.
A brief recap for those who have not read (or only skimmed) the article. The article’s primary claim is conceptual. The claim is that in making decisions pursuant to text, courts necessarily engage in a two-step process: (1) they read the text and try to discern its linguistic meaning, i.e. the meaning of the language, qua language, to members of the intended audience, and (2) they determine the legal effect of the text by rendering a decision. We call the former “interpretation” and the latter “construction.” The key point is that linguistic meaning and legal effect are not always the same, and therefore uncertainty over the legal scope of patent claims does not necessarily arise because of linguistic ambiguity.
A couple of clarifications at the outset. The first is that our definition of “construction” is simply giving text a legal effect—it does not care whether that legal effect is consistent with statute, precedent, and text or not. Camilla’s articulation (where judges “‘construe’ the claims according to their normative beliefs about what the claims’ legal effect should be”) potentially gives the impression that we define construction as only when judges act as super-regulators who engage in free-floating first-order policy analysis. This is not correct. Even the most restrained judge who makes decisions strictly in accordance with statute, precedent, and/or claim text engages in construction, because his decisions necessarily give the claim text a legal effect.
Second, our claim is conceptual, not normative. Our claim is not that interpretation is good or that construction is bad; our claim is that both are intrinsic to the activity of claim analysis, so there is really no “good” or “bad” about the matter. And because our claim is conceptual, our doctrinal prescriptions are very limited. We do not opine, for example, whether courts should give more or less weight to dictionaries, or specifications, or anything else.
The sum is that we are not trying to push a textualist or plain meaning agenda, and we are not trying to suggest that judges who engage in construction are judicial activists who ought to be impeached. Our goal is to simply provide more clarity in analysis and a better diagnosis for what ails patent claims. Specifically, we argue that the uncertainty in patent claims is not caused by linguistic ambiguity, but by methodological disagreement among judges about whether to construe claims according to the patentee’s actual inventive contribution, or instead according to the linguistic meaning of claim text written by the patentee (which is likely to be broader than the real invention).
It is at this point that Hrdy and Picozzi’s response comes in. Hrdy and Picozzi agree with us that methodological disagreement, rather than linguistic ambiguity, underlies disagreements over claim construction. But, they argue, the root source of methodological disagreement in patent law is more narrow than in other areas of law. That is, according to Hrdy and Picozzi, in areas such as constitutional law, “judges are relatively free to interpret and construct the text based on their prior beliefs about which policy outcomes or which form of interpretive methodology (e.g., textualism or anti-textualism) is superior.” But in patent law, they argue, judges are constrained by the Patent Act, and so disagreements over claim construction do not stem from free-floating beliefs about desirable policy outcomes, but instead are tethered to more mundane and legalistic disagreements about the proper statutory construction of the disclosure requirements of 35 U.S.C. § 112. In particular, they argue that “judges’ approaches to claim construction closely followed—and were likely influenced by—their constructions of section 112(a)’s written description requirement and section 112(b)’s definiteness requirement.”
It might perhaps surprise Hrdy and Picozzi to discover that I agree that judges’ approaches to claim construction largely parallel, and are likely influenced by, their positions on § 112. Yet I do not think this defeats my argument. As an initial matter, I would note that, even if Hrdy and Picozzi are correct that disagreements over claim construction stem from narrow legalistic disagreements about the proper understanding of § 112, that would not contradict the core claims of the article—the basic point that disputes about claim analysis are mostly disputes about legal methodology rather than linguistic meaning would still stand. In that sense Hrdy and Picozzi’s argument is less a refutation and more in the nature of a follow-on contribution. But I also think there is a deeper disagreement: Implicit in Hrdy and Picozzi’s con-law-is-different argument seems to be the contention that disagreements about § 112 are not based on beliefs about desirable policy but instead are narrow, mundane, and legalistic debates unencumbered by ideological priors. I think this contention is flawed and that their evidence is more consistent with an alternative hypothesis, namely that judges show a consistent pattern across several doctrinally-distinct issues—claim construction, § 112(a), § 112(b)—because they are influenced by their underlying beliefs about desirable policy (specifically, beliefs about desirable patent scope) in all of them. Judges who favor broad patents tend to construe claims broadly and read § 112 narrowly; judges who favor narrow patents tend to do the opposite. Rather than proving that patent law is not like con law, I submit that the Hrdy and Picozzi’s evidence tends to show that it is.