Claim Construction or Statutory Construction?: A Response to Chiang & Solum
“Claim construction” is the process by which a court determines the meaning of a patent’s claims—a process that in turn determines the scope of the covered invention. This process is extremely important because a court must determine what the patent covers before it can determine whether the patent is invalid or infringed.1 Uncertainty over how a given court will construe a patent’s claims is one of the main problems facing patent professionals.2 This problem was on the Supreme Court’s mind in Nautilus, Inc. v. Biosig Instruments, Inc., where the Court overturned the Federal Circuit’s standard for claim definiteness,3 and is also at the forefront of the Court’s upcoming decision in Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., where the Court will consider whether the Federal Circuit, which has exclusive jurisdiction over patent appeals, should continue to review district court claim construction decisions de novoor apply the more deferential clear error standard.4
In The Interpretation-Construction Distinction in Patent Law, Tun-Jen Chiang and Lawrence Solum offer an innovative analysis of the problem of uncertainty in claim construction outcomes.5 First, they propose that constitutional law’s interpretation-construction distinction, which separates “interpretation” of a legal text’s linguistic meaning from “construction” of its legal effect, can be employed to analyze the process through which judges “interpret” and “construct” patent claims.6 Second, by applying the distinction to several canonical claim construction cases in which judges have contested the meaning of claim terms,7 Chiang and Solum convincingly argue that uncertainty in claim construction outcomes is frequently caused by disagreements over policy, not disagreements over the objective meaning of claim language. In other words, claim construction really is about construction, not interpretation, of claims.Third, they conclude that the underlying policy debate in claim construction is a conflict between two distinct theories of claim construction: “textualism,” which pursues the linguistic meaning of claim text and often leads to a broader claim scope, and “anti-textualism,” which “pursues the true invention” and often leads to a narrowing of claim scope.8
Chiang and Solum’s central example, used throughout their article, is the Federal Circuit’s en bancdecision in Phillips v. AWH Corp.9 In Phillips, the judges disagreed over whether the claim term “baffles” should be construed according to its ordinary meaning—a structure—or more narrowly, as a structure extending at a particular angle. The majority adopted the broader construction, holding that “baffles,” when read in light of the description and illustrations in the patent’s specification, did not expressly require “baffles” to have any particular angle.10 Conversely, the dissent argued the court should have affirmed the district court’s narrower construction because the specification indicated that the invention’s intended purpose was to deflect bullets, which could only be done effectively using baffles extending at acute or obtuse angles.11
According to Chiang and Solum, the debate in Phillips was not over interpretation, but over construction. The judges agreed on what the linguistic meaning of “baffles” should be, but not on its ultimate legal meaning.12 The reason for the disagreement, the authors contend, is that the judges applied fundamentally different methodologies: the majority adopted a “textualist” approach, pursuing only the linguistic meaning of claim text; the dissent adopted an “anti-textualist” approach, going beyond the text to find the “true invention.”13 Chiang and Solum further link this methodological debate to a policy debate between judges who prefer broader patents and those who prefer narrower ones. Textualist methodologies, they suggest, tend to produce broader rights than anti-textualist methodologies because claims are drafted “by self-interested patentees” who “write the claim text broadly as a general matter.”14 These disagreements, Chiang and Solum conclude, explain the divided outcome in Phillips15 and are the “most significant source of uncertainty in patent law today.”16
Although Chiang and Solum’s methodological approach is different, their diagnosis is consistent with those of other commentators, who have recognized that the Federal Circuit’s claim construction decisions are increasingly panel-dependent.17 For instance, in a recent empirical study, Thomas Krause and Heather Auyang analyze cases in which judges disagreed over claim construction and conclude that “there are striking differences” among the judges with respect to how they come out in claim construction.18 Specifically, they find that Judges Linn, Clevenger, and Rader (all members of the Phillips majority) are more likely to broaden patent scope, while Judges Lourie and Newman (both members of the Phillips dissent) are more likely to narrow patent scope,19 and that, with the exception of Judge Moore, this distribution aligns with the judges’ pro- and anti-patentee holdings.20 Like Chiang and Solum, Krause and Auyang link these observed differences to the debate over whether judges should “determine what the inventor actually invented” and “limit the claims accordingly” or construe claims according to their linguistic meaning.21
Chiang and Solum’s conclusions concerning uncertainty in claim construction—that claim construction debates are debates over construction, not interpretation, and that the core policy debate concerns patent scope—are persuasive. However, in applying the interpretation-construction distinction, Chiang and Solum should be more attentive to the institutional context in which patent claim construction occurs. Unlike in constitutional law, where judges operate within a broad “construction zone”22 and outcomes are “essentially driven by normative concerns,”23 in patent law, judges do not construct patent claims in isolation. Instead, their actions are constrained by the Patent Act’s requirements. At a conceptual level, the Patent Act restricts judges’ discretion during claim construction. The Act establishes minimum standards of disclosure that patentees must meet when drafting claims.24 Under 35 U.S.C. § 112(a), a patent claim is legally invalid if it is not supported by a specification that includes, at minimum, details of the invention that are sufficient to permit “any person skilled in the art” to make and use it.25 Under § 112(b), a claim is invalid if it does not conclude with “one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or joint inventor regards as the invention.”26 Regardless of whether a judge adopts a “textualist” or “anti-textualist” approach to claim construction, he or she must enforce these minimum standards of disclosure when deciding what a patent’s claims mean.
This is not to say that, in practice, debates over policy do not enter into claim construction and affect outcomes. To the contrary, even though these requirements are statutory, judges have very different views about how much disclosure or definiteness the law requires, and they engage in significant statutory construction when defining the precise contours of § 112. This is consistent with the perception that patent law, like other areas of intellectual property law, evolves largely through the common law method. As Craig Nard puts it, the Patent Act serves as the “enabling statute,” leaving “ample room” for courts to fill in gaps and generate explanatory doctrine.27 We argue that what Chiang and Solum identify as “construction,” as opposed to “interpretation,” of patent claims is better understood as statutory construction of § 112, which, in turn, influences how judges construct claims. On this view, differences of opinion over the precise contours of the disclosure requirement—not an adherence to different methodologies of textual analysis—cause uncertainty in claim construction outcomes.
In Parts I and II, we illustrate this subtle, but important distinction between claim construction, in Chiang and Solum’s terminology, and statutory construction. In Part I, we show that several of the cases Chiang and Solum analyze using the interpretation-construction distinction appear to be the direct result of judicial disagreements over whether § 112(a) contains a separate “written description” requirement. In Part II, we suggest that judicial understandings of § 112(b)’s definiteness requirement have also historically influenced judges’ claim construction decisions, representing another example of statutory construction influencing how judges construe claims. Although, like Chiang and Solum’s, our evidence is limited to individual cases, our analysis suggests that statutory construction of § 112 may be partly responsible for the uncertainty in claim construction outcomes motivating Chiang and Solum’s project. In Part III, we discuss the implications of this statutory view of claim construction for the Supreme Court’s upcoming decision in Teva addressing whether claim construction is a question of law or fact and the amount of deference that should be granted to lower courts’ claim construction decisions.
Chiang and Solum argue that debates over policy are responsible for uncertainty in claim construction and represent the “age-old conflict between textualism and anti-textualism” with respect to patent claims.28 However, many of the cases cited by Chiang and Solum are better read as debates over the proper statutory construction of § 112(a). This section provides:
[A patent’s] specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same . . . .29
Over the course of history, courts have debated the precise contours of this requirement. Since the late nineteenth century, § 112(a) and its predecessors have been read to require that every patent claim be written so as to “enable” a person of ordinary “skill in the art” to practice the invention without undue experimentation.30 This enablement requirement limits the scope of the inventor’s exclusive right by creating a general rule that the inventor cannot claim what he or she cannot clearly describe.31 It also facilitates the patent system’s goal of teaching useful information to others in the field.32
Despite general agreement on this basic framework, § 112(a)’s language has generated significant controversy.33 The most recent debate concerns whether the section requires that a patent specification contain a “written description” of the invention separate from the enabling disclosure discussed above. Pursuant to this written description requirement, a patent must show that the inventor “had possession” of the invention at the time the patent application was filed.34
The Federal Circuit debated this issue for over a decade35 until finally deciding in Ariad Pharmaceuticals, Inc. v. Eli Lilly & Co. that § 112(a) contains a separate written description requirement.36 After discussing the statutory text, Supreme Court precedents, and the distinct roles of the enablement and written description requirements, Judge Lourie, writing for the majority, explained that pursuant to this requirement a patent must show “possession” at the time of filing and “show that the inventor actually invented the invention claimed.”37 Judges Rader and Linn dissented, arguing that a separate written description requirement was not justified on legal or policy grounds.38
As explained in the Introduction, Chiang and Solum classify the Federal Circuit’s leading claim construction case, Phillips,39 where the judges argued over the meaning of the claim term “baffles,” as a debate over claim construction methodology—textualism versus anti-textualism—and policy—broad versus narrow claiming. However, we argue that Phillips is better understood as part of the debate over whether § 112 contains a written description requirement, which is ultimately a debate over statutory construction. Phillips was decided squarely in the midst of this debate.40 More importantly, analysis of the case shows that the two issues—whether the meaning of “baffles” should be narrowly construed according to the embodiments disclosed in the specification and whether § 112(a) requires the patentee to describe, as well as enable, the invention—are intertwined.41 In Phillips and Ariad, the breakdowns of the judges and their rationales for narrowly/broadly construing “baffles” and adopting/rejecting a written description requirement, respectively, were similar. The Phillips majority, joined by Judges Linn, Clevenger, and Rader, the last of whom wrote the Ariad dissent, construed “baffles” broadly to cover structures beyond those the specification revealed the inventor possessed.42 The dissent, written by Judge Lourie and joined by Judge Newman, construed “baffles” narrowly. Consistent with his construction of § 112(a) in cases establishing the written description requirement,43 Judge Lourie sought to limit the claim according to what the inventor “actually invented.” Because “the specification contains no disclosure of baffles at right angles,” he wrote, the claim could not be construed to include this embodiment.44
In other words, the Phillips dissent narrowed the claim because the claim did not, in their reading of § 112(a), meet the statute’s requirements; the Phillips majority broadened the claim due to their perception that it did. This is a debate over statutory construction. Classifying it simply as “claim construction” versus “claim interpretation,” while this might be analytically accurate, misses the meat of what the debate was about—namely, how much disclosure does the Patent Act require?
Retractable Technologies, Inc. v. Becton, Dickenson & Co.45offers a more recent example of the same phenomenon. As in Phillips, the judges’ constructions of the claim term, “body,” mirrored their constructions of § 112(a). Judge Lourie, now writing for the majority and joined by Judge Plager, concluded that “body” referred to only a one-piece syringe body because “a construction of ‘body’ that limits the term to a one-piece body is required to tether the claims to what the specifications indicate the inventor actually invented.”46 This language, although written in the context of claim construction, precisely tracks Judge Lourie’s language in Ariad explaining that the specification must contain a written description of the invention in addition to an enabling disclosure. In his concurring opinion, Judge Plager agreed with Judge Lourie’s view that the court should seek to construe claims in accordance with what the patentee “actually invented,” and he made the connection between this standard and § 112(a)’s written description requirement explicit, writing that “claims cannot go beyond the actual invention that entitles the inventor to a patent. For that we look to the written description.”47 “[M]ore stringent rules,” he concluded, were “need[ed]” to “control” “the curse of indefinite and ambiguous claims, divorced from the written description . . . .”48 In Judge Plager’s view, and likely those of some members of the majority, the “actually invented” standard for claiming and the written description requirement for disclosure go hand in hand.
In the dissent’s view, in contrast, neither standard is mandated by § 112 or necessary as a policy matter. “The ordinary and customary meaning of ‘body,’” wrote then-Chief Judge Rader, “does not inherently contain a one-piece structural limitation.”49 “[T]he claims themselves, not the written description portion of the specification, define the patented invention.”50 It would be “improper to import limitations from the specification into the claims, and th[e] court ha[d] expressly and repeatedly warned against confining claims to specific embodiments of the invention set forth in the specification.”51 Consistent with his opposition to a separate written description requirement and his view that § 112(a) requires less disclosure than the majority demanded, Chief Judge Rader favored a broader construction of the claim term and effectively rejected the “actually invented” standard for claiming.52
We are not alone in perceiving statutory construction of § 112(a) as relevant for debates over claim construction. Following Retractable, Judge Moore dissented from the court’s denial of rehearing en banc. Review was necessary, she argued, because the panel’s decision in Retractable demonstrated the court’s continuing disagreement over whether claims should be construed according to what the inventor “actually invented.”53 Significantly, Judge Moore noted the connection between the debates over the “actually invented” standard in claim construction and the written description requirement.54
As just explained, Chiang and Solum’s illustrations of “claim construction” and “claim interpretation” in cases like Phillips and Retractable55 are better read as disagreements over the proper construction of § 112(a). Section 112(b)’s definiteness requirement presents another example of statutory construction’s influence on claim construction.
In 1870, Congress amended the Patent Act to specify that inventors must “particularly point out and distinctly claim” their inventions.56 This language is preserved in the modern definiteness requirement.57 In Merrill v. Yeomans, decided six years after the Patent Act’s amendment, the Supreme Court praised the “developed and improved” claiming provision,58 stating that the “distinct and formal claim” is “of primary importance, in the effort to ascertain precisely what it is that is patented.”59 A patent’s claims, the Court reasoned, should give the public “fair notice” of what the patent covers and ensure that “[t]he public should not be deprived of rights supposed to belong to it, without being clearly told what it is that limits the rights.”60
Today, courts continue to apply this rationale, requiring that claims provide the public “clear notice”61 of what is claimed by the patent and of “what is still open to them.”62 Upon motion, courts will invalidate claims that do not meet this standard.63 However, like § 112(a), § 112(b)’s language has generated significant disagreement. The Supreme Court’s decision in Nautilus, Inc. v. Biosig Instruments, Inc. represents the most recent example of judges disagreeing over “how much imprecision § 112[(b)] tolerates.”64 In Nautilus, the Court overturned the Federal Circuit’s definiteness standard, which held that a claim was invalid for indefiniteness only if it was not “amenable to construction” or “insolubly ambiguous.”65 Instead, the Court announced a different, apparently stricter standard and held that a claim is invalid if it “fail[s] to inform, with reasonable certainty, those skilled in the art about the scope of the invention.”66
Jurists have disagreed over the degree of claim definiteness mandated by § 112(b) for more than a century. Merrill was an early attempt to grapple with this question. In Merrill, the parties disputed whether the claim covered a method for making hydrocarbon oils or the hydrocarbon oil product.67 The Supreme Courtheld that the claim covered the method,68rejecting the petitioner’s arguments that the patent be construed “liberal[ly]” to cover the product because the claim language did not provide the public with “fair notice” of this construction.69 The petitioner’s construction, the Court wrote, would undermine the purpose of the Patent Act’s new distinct claiming requirement by allowing patentees to exclude the public through “ambiguous language or vague descriptions.”70Justice Clifford dissented tersely, concluding that the claim, “when properly construed,” covered the product, not the method.71
Chiang and Solum cite Merrill to support their conclusion that most linguistic ambiguity can be “resolved through interpretation” using ordinary interpretive tools.72 We emphasize instead the role of the definiteness requirement in influencing the Court’s construction of the claim at issue. The Court was not only concerned with resolving linguistic ambiguity, but also with the effects of this ambiguity on public rights.73 Upon considering the public goals underlying the amended definiteness requirement, the Court rejected the petitioner’s construction because it was inconsistent with those goals.
To us, Merrill suggests that judges’ understandings of the meaning and purpose of § 112(b)’s definiteness requirement, like their understandings of § 112(a)’s written description requirement, can affect their claim construction decisions and result in divergent claim construction outcomes. Again, framing this disagreement over statutory meaning as one over “claim construction” versus “claim interpretation” misses what the debate was about—namely, how definite § 112(b) requires claims to be. As discussed in the next Part, it also ignores an important pathway for resolving the debate—clarifying § 112(b)’s meaning or amending the statute.
This “statutory” view of claim construction, in which judicial disagreements over the law affect judges’ constructions of patent claims, has implications for the ongoing debate over whether claim construction is a question of law or a question of fact. In Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., the Supreme Court will decide whether the Federal Circuit should continue to review district court claim construction decisions de novo, or whether it should apply the more deferential clear error standard, applied to district courts’ factual findings.74 In Markman v. Westview Instruments, Inc., the Court held that claim construction is a “mongrel practice,” with both factual and legal components, but that judges, not juries, should determine claim meaning because of judges’ greater expertise in interpreting legal documents and establishing uniformity.75 In Cybor Corp. v. FAS Technologies, Inc., the Federal Circuit interpreted Markman to hold that claim construction is a “purely legal” issue and that appellate courts should therefore not defer to a “trial judge’s asserted factual determinations incident to claim construction.”76
In Teva, the petitioners have asked the Court to overrule Cybor and hold that although claim construction is “ultimate[ly]” a question of law, it involves “subsidiary” questions of fact, which must be reviewed under the clear error standard.77 The government also recommends this outcome.78 In contrast, the respondents argue that claim construction is a “pure question of law,”79 or, alternatively, is “ultimate[ly]” “a legal question, subject to de novo review” with “all disputes, factual” or otherwise “subsumed within” that question.80
Many commentators disagree with the current institutional allocation and support granting greater deference to district courts because, they argue, district courts are “better positioned” to examine expert testimony and other evidence regarding how skilled artisans would understand the patent’s claims.81 Chiang and Solum, in contrast, defend the status quo, arguing that claim construction really is a legal matter. Applying the interpretation-construction distinction, they argue that because most disputes over claim meaning involve construction and not interpretation, claim construction should continue to be treated as a question of law and not reviewed deferentially.82
Our view, which recognizes the role of statutory construction in influencing claim construction, partially aligns with Chiang and Solum’s. Although we agree with the Teva petitioners that claim construction involves many factual issues,83 we also see legal issues lurking in judges’ claim construction decisions. Admittedly, the precise boundary between fact and law is very difficult to determine in this context.84 However, we identify at least two significant legal questions over which judges currently disagree and which appear to affect claim construction outcomes.
The first question, discussed in Part I, is whether § 112(a), which Ariad held to contain a written description requirement separate from enablement, also requires claims to be construed according to the “actually invented” standard.85 The second question, discussed in Part II, is whether and how § 112(b) and its definiteness requirement should influence claim construction.86 For example, does § 112(b), as construed by the Supreme Court in Nautilus, require courts to review more expert evidence in order to determine whether a patent’s claims would inform those skilled in the art “with reasonable certainty” about the scope of the invention?87
These questions are questions of statutory interpretation and are, therefore, purely legal.88 The answers would greatly help to define the kinds of “subsidiary” factual questions that district courts could or should consider when construing claims and which should be treated as questions subject to deferential review on appeal. Although the Supreme Court is unfortunately unlikely to address these questions in Teva, the Court (or the Federal Circuit) should provide answers in subsequent cases.89
Our view also leads us to be less pessimistic than Chiang and Solum about the prospect of deference to district courts on issues that do not clearly involve statutory construction. The authors argue that giving district courts an open-ended license to construe claims based on their policy preferences would increase uncertainty and disuniformity,with “hundreds of district court judges across the nation each following their own individually preferred methodologies in the individual cases that come before them.”90 However, if, as we have suggested, the disagreements over claim construction methodology that concern Chiang and Solum are really disagreements over § 112’s disclosure and definiteness requirements, then resolving those legal disagreements should theoretically guide district courts during claim construction and increase certainty in claim construction outcomes. This, of course, would require the Federal Circuit and the Supreme Court to agree on a claim construction standard and district courts to decide cases consistently under this standard, but agreement and consistency would be needed regardless of the standard of review. If the courts are not up to the task, then Congress can amend the statute to provide greater clarity. This is how patent law has historically progressed,91 and, we believe, this is how lawmaking should work within a well-functioning statutory framework.
Chiang and Solum’s article is a welcome example of cross-disciplinary legal scholarship, and their framework is useful for shedding light on the policy debates underlying claim construction decisions. However, applying the interpretation-construction distinction to patent law requires taking into account the statutory context in which claim construction is performed. In this Response, we have argued that the debate the authors identify—between “textualist” and “anti-textualist” approaches to claim construction—is better classified as a debate over statutory construction, with judges construing claims in light of their divergent constructions of § 112’s disclosure and definiteness requirements. We have shown that recognizing the prevalent role of statutory construction in framing judicial debates during claim construction is consistent with case history and leads to several useful suggestions for improving certainty in claim construction outcomes and resolving whether claim construction is a question of law or fact. Most importantly, it provides a clearer and more optimistic framework for improving the law of claim construction than Chiang and Solum provide. The fact that a statute limits both patentees’ discretion in drafting claims and judicial discretion in construing claims distinguishes patent law claim construction from constitutional construction, where clear rules and consensus are neither possible nor necessarily desirable.
Camilla Hrdy is the Center for Technology, Innovation and Competition Fellow at the University of Pennsylvania Law School and a Visiting Fellow at the Yale Information Society Project. Ben V. Picozzi is a member of the Yale Law School J.D. Class of 2016 and a Student Fellow at the Yale Information Society Project. They would like to thank T.J. Chiang for his generous comments on this Response and for traveling to New Haven to discuss his article with the authors and other members of the Patent Law Reading Group on December 12, 2013. They would also like to thank Erica Newland and the other editors at the Yale Law Journal for their helpful comments on earlier versions of this Response.
Preferred Citation: Camilla Hrdy & Ben V. Picozzi, Claim Construction or Statutory Construction?: A Response to Chiang & Solum, 124 Yale L.J. F. 208 (2014), http://www.yalelawjournal.org/forum/claim-construction-or-statutory-construction.