For almost 100 years, Canadian Courts have consistently held that the interpretation of a patent claim is a question of law. In almost every context, questions of law are reviewed on a de novo basis, without deference. As the Canadian courts of appeal routinely describe claim construction purely as a "question of law,"1 in practice it is unclear whether the factual findings of the trial judge are afforded deference in claim construction (for example, in the assessment of the common general knowledge that informs the meaning of claim terms). In light of the United States Supreme Court opinion in Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc.2 and recent developments in Canadian law, this principle is ripe for reconsideration in Canada.
The SCOTUS Teva decision: Subsidiary Fact-Finding Attracts a Deferential Review
The United States Court of Appeals for the Federal Circuit has long reviewed a district court judge's interpretation of a patent claim without deference—de novo—as a pure question of law. But far from being divorced from fact, claim construction is a practice with "evidentiary underpinnings." Those underpinnings inform the meaning of claim terms in the context of the patented invention and to those skilled in the associated technology. Claim construction establishes the "metes and bounds" of a claim, which in turn defines the limits of a patentee's exclusive rights. Consequently, claim construction can profoundly affect the ultimate issue of every patent case: whether the defendant is liable for infringement or whether the asserted claim is even valid. De novo review of a Court's claim construction yields a higher reversal rate than review under a more deferential standard. It may therefore have enormous economic impact; reversal dictates the difference between competition in the market versus none, and liability for damages versus none. In this regard, the circumstances of Teva v. Sandoz are emblematic.
The dispute in Teva pertained to the meaning of the phrase, "molecular weight," which defined an ingredient used in the claimed method of manufacturing a drug. While a commonplace phrase to the lay scientist, the Court consulted expert evidence to resolve the issue as to whether or not "molecular weight" admitted of three different meanings to those skilled in the relevant science. If so, the claim was fatally indefinite—failing to establish the metes and bounds—and therefore invalid.
The district court resolved the dispute in favour of a single meaning for "molecular weight," holding the claim valid and infringed. The appellate court reversed: the patent was held to be invalid and not infringed. As was typical, the Federal Circuit reviewed de novo the district court's claim construction, including the underlying facts.
The Supreme Court vacated the holding, announcing that the factual findings underlying the claim construction should not be reviewed de novo. In matters of claim construction, there is no "exception from the ordinary rule governing appellate review of factual matters."3 Appellate courts are therefore now required to review all factual findings subsidiary to claim construction under the "clearly erroneous" standard. The Supreme Court explained that where the background science is consulted, or the history of how the term is understood, these are "evidentiary underpinnings", and this "subsidiary factfinding must be reviewed for clear error on appeal".4
Deference, however, will not always be due. As the majority stated, "when the district court reviews only evidence intrinsic to the patent (the patent claims and specification, along with the patent's prosecution history), the judge's determination will amount solely to a determination of law, and the Court of Appeals will review that construction de novo."5 Further, "the ultimate question of construction will remain a legal question", to be reviewed de novo.6
That the factual finding "may be nearly dispositive" does not render the subsidiary question a legal one.7 However, this near-dispositive character, coupled with deferential review, renders the new standard of review an important change and one that, serendipitously, is ripe for consideration in Canada.
Canada's Adherence to Claim Construction as a Question of Law: All Root, No Ground?
In Canada, many patent cases cite the Supreme Court's decision in Whirlpool Corp v Camco Inc8 for the principle that claim construction is a question of law. This has been accepted and applied in the Federal Court, with little detailed scrutiny. This principle finds its roots in Supreme Court case law and the Interpretation Act, but with little express consideration of the appropriate standard of review.
Whirlpool was a landmark case, in part because it rejected the long-standing doctrine of substantive infringement—i.e., a claim could be infringed if there was literal infringement, or if the activity of the defendant appropriated the "pith and substance" of the invention. Substantive infringement gave the trial judge tremendous discretion, and made the infringement analysis unpredictable. In rejecting substantive infringement, courts were directed to construe the claims in a "purposive" way and apply that same construction to determine patent infringement and validity.
Whirlpool did not include a detailed discussion of appellate standards of review. Specifically, the Court did not address whether an appellate court can or must review the factual findings of the trial judge on a de novo basis. For example, should the trial judge's findings as to the common general knowledge of the skilled person to whom the patent is addressed (which must inform if not define all claim construction) be afforded deference? The Court's direction is: "[h]owever, claims construction is a matter of law for the judge, and he was quite entitled to adopt a construction of the claims that differed from that put forward by the parties." While no authority was cited here, the statement is consistent with the Supreme Court's 1934 decision in Western Electric Co v Baldwin International Radio of Canada.9 There, the Supreme Court was clear that claim construction is a question of law and not for an expert or jury to decide.
In this respect, Western Electric accords with modern practice. In contrast with the United States, almost all Canadian patent cases are commenced in the Federal Court of Canada, where juries are prohibited.10 Further, Western is consistent with the modern principle that experts should not opine on the ultimate issue,11 as this is the role of the Court. However, other aspects of Western Electric have not withstood the test of time. Specifically, the Court in Western Electric stated that after the claims have been construed, the next step is to determine whether the invention has been "in substance" taken by the defendant—the approach to determining infringement that the Supreme Court expressly rejected in Whirlpool. If substantive infringement is no longer good law, then other issues in Western Electric, such as standard of review, are surely equally amenable to reconsideration.
The principle that claim construction is a question of law is also rooted in statute. The term "letters patent" is included within the definition of "regulation" in Canada's Interpretation Act. As the name suggests, the Interpretation Act provides guidance as to how government "enactments" are to be read (e.g., when they come into force and computation of time periods). The Interpretation Act also requires that an enactment will be given such fair, large and liberal interpretation as best ensures the attainment of its objects (a direction that is very consistent with what the Supreme Court said in Western Electric). However, while the Interpretation Act provides valuable guidance as to how a patent should be read, it is silent on standards of appellate review.
Should Canadian Practice be Considered De Novo?
Recent decisions addressing the role of appellate courts suggest that describing and treating claim construction as a question of law (where no deference is shown to the factual findings of the trial judge) is increasingly out of step with Canadian jurisprudence.
Errors of law are reviewed de novo, without deference under the "correctness" standard, because "the principle of universality requires appellate courts to ensure that the same legal rules are applied in similar situations," and to allow the appellate Court to fulfill its "law-settling" role."12 To overturn factual findings, the error must be palpable and overriding. This more deferential standard of review is colourfully described as follows: "[w]hen arguing palpable and overriding error, it is not enough to pull at leaves and branches and leave the tree standing. The entire tree must fall."13
Questions of mixed fact and law generally attract the standard of palpable and overriding error, unless the trial judge made an extricable error (of law) in the characterization of the legal standard or its application.14 The Supreme Court described questions of mixed fact and law as those that involve the application of a legal standard to a set of facts. This precisely describes the exercise of claim construction, but this is not how the exercise is described in the case law.
Claim construction necessarily requires two steps: the trial judge (1) reviews the relevant law on claim construction and then (2) applies that law to the "facts" of the case, i.e., what the patent states and claims is the invention, scope of exclusivity in light of the understanding of the skilled person. Arguably, because patents are written documents, their interpretation, like statutory interpretation, is within the province and expertise of the Court. But statutes, or under the Interpretation Act, "regulations", are not necessarily apt comparators or designations for patents.
Turning back to the United States Teva decision, the Supreme Court contrasted the fact-finding exercise that underlies statutory interpretation—a pure question of law—with the claim construction exercise. While statutes address themselves to a general public and concern themselves with facts related to a "reasonably broad set of social circumstances", patents typically "rest upon consideration by few private parties, experts, and administrators of more narrowly circumscribed facts related to specific technical matters".15 By contrast, the Court has "repeatedly compared patent claim construction to the construction of other written instruments such as deeds and contracts".16
A patent is more contractual than statutory in nature. In the ordinary course, the scope construction of the patent affects a handful of potential competitors, and not the public at large. Patent interpretation will almost always involve resolving competing expert evidence to establish the circumstances informing the interpretation. These circumstances may not cry out for the general, law-settling role of the appellate Court. Rather, under Canadian law, the circumstances favour a more deferential standard in the context of claim construction.
The Supreme Court of Canada recently abandoned the historical approach of reviewing, as a question of law, the rights and obligations under a written contract. Whereas, with limited exceptions, parol (or extrinsic) evidence was inadmissible for purposes of interpreting a contract, now the circumstances surrounding the contracts—facts—may always be considered, and attract a more deferential standard of review.17
Similarly, in administrative law, there has been a clear shift to affording deference to administrative decision-makers, even on questions of law. Beginning with Dunsmuir18 and continuing with subsequent decisions,19 the scope of questions of law where de novo review is warranted is becoming increasingly limited. Reasonableness has become the dominant standard of review—provided the result comes within a range of acceptable outcomes, it will not be disturbed on appeal. While the Supreme Court has been clear that standards of review for decisions of administrative tribunals must not be conflated with standards of appellate review,20 the approach is instructive as to how appellate courts view their role.
The Facts Speak for Themselves
Unlike appellate review of administrative and contract law, the Supreme Court has not provided clear direction as to how appellate courts should review a trial judge's factual findings when considering claim construction. In practice, the Federal Court of Appeal has described claim construction as a question of law, however it has been unclear whether the underlying facts have been swept up into a de novo review, or whether a demonstration of palpable and overriding error has been effectively required. The underlying facts are seldom discussed in terms of standard of review. Claim construction is often deemed "correct" or "proper" or "incorrect." However, deference to the underlying facts is either not express or not apparent, which seems increasingly out of step with the modern approach to appellate review in Canada. We anticipate that, over time, appellate review of claim construction will be clarified by the Supreme Court, and the practice will converge with the recent result in the United States.
1 With passing reference to the principle that "[a] judge's assessment of the expert evidence will not be reversed on appeal absent palpable and overriding error". Bell Helicopter Textron Canada Limitée v. Eurocopter, société par actions simplifiée 2013 FCA 219 at 74-75
2 Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc. 574 US____ (2015) (Case No. 13-854)
3 Id. at p. 6
4 Id. at p. 12
5 Id. at pp. 11-12
6 Id. at p. 13
7 Id. at p. 13
8  2 SCR 1067
9  SCR 570
10 Federal Courts Act (RSC, 1985, c. F-7), s. 49
11 R. v Mohan,  2 SCR 9
12 Housen v Nikolaisen,  2 SCR 235, 2002 SCC 33
13 Canada v South Yukon Forest Corporation, 2012 FCA 165
14 Housen v Nikolaisen at ¶36
15 Teva at p. 11
16 Id. at p. 11
17 Sattva Capital Corp v Creston Moly Corp, 2014 SCC 53
18 Dunsmuir v New Brunswick,  1 SCR 190
19 Alberta (Information and Privacy Commissioner) v Alberta Teacher's Association, 2011 SCC 61; McLean v British Columbia (Securities Commission) 2013 SCC 67
20 Agraira v Canada (Minister of Public Safety and Emergency Preparedness), 2013 SCC 36
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.