“Secondary” Evidence of Obviousness Is Not Secondary in Importance

Two types of evidence are normally adduced in the obviousness inquiry: evidence relating to the circumstances of the invention, such as long-felt need and commercial success, and evidence of expert witnesses as to how a person skilled in the art at the time would have approached the problem. It is now normal to refer to the former as “secondary” evidence. There is nothing wrong with this as a label, but in Mölnlycke v Procter & Gamble Ltd, [1994] RPC 49 (CA) at 113, the English Court of Appeal stated that “[t]he primary evidence will be that of properly qualified expert witnesses who will say whether or not in their opinions the relevant step would have been obvious to a skilled man having regard to the state of the art. All other evidence is secondary to that primary evidence,” and that secondary evidence “must be kept firmly in its place. It must not be permitted, by reason of its volume and complexity, to obscure the fact that it is no more than an aid in assessing the primary evidence.” In other words, “secondary” is not merely a label, but indicates that such evidence is intrinsically of lesser significance than the evidence of expert witnesses. This view, which has since been repeated in a some English decisions, is wrong both in principle and as a matter of Canadian law. The true rule, in Canada, as well as Australia, and indeed in England, is that the court must consider all relevant factors including circumstantial evidence, giving weight to each according to the circumstances of the case: see e.g. 2007 FCA 217 at [27]. It is nonetheless useful to distinguish these two types of evidence, as each is subject to different frailties: assessment of obviousness based on the evidence of expert witnesses is subject to the hindsight bias, while establishing causation is the main problem with circumstantial evidence. While there is no formal hierarchy of evidence in Canadian law, the general tendency is the opposite of that suggested by the “primary / secondary” labels. When circumstantial considerations point strongly in the direction of inventiveness, an argument in favour of obviousness based on expert opinion evidence will rarely prevail.

The “secondary” terminology originated with the 1966 US Supreme Court decision in Calmar v Cook Chemical, the companion to Graham v John Deere 383 U.S. 1, 26 (1966): see John Duffy’s excellent article “A Timing Approach to Patentability” (2008) 12 Lewis & Clark L. Rev. 343 at 364. It began to be used in Canadian and English courts about 1990, but has become common in Canada only in the last decade. Prior to Cook Chemical it was routine in the US to refer to “secondary” evidence as “objective” evidence, while in Canada there was no consistent terminology.

The first use of the “secondary” label in England was in the English Court of Appeal in Mölnlycke, in the passage quoted above. The Court of Appeal did not explain why the opinion of experts should be considered primary evidence, and indeed, it seems have to misunderstand the role of the expert witness. In the first place, an expert’s opinion as to what a skilled person would have thought at the time is clearly not direct evidence as to what a person of ordinary skill actually thought at the time, but rather opinion as to what such a person would have thought, and this, as the courts have repeatedly pointed out, is tainted by hindsight: a well-known statement is that in British Westinghouse v Braulik (1910) 27 RPC 209 at 230 disparaging “ex post facto analysis of invention,” but a similar observation appears in almost any leading obviousness decision. Consequently, while expert opinion on the ultimate issue of obviousness is admissible, “it must be treated with extreme care” (Beloit Canada Ltd. v. Valmet Oy, (1986) 8 CPR(3d) 289 at 295). The role of the expert is to educate the court as to the state of mind of the person skilled in the art at the relevant date. It is for the court to assess obviousness based on that information. The expert evidence as to the state of mind of the posita, while essential, is not even nominally direct evidence as to obviousness. Clearly the court’s reconstructive conclusion is just that, a conclusion based on the evidence; it is not in itself direct evidence as to what a posita would have thought. Despite its education in the course of the trial, the court is even less a person skilled in the art than the expert witness, and the court’s opinion is also subject to hindsight bias. Thus there is no particular reason to believe that the judge’s opinion as to obviousness based on his education by experts will be more reliable that inferences based on circumstantial factors.

It is true that the ultimate statutory question is whether a person of ordinary skill in the art would have thought the invention obvious at the time, but this does not mean that either the opinion of expert witnesses, or the attempts by the court to assume the mantle of such a person, is the best evidence on the issue. The Hon Giles S Rich, a long-serving member of the US appellate patent courts, and one of the drafters of the US obviousness provision, has stated that “the best evidence and the most reliable may be the circumstantial evidence,” explaining colorfully that “If a man is observed coming away from the scene of a murder with a bloody knife or a smoking pistol, the evidence thereof may be more convincing than what he says” ((1972) 1 AIPLA J 26 at 38, emphasis in original). Accordingly, when circumstantial considerations point strongly in the direction of inventiveness, an argument in favour of obviousness based on expert opinion evidence will rarely prevail. So, as a well-known passage put it:

The truth is that when once it has been found, as I find here, that the problem had waited solution for many years, and that the device is in fact novel and superior to what had gone before, and has been widely used, and used in preference to alternative devices, it is, I think, practically impossible to say that there is not present that scintilla of invention necessary to support the Patent.

Samuel Parkes & Co Ltd v Cocker Bros, Ltd. (1929) 46 RPC 241 at 248

In many of the leading cases on obviousness the reason for appellate reversal of the trial judge is that undue weight was given to evidence of expert witnesses as opposed to objective factors. For example, in Beloit Canada Ltd. v Valmet Oy, long the leading Canadian decision on obviousness, the trial judge held that the invention was obvious, basing his finding primarily on the evidence of expert witnesses to that effect. The Court of Appeal reversed, noting at 296 that the trial judge “appears to have given no weight to the uncontested objective facts and preferred instead the subjective ex post facto opinion of a hired expert.” The Court noted at 296 that “[w]hile none of these [circumstantial] factors taken in isolation is necessarily determinative on the issue of obviousness, each of them argues more eloquently than any ex post facto analysis can ever do that the patent is inventive; cumulatively their effect is simply irresistible.”

With that said, circumstantial evidence cannot be accepted any more uncritically than expert evidence, as it is subject to its own frailties, primarily the need to show a causal connection between the circumstantial evidence and inventive ingenuity. When these hurdles can be overcome, circumstantial evidence speaks powerfully, and will generally prevail over an argument based on reconstruction of the inventive process, which will always be plagued by hindsight. It is perhaps fair to say that the reconstructive approach can always be attempted, but at the same time it will always be unreliable because hindsight bias is so powerful and inescapable. On the other hand, the utility of circumstantial evidence varies widely. Because it is not subject to the problem of hindsight, it may be much more powerful than reconstructive evidence; but when the necessary relevance cannot be established it may be entirely unhelpful. The reconstructive approach is “primary” in the very limited sense that when helpful circumstantial evidence cannot be established, it is necessary to rely on the reconstructive approach, faute de mieux. Of course, in many cases, perhaps most, circumstantial evidence will be helpful but not so strong as to be determinative, in which case reconstructive and circumstantial evidence will both be considered, with due allowance for the strengths and weaknesses of each.

Finally, I note that in Janssen-Ortho v Novopharm Ltd / levofloxacin [2006] FC 1234 Hughes J suggested at [113] that factors such as commercial success and meritorious awards are of “secondary importance.” However, he distinguished between factors which can be established as of the date of the invention and factors arising after the time that the alleged invention is made, referring to the latter as “secondary.” This does not correspond to the usual distinction, between evidence of expert witnesses and all other evidence, regardless of the date at which it arises. Hughes J explained that secondary evidence, as he used to the term, is of “secondary importance” on the basis that “the Court is to be concerned with ‘inventive ingenuity’ exercised at the time of making the invention.” However, it has long been thought that evidence arising after the date of the claim, such as commercial success and meritorious awards, can shed light on inventive ingenuity at the time, just as behaviour of an accused criminal after the fact may be highly probative of guilt or innocence, and Hughes J did not explain why this is not so. The Court of Appeal 2007 FCA 217 at [27], in emphasizing that all the evidence must be weighed according to the circumstances, stated that Hughes J’s “ list is a useful tool, but no more.”