Puzzles of Disclosure and Support

Paid Search Engine Tools, LLC v Google Canada Corporation 2021 FC 1435 McDonald J

2,415,167 / Paid Search Engine Bid Management

In my earlier post on this decision, I noted that McDonald J held that in light of the parallelism between 28.1(1), the claim date provision, and 28.2(1), the novelty provision, priority can only be claimed if the priority application discloses the same invention; it is not enough that the claimed subject matter is reasonably inferable from priority document [221]. Specifically, 28.1(1) says priority may only be claimed to a prior application “disclosing the subject-matter defined by the claim.” But as a comment on that post noted, 38.2 prohibits amendments that “add matter that cannot reasonably be inferred from the specification. . . on its filing date”; that means, conversely, that an amendment is permitted if the added matter can be reasonably inferred from the original application. Since the specification is part of the claim, it seems that 38.2 means a new claim is permitted so long as the subject-matter can be reasonably inferred from the original specification. On its face, there seems to be a gap between “reasonably inferred” and “disclosed” with the latter being more stringent than the former: see Western Oilfield 2021 FCA 24 on the difference between the Canadian and UK standards for added matter, discussed here. How can these differing standards be reconciled?

There is a straightforward answer, which is that 38.2 goes to amendments, and 28.1 goes to priority. So it is consistent to have a claim which is permitted (under the less stringent 38.2 standard), but which is not entitled to priority (under the more stringent 28.1 standard).

But problems lie not far below the surface. Consider a scenario in which the description in the application as filed is identical to that in the priority document. A claim is amended during prosecution in response to an objection by the examiner. The amended claim is reasonably inferable from the application as filed, so is permitted by 38.2, but it is not disclosed by the application as filed. The claim date for that claim must be either the filing date or the priority date: there are no other options. In particular, it is not an option under the Act or Rules for the claim date for that claim to be the date on which it was amended. (Compare the old Rule 53 under the old Act.) Section 28.1 provides that the claim date is the filing date unless priority is claimed. If the applicant does not claim priority, then the claim date is the filing date. The claim must also be supported by the description, per Rule 60. It must be implicit that if the amended claim can reasonably be inferred from the description, it is also supported by the description. Otherwise, the “reasonably inferred” standard for permitted amendment would be pointless, as any amended claim that was only reasonably inferred from the description would be invalid for lack of support. This is supported by Hughes J’s passing statement in Bartley 2011 FC 873 [15] equating “supported” with “reasonably inferred.” Because the description is the same in the two documents, this means that the priority document also supports the amended claim.

Now suppose the applicant does claim priority. If the amended claim is not disclosed by the application as filed, it is not disclosed by the priority document, given that in our scenario the descriptions are same in the two documents; but because the descriptions are the same, it also follows that the claim is supported by the priority document. It seems very counter-intuitive that a claim should be properly supported by the application as filed, and yet not entitled to claim priority to a priority document with exactly the same disclosure.

Now change the scenario slightly, and suppose that the claim in question was present in the application as filed. The added matter provision (38.2) is now irrelevant; but by hypothesis, the claim is not disclosed in the priority document even though it is supported by the Canadian application with exactly the same description, which means that priority cannot be claimed.

Refusing a claim to priority on that basis is pointless formalism. The quid pro quo for which the patent exclusivity is granted is the new and non-obvious information disclosed in the description: see Pope Appliance (1929) 46 RPC 23 (JCPC) 55; Halocarbon [1979] 2 SCR 929, 944-45; Pfizer v Ratiopharm 2006 FCA 214 ¶ 23; American Cyanamid Co (Dann's) Patent [1971] 88 RPC 425 (HL) 452; Consolboard [1981] 1 SCR 504, 517; Amazon.com 2011 FCA 328 ¶ 36. The new information disclosed in the specification—the “mental result” or “new idea” or “novel mental concept”—cannot be patented as such: “A disembodied idea is not per se patentable. But it will be patentable if it has a method of practical application”: Shell Oil [1982] 2 SCR 536, 554. The claims define practical embodiments of that information: as the SCC said long ago: “A patentable invention is a mental result. . . . The machine, process or product is but its material reflex and embodiment”: Ball v Crompton Corset Co (1887) 13 SCR 469, 477 quoting with approval Smith v Nicholls, 88 US (21 Wall) 112 (1874) 118. The requirement that the claims are supported by the description ensures that the claimed subject matter reflects the new information: an inventor cannot disclose a new way of peeling potatoes, and claim a cure for cancer. But if the original disclosure was sufficient to support the claim, the patentee has met its part of the bargain, and that is just as true whether the disclosure was made in a foreign filing or in a Canadian application in exactly the same terms. In this scenario, the information content of the two documents is exactly the same; the only difference is that the exact words of the claim do not appear in the priority document. This is an absurd basis for refusing priority—and I mean that in the statutory interpretation sense, that it cannot possibly be right. A claim that is fully supported by the Canadian application should be able to claim priority to a priority document with exactly the same disclosure.

However, this absurdity flows from a fairly straightforward reading of the text of the Act. Of course, while the text is dominant in statutory interpretation, the text must be read in light of context and purpose to arrive at a harmonious result. So, I looked more deeply into the context, including the modern and historical case law, the history of the rule changes, the concept of support, the law in other jurisdictions and so on. That context provided some insights, but added some complexity, and while I have some leads, I have found no easy answer, at least at this point. Unfortunately, I am not in a position to chase this down right now. Instead, I’m just adding this update to say that the position I stated in the post above follows from the text of the Act, but it leads to enough problems down the road that I no longer have the same confidence in it that I did when I wrote the post.

Now consider another implication of the above hypothetical, in which the priority document and the Canadian application have exactly the same description, and the Canadian application has a claim that is supported by the description but is not disclosed by the description. As discussed, that means that the claim is valid but priority cannot be claimed. But at the same time, it is almost certain that a description sufficient to support the claim would also be sufficient to anticipate the claim, with the counter-intuitive result that the priority document would anticipate the claim (but for the grace period), even though priority cannot be claimed. This result is very difficult to reconcile with the view that “disclose” has the same meaning in 28.1 as in 28.2. That is not to say that disclose has different meanings in those two provisions, as the analysis also turns on the meaning of “priority”. The point is only that this is another absurd result flowing from a straightforward reading of the text.

I should emphasize that in the foregoing, I do not mean to endorse any particular test or definition for “disclose” and “support” and “reasonably infer.” So, in agreeing with McDonald J that “disclose” has the same meaning in 28.1 as in 28.2, this was not to endorse the practice of any particular patent office as to the interpretation of that term. I am given to understand that some patent offices interpret “disclose” to require a verbatim or almost verbatim reproduction of the claim in the description. I doubt that is right; I am inclined to prefer the approach taken by Jacob LJ in Unilin Beheer [2004] EWCA Civ 1021 [48].“Disclose” of course is subject to extensive caselaw in the context of anticipation, but even if it has the same meaning in the priority context, it is not obvious how it should actually be applied. For example, a claim to “PDEV inhibitors for the treatment of erectile dysfunction” would be anticipated by the prior use in traditional medicine of epimedium to treat ED; but that does not mean that the disclosure in the priority document of the use of epimedium to treat ED would disclose the use of PDEV inhibitors to treat ED. More broadly, there is very little caselaw on the meaning of any of these terms: see the very interesting discussion of the meaning of “support” in CA/PL 27/99, part of the Travaux préparatoires for the EPC 2000. I expect a proper resolution of this problem will show that some aspects of current practice, or implicit understandings of the key terms, are wrong.

Finally, I note that the result in this case does not seem to have turned on the point. PSETs experts argued that certain claims were “reasonably inferable” from the 976 patent [216]–[217]. McDonald J rejected this argument in terms which seem to me to reject the argument of reasonable inference [218]–[219]. McDonald J did go on to say that the priority document must “disclose” the invention [220]–[221] as discussed above, but, so far as I can tell, this was not necessary to the result. (I want to emphasize that I am not saying that I am of the view that reasonably inferable is the test for claiming priority—simply that the point was not determinative in this case.)