A talk for Fordham IP Conference, 9 April 2015, Cambridge UK
In preparation for this talk, I asked a venerable patent attorney if he could share his deepest thoughts on priority with me. He pondered, then said: "Priority is a very good thing".
And so it is, or at least it can be. The origins of the rights of priority can be traced back to the great international exhibitions of the 19th century, particularly in London, Paris and Vienna. At a time of national rules on novelty, and strict rules invalidating patents if they were not worked nationally and quickly, potential exhibitors - especially from the US - were reluctant to come without better IP protection. Specific laws were enacted around the London and Paris Expositions in the 1850s and 1860s, and discussions started in the 1870s in Vienna that ultimately led to the Paris Convention and the adoption of a right of priority.
The purpose of the Convention was to avoid the burden of applicants being forced to make multiple simultaneous international applications. Strangely, the Paris Convention is silent on the subject matter of the subsequent application, but this is largely resolved by Bodenhausen's authoritative Guide1: "The subsequent filing must concern the same subject as the first filing on which the right of priority is based. This means that in the case of patents … it must concern the same invention or innovation".
The first question I set out to address in this talk is whether the current strict test for priority fails to recognise the early stage in enunciating an invention at which the priority document is written.
In European Patent Convention member states, the leading authority on claims to priority is G2/98. The Enlarged Board addressed the question of whether to adopt a narrow or broad interpretation of the concept to "same invention", holding that "the skilled person can derive the subject-matter of the claim directly and unambiguously, using common general knowledge, from the previous application as a whole".
A particularly interesting case applying G2/98 is T1213/05 Breast and ovarian cancer/University of Utah. This was a gene patent. The priority document contained a sequence that was 99.73% correct. The sequence was corrected on filing the patent application. It was accepted that the errors were within the range of normal experimental measurement, that 180,000 subsequent tests showed that the differences were of no effect, and that the original sequence was a good basis for the intended use (detecting a BRCA1 mutation). Even so, the TBA refused priority. The decision placed a heavy reliance on the fact that the Enlarged Board required a narrow approach.
This may sound mere semantics, but faced with a choice between a narrow or broad approach, I would suggest that the best course is to reject both. Like Goldilocks: do not go too broad, or to narrow, but just right.
Now it is easy, if you are English, to think that you are right. And at least sometimes we can be: see HTC v Gemalto2. Relying on earlier English judgments3, the judges at first instance and on appeal emphasise that the approach should not be formulaic: "priority is a question about technical disclosure, explicit or implicit. Is there enough in the priority document to give the skilled man essentially the same information as forms the subject of the claim and enables him to work the invention in accordance with that claim." And "the important thing is not the consistory clause or the claims of the priority document but whether the disclosure as a whole is enabling and effectively gives the skilled person what is in the claim whose priority is in question." On that basis, the court upheld claims that included features that included features that were not expressly in the priority document. The court held these were things that the skilled person "would not think they were doing anything different from the principles of general application the inventors of the priority document have taught."
There's good reason to be cautious about broad readings of priority documents. It is worth recalling that the question addressed in G2/98 was precisely whether a line of cases introducing a broad approach could stand. The Enlarged Board warned that allowing priority claims to stretch illegitimately would mean potentially dispossessing the true later inventor. It could be particularly unfair on selection inventions: if priority is not kept in its place, the earlier filer can simply hijack the work of later inventors.
On the other hand, there is also good reason not to be overly strict. It's hard to see a proper justification for denying priority where the differences are accepted to make no difference as appears to have been the case in T1213/05. But I'd argue that these problems derive from a misreading of G2/98. The problems arise from misunderstanding what the "strict" test is in G2/98: it should be holding patentees strictly to the meaning of their priority document - not holding them to its strict meaning. Going down the latter path means denying priority for something within the true meaning of the priority document, a course which is impossible to justify. Priority should be a tool to benefit inventors, not a bear-trap to take away their inventive efforts.
And that brings me to my second subject, the poisonous divisional. If anything brings the law into disrepute, this is it. The priority system exists, at heart, to ensure that inventors can keep the benefit of their inventions in the face of a complex international patent system. Divisional patents exist primarily because the patent office has a rule of unity of inventions. Putting the two together, it is surprising to find that a loss of priority means a loss of coverage whereby a patentee, who has only done what the patent system should allow him to do, finds himself unable to claim the generalisation of his invention.
This is a very real problem. Both the EPO4 and in the UK domestic court5 have held patents invalid over the disclosure of the priority document.
Example 1 - Nestec
- The priority document set out one method for the arrangement of a coffee capsule extraction system.
- The patent's claims set out a more general set of arrangements for coffee capsule extraction but they could not be considered as a set of clearly defined alternatives.
- It followed that the claim was not entitled to priority
- As a result the published priority application constituted prior art and the patent was therefore invalid for anticipation
Example 2 - T1496/11
- Claim 1 of the Main Request included a security document having a security device.
- The priority document described a security document as including "a printed or embossed" feature.
- Claim 1 of the Main Request defined subject-matter generalised from the priority document, did not relate to the same invention, and so was not entitled to priority
- A published divisional featured the description that was entitled to priority. The description in the divisional therefore anticipated the parent's claims.
The consequence is that it is frequently impossible to claim a generalistion of one example in a priority document by reference to further examples developed over in the priority year. The good news is that the TBA is also troubled by this. Questions (not yet available) have been referred to the Enlarged Board in T557/13.
I very much hope to see that the Enlarged Board unravels this particular problem. In my view it is incompatible with Paris Convention obligations. The Convention provides at Article 4G that "(1) If the examination reveals that an application for a patent contains more than one invention, the applicant may divide the application into a certain number of divisional applications and preserve as the date of each the date of the initial application and the benefit of the right of priority, if any." So the Convention expressly contemplated divisionals which may or may not enjoy a right of priority and it seems incredible to think that an inventor can be refused a patent on the basis of a system developed to deal with problems of unity of invention. I suppose one answer is to be less inventive but one hopes the Enlarged Board can do better than that.
A solution that I favour and perhaps the only plausible solution is to be more expansive in the use of partial priorities. G2/98 acknowledges that claims can have partial priorities "provided that it gives rise to the claiming of a limited number of clearly defined alternative subject-matters."
The approach is as follows. Define X as anything clearly and unambiguously supported by the priority document - which on the face of it must be a clearly defined group because, not being entirely flippant, even the definition says "clearly". Then everything must be X or not X. So now claims can be qualified by adding "and characterised in that it is X or not X". And now the claim includes two clearly defined alternatives. The X alternative is entitled to priority precisely coextensive with the priority document and therefore is not capable of being defeated by a divisional. And the rest isn't. And on that basis, all claims entitled to partial priority can be regarded as consisting of a limited number of clearly defined alternative subject-matters.
Brilliant and attractive as that solution may appear, a real difficulty arises when dealing with intervening art. Consider a journal article published between the priority document and the patent. The question is whether that article renders obvious anything within the claim that is not disclosed and enabled by the priority document. So that requires a comparison of document with document: something that patent law has not had to face since the creation of claims.
But where does the greater difficulty lie? Depriving inventors of the benefits of their inventions, or posing a difficult problem for courts to solve when assessing prior art. I am afraid on this, until a better solution appears, I hope for an Enlarged Board decision favouring inventors.
Finally, just a few words on formalities. Decisions in the TBA (T788/05) and in the English courts (Edwards Lifesciences v Cook  EWHC 1304 (Pat)) denied priority on the basis of a faulty assignment of the right. A claim to priority only holds good if it is made by the same person who filed the application from which priority is claimed or the successor in title. That means, in the cases of co-owned priority applications, that all the co-owners must have completed assignment to the applicant at the time the claim is made. Title cannot be perfected later, nor can one co-owner acting alone assign the right.
But it is fair to say that this rather harsh result is one dictated by statute. It is hard to see what mischief such a strict law seeks to overcome; and judges try hard to ameliorate its effects. The real difficulties faced by the patentee in the Edwards case were circumvented in HTC v Gemalto, where Birss creatively assumed (in the absence of evidence on the point) that Texan law on employee ownership was the same as UK law of employee ownership thereby making the European patentee also the owner of the US invention. In KCI v Smith & Nephew  EWHC 1487 (Pat), Arnold J held obiter that a confidentiality agreement would have been enough to transfer beneficial ownership of the right of priority. The Legal Board of Appeal in J19/87 Burr-Brown/Assignment held that an equitable transfer of title would be good enough to base a claim to priority.
Those are partial, and fact-dependent solutions. Whether the problem is on a scale that calls for treaty change or legislative intervention is a matter for debate. It is worth bearing in mind that the Paris Convention lays down circumstances in which priority must be acknowledged, but does not rule acknowledging priority in wider circumstances.
Drawing those threads together, the picture that I think emerges is that the priority system is a technical one and prone to technical failure. Overly strict lines can be drawn around the teaching of priority documents; the existence of poisonous divisionals is an embarrassing failure of the patent system; and formalities can all too easily defeat patents.
But all of these are points that can be, and are, addressed by tribunals seeking to get to a fair answer. G2/98 does not mandate strict interpretations, but a more holistic approach. Solutions exist to poisonous divisionals (though perhaps at some price of complexity), and we can hope the Enlarged Board adopts them. Creative judges try to help parties sort out their imperfect chains of title.
And so to answer the title question: is the priority system failing us? We have the tools we need to make sure it doesn't.