From March to September 2016 the team is joined by Guest Kats Emma Perot and Mike Mireles.

From April to September 2016 the team is also joined by InternKats Eleanor Wilson and Nick Smallwood.

Thursday, 13 August 2015

Partial Priority - questions for the Enlarged Board now revealed

Thanks to the comments posted by keen-eyed readers here and here, it has come to this Kat’s attention that EPO Board of Appeal decision T 0557/13 has finally been published (despite the patent in question having already expired...). Tufty the Cat covered this appeal in an article in January, as he pawed through the minutes of the Board of Appeal hearing, and has since already published an in-depth analysis of the recently published decision here. This Kat will therefore give but a brief overview of the situation.

But first, some background.

Even Merpel is subject to partial priority
This decision relates to patent number EP 0 921 183 (which is the result of a divisional application) in the name of Infineum. The patent is directed towards the use of cold flow improvers in fuel oil compositions - essentially, introducing additives to diesel such that the freezing point of the diesel is lowered, thereby avoiding undesired characteristics in cold temperatures.

The appeal itself derives from a decision by the opposition division to revoke the patent, based on the fact that the claims were not fully entitled to their priority - the granted independent claim was a generalisation of a more specific disclosure in the priority document. As a result, the published parent application constituted prior art under Article 54(3), and was novelty-destroying.

Enlarged Board decision G 2/98 was cited by the opposition division. The independent claim of the patent features a generic chemical formula which, due to the variable nature of the described substituents, encompasses a number of alternatives but does not specifically define each alternative. This can be viewed as a 'generic "OR"-claim', in the language of G 2/98. With respect to this type of claim, Reason 6.7 of G 2/98 states:
The use of a generic term or formula in a claim for which multiple priorities are claimed [...] is perfectly acceptable under Articles 87(1) and 88(3) EPC, provided that it gives rise to the claiming of a limited number of clearly defined alternative subject-matters.
The opposition division did not believe that the claim gave rise to the claiming of a limited number of clearly defined alternative subject-matters, and so revoked the claim.

Appeal

In the present decision, the Board of Appeal recognised that first and foremost, the issue to be resolved was whether partial priority could be claimed:
The concrete question to be resolved here is whether Claim 1 enjoys partial priority to the extent that the use of [a specific compound] as disclosed in [the priority document] is encompassed by the more generic definition of Claim 1, rather than being spelt out in it.
It therefore follows that there needs to be clarification as to what the "limited number of clearly defined alternative subject-matters" of G 2/98 means.

The questions referred are as follows:
1. Where a claim of a European patent application or patent encompasses alternative subject-matters by virtue of one or more generic expressions or otherwise (generic "OR"-claim), may entitlement to partial priority be refused under the EPC for that claim in respect of alternative subject-matter disclosed (in an enabling manner) for the first time, directly, or at least implicitly, and unambiguously, in the priority document?
2. If the answer is yes, subject to certain conditions, is the proviso "provided that it gives rise to the claiming of a limited number of clearly defined alternative subject-matters" in point 6.7 of G 2/98 to be taken as the legal test for assessing entitlement to partial priority for a generic "OR"claim?
3. If the answer to question 2 is yes, how are the criteria "limited number" and "clearly defined alternative subject-matters" to be interpreted and applied?
4. If the answer to question 2 is no, how is entitlement to partial priority to be assessed for a generic "OR"-claim?
5. If an affirmative answer is given to question 1, may subject-matter disclosed in a parent or divisional application of a European patent application be cited as state of the art under Article 54(3) EPC against subject-matter disclosed in the priority document and encompassed as an alternative in a generic "OR"-claim of the said European patent application or of the patent granted thereon?
This Kat is pleased to see that the Board purposely formulated the questions in terms broader than those suggested by the parties, such that they reflect the issues of partial priority generally, which can arise whenever there is state of the art potentially relevant under either Article 54(3) or 54(2) EPC, and not only in situations of "poisonous priority".

This Kat would like to thank his colleague Rob Barker for authorial assistance with this post.

154 comments:

Anonymous said...

May be we should remember why the right of priority has been created in 1883: it was simply to avoid self-collision.
Whatever the Enlarged Board might decide following the present questions, it should remember this important function of the priority right. Using the prior application as prior art under Art 54(3)seems to go contrary to the concept of avoiding self-collision.
G 2/98 was far from perfect, but has to be seen in the context at the time, i.e. with respect to the Snackfood T 73/88 and the similar following decisions. As it made clear that the Snackfood decision should not be followed, it brought some clarity for applicants and third parties.
Can it really be correct to claim that the FICPI memorandum can be said to express the legislative intent underlying Art 88(2), second sentence? Why did the EBA at the time have to go into this?

Anonymous said...

Kobus says:
@anonymous of 21:43
I do not agree that the (sole) purpose of a priority right is to avoid self-collision. The Paris convention dates from the 19th century when it would be impossible to file patent applications in different countries on the same date. It allowed an applicant to protect an invention in many countries provided he filed the first (or priority) application before anyone else. It would then not be necessary to race to beat competitors to file at patent offices in the other countries because the date of the priority application would already be established. The applicant could then comfortably travel around the various countries by train or ship at his leisure in order to file further applications. The question of whether a claim to priority is valid relates more to whether the applicant is truly filing a further application for the same invention or has altered the subject matter since the priority date. Such an application should only be accorded a filing date when the amended subject matter was invented and attempting to claim the date of the priority would be cheating.

Anonymous said...

Self-collision? Is that the case?

"Any person who has duly filed an application for a patent, or for the registration of a utility model, or of an industrial design, or of a trademark, in one of the countries of the Union, or his successor in title, shall enjoy, for the purpose of filing in the other countries, a right of priority during the periods hereinafter fixed."

Where is the self-collision in other countries?

MaxDrei said...

Useful thread. Bear in mind that outside Europe your own earlier (and as yet unpublished) filings don't hurt the patentability of your later filings. But the earlier unpublished filings of others do. Under the AIA even for obviousness.

My own view is that EPC Art 54(3) gets it exactly right: novelty but not obviousness; your earlier filings just as prejudicial as the filings of others.

That's because obviousness alone is enough to set patentability relative to the publicly-known state of the art. Novelty is needed simply to i) stop double patenting and ii) mediate between rival filers none of whom have yet published. Art 54(3) does this job, as optimally as it can be done.

Anonymous said...

It is certainly not self-collision as many (I would venture to say most) countries in the world do not recognise internal priority. You can only claim priority from applications that have been filed abroad. That the EPC does recognise EP files as priority rights is an exception to the rule due to the nature of the European patent. Priority rights were established to allow filing with ease in other countries without having to race to beat the competitor, as was mentioned by another commenter yesterday.

MaxDrei said...

And why is it that, outside Europe, your own earlier unpublished filings don't get applied against your novelty? Anybody know? OK, I'll tell you. It's in the Tegernsee Group Report on Art 54(3). Japan reported that it is just "too difficult" for a corporation filing streams of patent applications to keep track of whether they have novelty over all their own earlier filings.

And how "difficult" is it for the public to keep track, I wonder, of which earlier unpublished filings shall "count" under Art 54(3) and which do not. Does that count for anything?

Meanwhile, the USA always gave a "Free Pass" to earlier filings of the same inventor. I guess Japan felt constrained to follow the Americans, when writing that part of its patent law.

The snag is, of course, deciding what is "my" earlier filing and what is in fact the earlier filing of "another". The simplicity of the EPC's Art 54(3) cuts through all that nonsense. Let's see now if the EBA can continue to "Keep It Simple"!

MaxDrei said...

A further thought occurs.

Prior to the EPC, the UK had ist Patents Act of 1949. There were some nice provisions in there, that might have helped Inventors if they had been carried through ito the EPC.

One such was Section 52, which provided that disclosure during the Paris year of the content of the "earlier" priority-providing application for the "same invention" would not be available to attack the patentability of any claim that has the bad luck not to be entitled to the priority date.

So the UK, even back then, was on the side of the world today outside EPC-land, in softening the horror of "self-collision". The Point of a properly functioning patents system is to get early filing and early A-publication, to "Promote the Progress of Useful Arts". If Inventors are going to be deterred from early filing, perhaps even cautioned by their patent attorneys to hold back till the invention is perfect, that is not promoting the progress.

So, if the EBA has difficulty deciding, whether to hold to a strict line, or to be a bit more tolerant towards filers who have collided with themselves, perhaps a reflection what patent laws outside EPC-land think is right might help them to get to a balanced clean and clear line that is fair to i)inventors who find they have to divide, ii) inventors who file on every scrap and so don't need to divide and iii) the public, and which the world can applaud.

THE US anon said...

MaxDrei at 7:48 - what meaning are you using with "unpublished" and the AIA?

Why do you seemingly want to use only novelty and not obviousness as a screen between filers? Does only the one and not the other control what is awarded a patent? Do you leave out as well any sense of enablement of other particulars of patent law?

As to your rather presumptuous KISS at 8:39, I am reminded of one of my favorite quotes from Einstein (you are trying to be too simple, my friend).

Lastly, the argument you make at 10:44 harkens to the US system pre-AIA. As I recall, you often chided us Americans to join ROW and eliminate that which you now appear to pine for.

I am left with the distinct impression that your view on "best system" is a bit muddled.

Proof of the pudding said...

Is not the most pressing point to deal with here the absurdity that the mere act of dividing an application and allowing the divisional application to publish can render unpatentable subject matter that would have been totally unobjectionable if the applicant had not taken the decision to divide?

Remember, a poisonous priority objection has no teeth if the divisional application is not published. However, what kind of public policy is served by punishing applicants for publishing divisional applications (whose subject matter cannot, in any event, extend beyond what has already been published in respect of the parent application)? It is THAT absurdity that has no counterpart anywhere else in the world.

Debate all you wish about which system is the most sensible when it comes to dealing with Article 54(3)-type prior art. However, any system that includes such a ridiculous, unfair and potentially inescapable "bear trap" for applicants to fall into is, at least in my book, a very broken system indeed.

Remember, applicants may have no choice but to file divisional applications (e.g. due to an unexpected non-unity objection based upon newly-discovered prior art). The purpose of a divisional application in that situation is supposed to be to help the applicant secure protection for all of the inventions that they have disclosed - and not to poison the well so that (suddenly) subject matter of both the parent and the divisional becomes unpatentable.

MaxDrei said...

All fine and good Pudding, but your wickedness only clicks in when there is a disconformity between your priority-date-defining earlier application and your one-year-later-PCT filing. When the two have identical disclosure content, the one cannot poison the other, can it?

Given that Paris promises a right of priority only for the "same invention" what makes you think you can file one disclosure on the priority date and a different one as your PCT without suffering any adverse consequences. You are abusing the system, aren't you?

I mentioned the 1949 UK Patents Act above. That was a different system, written in the full expectation that inventors would have more disclosure in their "complete specification" at the end of the priority year than in their "provisional specification" that starts their priority year. But all that stuff got swept away in 1973 when Europe settled on the EPC regime.

If, during your Paris year, you find that you have more to tell the Patent Office, file another application, with claims that are novel over your own earlier application.

I do understand that the EPC puts a premium on competent drafting. Every other jurisdiction is more lenient and forgiving of poor drafting. When EPO Examiners routinely urge it on Applicants that they are the authors of their own misfortune, they don't understand i) how tolerant the world outside Europe is, of poor drafting and ii) how hard it is to draft well.

But we want respect for good drafting, don't we?

Tim Jackson said...

MaxDrei: "Given that Paris promises a right of priority only for the "same invention" what makes you think you can file one disclosure on the priority date and a different one as your PCT without suffering any adverse consequences. You are abusing the system, aren't you?"

Max, you seem to have overlooked Paris Article 4F, which allows exactly that. Partial priority. You should get the benefit of priority to the extent that your claim overlaps subject matter in the priority application, not just if it's identical.

That ought to protect you from a novelty-only attack under Article 54(3). Unfortunately G 2/98 has spawned a line of case law which says otherwise, based on its requirement for "a limited number of clearly defined alternative subject-matters". Question 2 of the present referral asks if that's right.

You previously mentioned Section 52, Patents Act 1949. That went further. Not only were you protected against a novelty-only attack if your claim was wider than your priority document. Section 52 also said that an intervening publication of subject matter in your priority application couldn't be used in an inventive step attack.

So for example, if your priority application said "copper" and you published a paper the next day saying "copper", but your eventual claim said "metal", then your publication couldn't be used to argue that using another metal such as iron would be obvious.

maxdrei said...

Yes Tim, I had overllooked Art 4F. Thanks.

Anonymous said...

I am reminded of an old discussion:

http://tuftythecat.blogspot.co.uk/2011/02/article-543-and-ep-divisionals.html

Meldrew said...

Proof of the pudding has nailed it with use of the word "absurdity"

The EPO is already seen as expensive, slow, and formalistic.

Poisonous divisionals adds Kafkaesque to the list of unfortunate adjectives that could reasonably be applied.

MaxDrei said...

Meldrew writes excellent stuff but this time I cannot go along with his attitude that, when the law gets messy, it's all the fault of the EPO. Perhaps he is one of the crowd that would have it that DG3 has messed up not only the law of novelty but also the law of eligibility (Art 52) and obviousness (Art 56)? On that, count me out. I'm not in that crowd.

Fact is, Meldrew, the founding fathers of the EPC thought it right to include your own earlier filings in the Art 54(3) definition of novelty-prejudicial art. That's the root of the problem, I say, and the EPO didn't write the European Patent Convention.

Where the EPO does have to step up to the plate is to say whether those who bundle their disclosures and later file divisionals are to be advantaged relative to those who file a stream of new patent applications for each of their discrete disclosures. Agreed?

And as to Paris 4F, I think that contemplates successive filings with successive contributions to the art. I'm not sure it was written to legitimate successively wider claims (statements of invention) supported by the same unchanged disclosure in the description and drawings. We in the UK like to think we can draft claims better than anybody else. We cannot fairly exhort others to "Get it Right First Time" if we don't hold ourselves to the same standard.

Meldrew said...

Did I say it's it's all the fault of the EPO?

I merely said that the adjective Kafkaesque is one that is likely to be added to other adjectives reasonably applied.

Of course judges can create problems in the law by undue over-analysis of words rather than intent, and the judge-wrought wording “provided that it gives rise to the claiming of a limited number of clearly defined alternative subject-matters.” of G2/98 seems to account for most of the ills we are seeing. So yes, there is someone to blame: whether it is those who used these unfortunate words, or those who have over-interpreted them, I leave it to history to decide.

By the way, I think the EPO is great, which is why I care about its reputation.

Anonymous said...

I also hate the hypocrisy of attorneys who take advantage of the law of novelty the EPO has developed, then bemoan their approach to added matter. Rather than whinge, I suggest these attorneys take this long-standing law into account in their drafting. Yes, yes, it isn't possible to cover every eventuality, but life is a series of swings and roundabouts and you win some, you lose some, and if you excuse the use of an idiom, go take a chill pill.

Tim Jackson said...

MaxDrei: "... the founding fathers of the EPC thought it right to include your own earlier filings in the Art 54(3) definition of novelty-prejudicial art. That's the root of the problem, I say..."

Is that really the root of the problem? In this respect, no problem appeared for the first 20 or 30 years of the life of the EPC. Then came a line of case law which interpreted the phrase "a limited number of clearly defined alternative subject-matters". That phrase doesn't come from the EPC, but its gradual development is what has eventually led to poisonous priorities.

Notice, however, that Question 5 of the referral asks whether the founding fathers really did intend to include the subject matter of your own earlier filings in the Art 54(3) definition of novelty-prejudicial art.

MaxDrei said...

What a pleasure to read these eloquent contributions from Meldrew and Tim Jackson. They prompt a comment from me on stare decisis (binding precedent) a thing I used to think was vital for legal certainty and which the EPO lacks. Now I think the "survival of the fittest" progress of DG3 law delivers more certainty on the substantive law of validity under the EPC. G2/98 has to go!

I agree that "limited number of clearly defined alternative subject-matters" is Quatsch, unhelpful, and must be over-written by the EBA, at the first available opportunity. Like, now!

Just as DG3 got there in the end, on Art 52 and Art 56, so too, with this case, I expect it to find a line that everybody can sign up to, on Art 54(3).

And, yes, anonymous, attorneys do indeed work Art 54 jurisprudence to their advantage while disparaging the EPO when, for Art 123(2), it adopts exactly the same stance on what is the disclosure content of any particular document. Some attorneys seem to be adopting the Nelsonian approach, of putting the telescope to their blind eye. They would call it "zealous advocacy", I suppose. I deplore it as much as you.

The Cat that Walks by Himself said...

If it's 54(3) EPC and "OR", how about disclaiming the parent application alternative/embodiment?

The Cat that Walks by Himself said...

The added subject-matter test is more strict than the novelty test. Namely, a general is not new over a specific, however the general adds subject-matter to the specific. Therefore, the general is not new and without priority granted to the specific.

In case the specific can be delimited within within the general, there is no obstacle to claiming partial priority for the general.

Anonymous said...

Regarding question 1, if a claim merely encompasses (i.e. is broader than) subject-matter disclosed in the priority document, it follows immediately from the order of G 2/98 itself that in principle the claim cannot enjoy a right of priority.

Where the subject-matter of a claim is the union of a finite number of subject-matters disclosed in the application, there is of course no harm in recognising a separate priority right for each of those subject-matters. This is because the applicant could anyway have claimed them separately, and actually forcing the applicant to do so would be silly and unhelpful. That was the point of the famous memorandum and G 2/98 acknowledged that point. (Really, the memorandum is about a simple formal point: do we need to file each alternative as a separate claim, or will the EPO allow us to group them in one claim. Dissecting the memorandum's examples from the early 1970s using post-G 2/10 knowledge of the disclosure test is not the right approach.)

Note that being "generous" on partial priorities will not be of much use outside the area of poisonous priorities/divisionals. If the priority document discloses copper and the application claims metal, then even with the "generous" approach an intermediate document showing copper will be 54(2) prior art for the "metal minus copper" part. So it won't be novelty destroying, but it will normally take away inventive step. (And in the rare case where an inventive step is involved in replacing copper with any other metal, it would be rather strange that a patent is effectively awarded for inventing "metal minus copper" without the application actually disclosing that subject-matter. Of course IF the application did disclose "metal minus copper" together with "copper", then partial priorities should be recognised, see the previous paragraph.)

At the same time, the proposed "generous" approach will invalidate many priority claims due to the priority document not being the "first" application anymore (compared to the accepted approach). In fact, T 1222/11 is an example of just that happening.

Unintended consequences of Art. 54(3) should be fixed by carefully interpreting Art. 54(3). Shouldn't be too hard and won't mess up other areas of patent law.


@The Cat that Walks by Himself:
Very good point but see G 2/10, in particular point 4.7. The EBA decided that the "undisclosed disclaimer" approach of G 1/03 cannot be used to disclaim subject-matter that is disclosed in the application.

Anonymous said...

General encompasses the specific so CANNOT be novel by any stretch of patent law anywhere. This does not make the added-matter rules stricter.

'General excluding the specific' is novel. Draft this and you have a novel claim. Have a fallback to this and you won't add matter by amendment.

MaxDrei said...

So, have I understood the recent anonymous contributions aright, in the situation where you realise during drafting the PCT that you were too narrow in your priority document, you can avoid poisoning yourself by careful claim drafting during the preparation of the PCT. Thus: If your priority application described and claimed "It, using copper", but for your PCT a year later you want to claim "It, using metal", you should include in the PCT a precautionary statement (or a claim) that discloses your invention to be "It, using metal other than copper".

Anonymous said...

Max, I was only considering the strictness of added matter versus novelty comment.

However, what you say is certainly good practice, not only to deal with known potentially poisonous priority issues, but all 54(3) overlaps that are known at the time of drafting the PCT. If an attorney is aware of a 54(3) novelty issue where a species has been disclosed in an earlier application, and the attorney fails to either draft a claim as you have described (general disclaiming species) and/or fails to draft alternative embodiments that are novel, I believe that attorney is negligent. And, it happens. An attorney should be aware if 54(3) art if they are intimately involved in a project.

Up until recently, a failure to take into account poisonous priority issues when drafting a PCT, would not be negligent, in my view. The issue is now so well known (to those who bother to perform the most basic CPD) that it must now be taken into account. However, a good draft should have sufficient fallbacks to overcome unexpected issues, such that commercially valuable patent claims could be obtained.

MaxDrei said...

Thanks Anonymous. On drafting the PCT, one knows (or ought to know) the disclosure content of all your own earlier filings (including the earlier filed applications you will declare in your PCT), so one can take them into account when drafting the claims. It is the same issue, I suppose, whether the A-publications which will later self-collide are parent and divisional or just an ordinary 54(3) cite. You can see all the cards in your own hand.

But one can't know yet what 54(3) novelty issues are going to arise out of the near contemporaneous filings of others, when they go down, face up, on the A publication table. Those who see how to draft around those 54(3) disclosures are the really gifted ones. But DG3 does see the issue and does allow for it.

Anonymous said...

I saw a petition recently for some unjust-treatment complaint where a divisional application had been filed with a claim disclaiming subject matter of the parent case. No literal basis for the disclaimer wording and no basis to remove the added disclaimer because the divisional was an amended version of the parent case. Arguable whether the disclaimer should have been allowed, but not a situation that was wholly unavoidable, to say the least. I couldn't work out the reason for the disclaimer, because there was no poisonous priority issue that I could make out, this being the reason for my initial interest.

Proof of the pudding said...

Max - on 15 August you wrote the following.

"Where the EPO does have to step up to the plate is to say whether those who bundle their disclosures and later file divisionals are to be advantaged relative to those who file a stream of new patent applications for each of their discrete disclosures. Agreed?"

Can you please explain how you think that those who bundle disclosures together might (through proper allocations of partial priority) be advantaged relative to those who instead file streams of separate applications?

Partial priority is only applicable in cases of overlap between the disclosures of an EP application and its priority document. If you have a separate application with a broader disclosure that overlaps with a more specific disclosure in one of your earlier filings, then why on earth would you not arrange to claim priority from that earlier filing (thus placing yourself in exactly the same position as if you had bundled together and only later divided)?

Anonymous said...

It looks like T 1890/09 should have been stayed.
http://www.epo.org/law-practice/case-law-appeals/pdf/t091890du1.pdf

Maybe the representatives were not aware of this pending referral but even so for the board to deny a priority right and then find the priority publication novelty destroying does stretch the twisted logic of the EPO somewhat.

EdT

MaxDrei said...

Pudding, good day to you. You ask me to expand on my thought about Art 54(3) EPC and divisionals.

The specific "poison" issue is one I see as being slightly different from the general "self-collision" issue.

We all know that for Art 54(3) it matters not whether the reference is your own or somebody else's. So it has been a standard part of the "scenery" under the EPC that your own earlier filings are used as novelty attacks. Nobody was dubbing this ordinary self-collision "poison".

The notion of poison emerged only recently, with "poisonous" divisionals and poisonous priorities. Only recently has everybody started citing parent A publications against divisional claims, and vice versa. There is something about it that strikes one as "not right". Hence my thought, that the EBA might somehow "grace" parent/div A publications under Art 54(3) while leaving the general self-collision regime untouched.

In logic, there seems no good basis to distinguish poisonous divisional self-collision from ordinary self-collision, but who knows what thoughts might find favour with the EBA, when grappling with, say, Referred Question 5?

I'm not spending a lot of time thinking about all this, just running thoughts up the flagpole for others to think about, and comment on.

Proof of the pudding said...

Max - thanks for clarifying the origin of your original comment.

I am not sure that I agree with you that divisionals need to be treated as a special case. As far as I can see, all that is required to minimise the impact of a divisional upon a parent (and vice versa) is a common priority and a sensible approach to partial priorities.

By way of illustration, please consider the following scenario:
- both parent (A1) and divisional (A2) applications disclose a species (X) and a genus (G) that encompasses X; and
- X is disclosed in an earlier application (P1), which is the earliest priority document for both A1 and A2; but
- G is not entitled to priority from P1.

A1 and A2 are, in theory, 54(3) prior art against one another - meaning that a claim to G (in one A1 and A2) could be alleged to lack novelty over the disclosure of X (in the other of A1 and A2).

However, what if it is assumed that the claim to G is entitled to priority from P1 to the extent that it covers X? If one takes this approach, there is no collision between A1 and A2 and the claim to G is indisputably novel.

What you will note from the above analysis is that there is no need for A1 and A2 to be related as parent and divisional. Instead, all that is required is for both to share a common priority (P1).

No doubt there are more challenging scenarios than I have described above - though none that I can think of that would merit "special" treatment being given to divisionals.

The Cat that Walks by Himself said...

Re: @The Cat that Walks by Himself:
Very good point but see G 2/10, in particular point 4.7. The EBA decided that the "undisclosed disclaimer" approach of G 1/03 cannot be used to disclaim subject-matter that is disclosed in the application.

Indeed, the situation after filing of a divisional is more complicated. After filing, I would try to exclude a specific embodiment in order to restore novelty of a remaining general by arguing "avoiding double patenting",since the specific embodimen has been patented within the parent application.

The Cat that Walks by Himself said...

RE: What you will note from the above analysis is that there is no need for A1 and A2 to be related as parent and divisional. Instead, all that is required is for both to share a common priority (P1).

Well noticed.

Anonymous said...

If the EPO 'grace' parent/div under 54(3), what happens when someone decides to include numerous inventions in their PCT so as to split them out at a later date, thereby avoiding their own 54(3) art?

Partial priority to subject matter disclosed is the answer. Distinct alternative embodiments, or whatever was previously decided, is not. If the priority doc discloses 'copper' and the PCT claims 'metal', and 'preferably copper', then priority exists for 'copper' itself and the part of 'metal' that is 'copper'. A divisional to 'copper' alone should not destroy the novelty of 'metal', because the part of 'metal' that is anticipated by 'copper' is entitled to priority.

Note: It is much too late at night to be giving full consideration to one's random thoughts, so this will have to do. Warts and All.

Anonymous said...

Partial priority to subject matter disclosed is the answer.

An easier remedy: Art. 54(3) does not apply to subject-matter that is disclosed with the same effective date in the application itself. So if both the priority document and the PCT application disclose 'copper', then that disclosure of 'copper' in the PCT application or in its divisionals does not count as prior art under Art. 54(3).

If the priority doc discloses 'copper' and the PCT claims 'metal', and 'preferably copper', then priority exists for 'copper' itself and the part of 'metal' that is 'copper'. A divisional to 'copper' alone should not destroy the novelty of 'metal', because the part of 'metal' that is anticipated by 'copper' is entitled to priority.

The problem of this approach is that 'copper' does not cover all of 'metal' and that 'metal except copper' does not have 'its own date', unless it is disclosed in the application itself.

In this regard, see point 11.5 of T 557/13. The memorandum seems to say that if a first priority document P1 discloses copper and a second priority document P2 discloses metal, then 'copper' has priority from P1 and 'metal except copper' has priority from P2. It is then argued that the same should apply if metal is not in P2 but in the application, "the European patent application itself taking the place of the second priority document". That last sentence makes sense, but the step preceding it has a problem. According to the order of G 2/98 (in combination with G 2/10, point 4.5.3), 'metal but not copper' simply does not have priority from P2 if P2 only discloses 'metal' (or 'metal, preferably copper').

In my view, the rule that T 1222/11 extracted from the memorandum examples (that refer to first and second priority documents) is in conflict with the order of G 2/98. T 1222/11 applying that rule to the partial priority case (the EP taking the place of the second priority document) made the conflict less visible, but did not remove it.

See also point 17.3 of T 557/13: "whether all conceivable alternatives encompassed by the generic "OR"-claim which are not disclosed in the priority document can be considered in toto as one alternative, as was the approach in T 1222/11". This really is the crucial question. But G 2/98 (order) in combination with G 2/10 has answered it in the negative (at least in the multiple priority case: an alternative has priority from P2 only if P2 discloses that alternative.)

MaxDrei said...

There is a newer thread running now, on the PatLit blog:

http://patlit.blogspot.de/2015/08/all-change-for-priority-at-epo-referral.html

Anonymous said...

"The problem of this approach is... that 'metal except copper' does not have 'its own date', unless it is disclosed in the application itself."

That sounds like a re-statement of the current application by the EPO that is the problem. 'Metal except copper' is not specifically disclosed in the PCT application so there is no basis for including such wording in a claim. However, to the extent that the claim covers 'metal except copper', it cannot lack novelty if we give partial priority to the part of metal that is copper. There is no need for disclosure of specific wording/alternative embodiments. No rules require amendment - just a G decision.

Anonymous said...

"I'm not spending a lot of time thinking about all this, just running thoughts up the flagpole for others to think about, and comment on."

That's decidedly UNhelpful.

We do not need more unthinking thoughts cluttering up the flagpole. We need better thoughts upon which to build upon.

Tim Jackson said...

"The problem of this approach is... that 'metal except copper' does not have 'its own date', unless it is disclosed in the application itself."

Why does it need to have its own date? The claim doesn't say 'metal except copper'. It just says 'metal'. That has priority from P2.

Of course, that then overlaps with the partial priority from P1 for 'copper', rather than being a neat alternative subject-matter according to G 2/98. But the present referral is asking the Enlarged Board whether G 2/98 really had the right test anyway.

True, the FICPI memo suggests that if partial priorities cannot be claimed for one and the same claim, then the applicant would have to draw up a separate claim for (in this example) 'metal except copper'. But presumably the EP application would then also have to explicitly disclose 'metal except copper', and would not get priority for that unless P2 also disclosed it explicitly.

But that's hypothetical. The claim just says 'metal'. That has priority from P2. It also has a partial priority from P1, to the extent that the metal is copper.

Anonymous said...

"The problem of this approach is... that 'metal except copper' does not have 'its own date', unless it is disclosed in the application itself."

That sounds like a re-statement of the current application by the EPO that is the problem.


That current application is based on the order of G 2/98 and the systematic conception of disclosure and priority that lies behind it. Exceptions to that approach will lead to inconsistencies and surprising results like the outcome of T 1222/11 (a patent was revoked that would have survived under the normal EPO approach).

Poisonous divisionals are a problem of Art. 54(3). Why not solve it there?

However, to the extent that the claim covers 'metal except copper', it cannot lack novelty if we give partial priority to the part of metal that is copper.

You can give partial priority to copper, but then you still need to identify some alternative to copper.

As I tried to explain, T 1222/11 in fact argues, based on the memorandum and a sentence from G 2/98, that the complement of copper in metal should be given partial priority from P2 (in case of multiple priorities, P2 disclosing metal). I hope we can agree that this is in conflict with the order of G 2/98. So the basic premise of T 1222/11 is flawed.

Of course we could accept that the reasoning of T 1222/11 is flawed and still argue that its approach should be adopted, but what use would that have? All that just to find an antidote to poisonous divisionals and priorities? We can find that antidote elsewhere (Art. 54(3)) without making a mess of priorities.

Tim Jackson said...

"I hope we can agree that this is in conflict with the order of G 2/98. So the basic premise of T 1222/11 is flawed."

Or G2/98 is flawed - which is what the Enlarged Board is now being asked to consider.

Anonymous said...

Or G2/98 is flawed - which is what the Enlarged Board is now being asked to consider.

Well, it is one thing to say that the conclusion reached in points 6.4-6.7 of G 2/98 is flawed (even though T 1222/11 derives its approach from those points). But expecting the EBA to overrule the order of G 2/98 goes a bit far. That would create a considerable amount of legal uncertainty, in particular if a decision is taken that goes against G 2/98 but does not completely rewrite G 2/98.

In my opinion the order of G 2/98 is just fine and accepting the order and its implications means that T 1222/11's theory cannot stand.

That leaves the problem of poisonous divisionals. Why not solve it by giving a slightly restricted interpretation to Art. 54(3)? One that excludes part of an EP being cited against that same EP? Justification can be found in the reasons of G 1/03: if the priority document discloses copper, the EP discloses metal, preferably copper, it was not the intention of Art. 54(3) to prevent a patent being granted on metal.

On the other hand one could argue, as I think some have done in this thread, that there is no problem that needs to be solved: the drafter of the EP could and should have avoided the problem by always making sure that the genus minus the species finds a basis in the EP. But this would force the drafter of a PCT application to be aware of the EPC-specific problem of poisonous divisionals. According to the EBA that would be wrong. See G 1/03, point 2.4.3: "The same situation exists if an international application under the PCT is filed which has the same effect as a national filing in more than 120 Contracting States. In both situations, the applicant cannot be expected to check the substantive patent law of all possible states before filing the application and to introduce appropriate limitations in order to cope with any exclusions which he might later face in the respective states." This is about exclusions for non-technical reasons, but I do not see why it would be any better if an EP deriving from a PCT is refused for being anticipated by its divisional (or itself!) and only because the PCT-drafter was not aware he should have included a sentence creating an artificial basis for "metal but not copper".

(I'd say point 2.4.3 of G 1/03 also shows that G 2/10 was wrong in deciding that the G 1/03 approach does not apply to undisclosed disclaimers of disclosed subject-matter.)

Proof of the pudding said...

Anon 19 Aug 21:27 - there is no need "to make a mess of priorities". Indeed, as far as I can see, there is nothing "messy" about affording partial priority to a claim to the extent that it fully encompasses a more specific embodiment that is disclosed both in the priority document and in the application as filed.

What would be the implications for novelty assessments if we took this approach to priority? I guess that we would be saying that a more generic disclosure, to the extent that it encompasses a specific disclosure, is not novel over (i.e. is the same invention as) that specific disclosure. But then that conclusion is not exactly a surprise.

We would also be saying that a generic claim is conceptually sub-divisible in a huge variety of different ways - rather than just being divisible between discrete alternatives listed in an "OR" claim. But then what would be the harm in adopting that convention?

In any event, an alternative approach that assigns partial priority only to "OR" claims would result in the triumph of form over substance. This is because it is often the case that precisely the same technical subject matter can be claimed using either "generic" terms or a list of alternatives.

For example, consider the situation where:
(1) a priority document (P1) describes and claims only compounds containing the structural motif -[X]n-, in which n=3;
(2) the subsequent application claiming priority from P1 discloses and claims compounds in which n can vary from as little as 1 to as many as 5.

Under what circumstances would partial priority for n=3 be afforded to the claim to the broader genus of compounds?

We can all agree that partial priority would be afforded if "n" were defined as "1 or 2 or 3 or 4 or 5". In that instance, "3" is an explicitly listed alternative in an "OR" claim.

But what about if "n" were instead defined as "from 1 to 5". There is no use of "OR" in that definition, and so we might hesitate to reach the same conclusion. Yet that definition clearly relates to precisely the same subject matter as the alternative.

There are plenty of other examples that one could envisage, such as the numerous "generic" terms of art that, in reality, define a closed set of discrete alternatives.

An example of this would be "noble gas" - which can clearly only mean a gas selected from a list of 6 alternatives. Another example would be an (elementary) metal. Which elements count as "metallic" might be a matter of technical dispute, but there is no doubt that those skilled in the art only have access to a limited number of alternatives when it comes to selecting an element that is a "metal".

With this in mind, what is the harm in taking a pragmatic approach to partial priority and assigning it to anything that can be shown to be an embodiment of the more generic claim? It would appear to me to be a more defensible approach.

That does leave a problem for some claims that have been narrowed relative to the disclosure of the priority document. However, that would be entirely consistent with the approach to novelty (where more specific subject matter is novel over - i.e. not the same invention as - a more generic disclosure).

maxdrei said...

Pudding, write something will you,on where that solution leaves novelty of selection.

Anonymous said...

Priority application: n = 3.
PCT: n = 1 to 5, preferably 3.
Parent EP: n = 3.
EP divisional: n = 1 to 5.

Partial priority can be afforded to the EP divisional for the subject matter n = 3, as a species of the broader range.

n = 1 to 5, excluding the subject matter for which priority is valid (n = 3), does not lack novelty over the parent disclosure.


Novelty of Selection (Example as above without any priority claims/concerns):

Prior art discloses n = 1 to 5, therefore a claim a compound where n = 3 is novel.

I don't see a problem.

maxdrei said...

Indeed, perhaps there is no problem. I was only thinking that if you can find n3 in a claim to the range 1-5, then you can also find n4, thereby to deprive of novelty a later claim to n4, a purposive selection by another applicant.
Claims can disclose subject matter, can't they?

Proof of the pudding said...

Max - novelty of selection is a different beast entirely.

I am not sure that the EPO has got their rules on that quite right. Indeed, I think that the EPO might even admit this, as there appear to be signs of the "purposive" part of the selection rules being abandoned.

If you think about it, the EPO's current approach does not make sense in relation to the "pseudo-generic" claim language that I described previously. That is, why should you get novelty for n=3 when the prior art disclosure is "n = 1 to 5" but not when it is "n = 1, 2, 3, 4 or 5"? Remember, in the example that I gave, n can only be a whole integer - meaning that the two "generic" definitions are identical in terms of the technical information that they convey.

For my own part, I would rather have "special treatment" for selection inventions than for interrelationships between parents and divisionals. We already have various legal fictions employed in the assessment of novelty, so why not one more?

By the way, I am not sure that abandoning the "purposive" element of the selection test is necessarily the right way to go. If you do this then you could end up in a tricky situation where an earlier document disclosing "n = 1 to 5" is Article 54(3) prior art to 5 separate applications filed by the same party that claim (separately) n=1 through to n=5. In other words, if the Article 54(3) document was filed by the same party, then they could effectively achieve (term-extending) double patenting of precisely the same subject matter - without ever having to show that there is anything "special" about the inventions described in any of the 5 later applications.

Anonymous said...

Disclosure of a genus of compounds where n = 1 to 5 would not anticipate a future claim to n = 3, or to n = 4. The partial priority application would not affect this.

Still no problem.

Anonymous said...

@Proof of the pudding:
In any event, an alternative approach that assigns partial priority only to "OR" claims would result in the triumph of form over substance. This is because it is often the case that precisely the same technical subject matter can be claimed using either "generic" terms or a list of alternatives.

You are effectively arguing that certain generic terms disclose the same subject-matter as an OR-claim. Where that is the case, Art. 123(2) would allow one to replace the generic term with the OR-claim, so each alternative may have its own priority date.

I agree that there is no difference between "1, 2, 3, 4 or 5" and "1 to 5". I don't think the EPO distinguishes between "a method according to any one of claims 1 to 5" and "a method according to any one of claims 1, 2, 3, 4 or 5". (I also don't see why n=3 would be novel over a document disclosing "an integer from 1 to 5". Is there case law that says otherwise? For very long lists it might be different, though. And we're not talking intervals here.)

However, I don't think "metal" is commonly understood as being limited to elementary metals. In any event, in this thread "metal, preferably copper" is intended to represent the most general case of ", preferably ". One of the memorandum examples is "hollow body, preferably a bottle". The term "hollow body" certainly is not equivalent to a finite list of specific hollow bodies, so rewriting it as "bottle OR (hollow body but not a bottle)" adds subject-matter. (If it does not, then rewriting the claim is allowable and partial priorities work just fine even under the strictest approach.)

With this in mind, what is the harm in taking a pragmatic approach to partial priority and assigning it to anything that can be shown to be an embodiment of the more generic claim?

It would give the applicant the benefits of a disclosure of "a hollow body that is not a bottle" without his application actually disclosing that subject-matter. Why would that be justified?

Tim Jackson said...

Anon 09:56 - "Well, it is one thing to say that the conclusion reached in points 6.4-6.7 of G 2/98 is flawed ...."

That was the point I was making: the conclusion in point 6.7 is flawed, though points 6.4 - 6.6 may not be.

".... (even though T 1222/11 derives its approach from those points)."

As it had to. An ordinary BoA can't overrule the EBA, whereas in the current referral the EBA itself could do so.

"But expecting the EBA to overrule the order of G 2/98 goes a bit far.... In my opinion the order of G 2/98 is just fine and accepting the order and its implications means that T 1222/11's theory cannot stand."

I agree it's unlikely the EBA will make any change to the basic principle of what "the same invention" means in Art. 87(1) EPC. But I would hope for a change in how partial priority is determined under Art 88(3). There are several ways that might come about.

For example, the EBA could answer 'no' to Q2 of the referral. They could say that "a limited number of clearly-defined alternative subject matters" is not the correct test (overruling point 6.7 of G 2/98). In that case, T 1222/11's analysis of that phrase becomes irrelevant.

Or they may say that "a limited number of clearly-defined alternative subject matters" is correct, but that subsequent decisions have misinterpreted it (as T 1222/11 argues). See Q3.

Interesting to read document D26 in the referring decision. Or document D25 is a longer version in German. These articles are by Rudolf Teschemacher, who was one of the members of the EBA in G 2/98.

Dr Teschemacher's view is that "The sentence in G 2/98 that multiple priorities are acceptable provided that they give rise to the claiming of a limited number of clearly defined alternative subject-matters has been interpreted out of context". He hopes that other Boards will follow T 1222/11.

Anonymous said...

@Tim Jackson:
But I would hope for a change in how partial priority is determined under Art 88(3).
Why exactly? Just for the sake of solving the poisonous divisional problem? Are there any other real advantages?

Let me point out one consequence, assuming the order of G 2/98 survives.

P1: copper
P2: metal, preferably copper
application: metal, preferably copper. Claims metal.

D1 published between P2 and the application. It discloses metal.

Assume the answer to question 1 is "no".
Now P1 is the "first application" for the part copper of metal.
Therefore that part cannot get priority from P2 (it does from P1).
The remaining part is "metal but not copper". P2 does not disclose "metal but not copper", so the part "metal but not copper" of the claim does not have priority from P2. This is what the order of G 2/98 states.

So D1 is prior art under 54(2) for the claim part "metal but not copper".
Assuming D1 only discloses metal and does not exclude copper, that part is new over D1. However, it will most likely not be inventive to start from D1 and pick some other metal than pure copper. So the claim lacks inventive step.

I suppose this is not what you would like. But you can't pick and choose the consequences of answering "no" to question 1.

The Cat that Walks by Himself said...

There are two other interesting aspects in this subject.

First one is that there is a certain difference between "species vs genus" and "species vs a limited number of clearly defined alternative".

Second, patentee A of parent (species) and divisional (genus) applicatoin and patentee B (genus) should have the same rights with respect to 54(3)EPC.

Tim Jackson said...

To Anon at 9:30,

The answer to your point is in my post of 21:14 on 19 August.

If the answer to Q2 is 'no', then we no longer need to look for "alternative clearly defined subject matters". The EBA will have rejected that test, overturning point 6.7 of G 2/98.

So the alternative "metal but not copper" is now not relevant.

Instead we have overlapping priorities. 'Copper' has priority from P1. 'Metal' has priority from P2. The priority from P2 protects the claim from the intervening publication D1.

There's a residual complication under Art. 87(4) EPC. To the extent that the metal is copper, P2 is not the first application. So the claim is not entitled to priority from P2 to that extent. But to the extent the metal is copper, the claim has priority from P1. So there's no problem; D1 is still not prior art. (As in T 1222/11, there would be a problem if P1 was filed more than 12 months before the EP application under consideration.)

Note that throughout all of this, I'm looking only at the subject matters actually disclosed and claimed in the various documents. Namely, 'copper' and 'metal'. And I'm only looking at them to the extent that they are disclosed. There's no need to consider hypothetical subject matter which is not disclosed, like 'metal but not copper'.

Proof of the pudding said...

Anon 21 Aug 09:30 - are you sure that there is not another way of looking at this?

What part of the EPC (or the intention behind Articles 87 and 88 EPC) says that there needs to be a disclosure of "metal but not copper" in order for partial priority to "copper" to be effective without leading to the absurd situation that you describe?

What is wrong with instead viewing "copper" and "metal" as each having their own priority dates, but a claim to "metal" - when the application claims priority from P1 - as having both dates? This can be justified on the grounds that "metal" clearly encompasses copper and is a development (broadening) of the same invention.

Would not that "layering" approach (as opposed to your either / or approach) fulfil the legislative intent behind Articles 87 and 88 EPC?

I think that the Enlarged Board of Appeal needs to be bold here and provide an interpretation of Articles 87 and 88 EPC that serves the intended purpose behind those provisions - and the absurdities that arise from "strict" interpretations of G2/98.

Anonymous said...

"(I also don't see why n=3 would be novel over a document disclosing "an integer from 1 to 5". Is there case law that says otherwise?"

When I argue n = 3 was novel, I am arguing on the basis that there is a selection from more than one list, etc, and trying to stick to the point at issue. I may have a comma out of place somewhere also.

In the case of chemical compounds where there is a disclosure of a genus where the only variable is n (e.g. n = straight chain alkyl group -(CH2)n-), where n = 1 to 5 (subscripts for 2 and n should be imagined as I can't write them here), then n = 3 would not be a novel selection. However, such explanations distract from the priority/novelty discussion.

Anonymous said...

@Tim Jackson:
Instead we have overlapping priorities. 'Copper' has priority from P1. 'Metal' has priority from P2. The priority from P2 protects the claim from the intervening publication D1.

What you are saying here is that you want the usual strict approach when considering D1.

But you can't have it both ways. If you accept that "metal" has a partial priority from P1 to the extent that it covers copper, P2 is simply not the "first application" for metal to the extent that it covers copper. The applicant does not get to choose here. And unfortunately, G 2/98, order, prevents a partial priority for "metal but not copper" from P2.

(Maybe what you really want is a special rule for dealing with 54(3) prior art?)

The Cat that Walks by Himself said...

RE: Instead we have overlapping priorities. 'Copper' has priority from P1. 'Metal' has priority from P2.

Why 'Metal' would be new ?

RE: There's no need to consider hypothetical subject matter which is not disclosed, like 'metal but not copper'.

Tim Jackson, would you see any difference between 'metal but not copper' and 'metal other than copper'?

Anonymous said...

@Proof of the pudding:
What part of the EPC (or the intention behind Articles 87 and 88 EPC) says that there needs to be a disclosure of "metal but not copper" in order for partial priority to "copper" to be effective without leading to the absurd situation that you describe?

See the order of G 2/98:
"The requirement for claiming priority of "the same invention", referred to in Article 87(1) EPC, means that priority of a previous application in respect of a claim in a European patent application in accordance with Article 88 EPC is to be acknowledged only if 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."

Partial priority for an alternative subject-matter simply means that that alternative subject-matter enjoys a priority right. So that alternative subject-matter must be directly and unambiguously derivable from, i.e. disclosed by, the priority document.

There is nothing absurd about this. It just shows that the "relaxed approach" to multiple and partial priorities does not fit into the systematics of G 2/98. The usual interpretation of point 6.7, i.e. that each alternative subject-matter should be disclosed as such in the application as filed, does fit perfectly and, in my opinion, was clearly what the EBA had in mind. The usual approach follows not just from point 6.7, but from G 2/98 as a whole.

Again, poisonous divisionals should be fixed, but let's leave G 2/98 alone.

Anonymous said...

@Tim Jackson:
There's no need to consider hypothetical subject matter which is not disclosed, like 'metal but not copper'.

There is if you want to avoid poisonous divisionals.

The application discloses metal, preferably copper. Copper has priority from P1. A divisional is filed. Copper in that divisional has priority from P1.

Conceptually rewrite the claim as an OR-claim:
metal = copper OR X.
The claim is new only if copper is new and X is new.

If X is metal, then copper will be novelty destroying for X.

If X is "metal but not copper", then you are considering hypothetical subject-matter which is not disclosed.

Tim Jackson said...

Anon writes: "If you accept that "metal" has a partial priority from P1 to the extent that it covers copper, P2 is simply not the "first application" for metal to the extent that it covers copper."

Yes, I said that.

"And unfortunately, G 2/98, order, prevents a partial priority for "metal but not copper" from P2."

But we don't need one.

First of all, the EBA simply might not see any reason to get into this level of detail, leaving it as self-evident that between the full priority to P2 and the partial priority to P1 the applicant has all the bases covered.

However, if you do get into that detail, the issue would be that P2's full priority for 'metal' fails to the extent that the metal is copper. That's the extent to which P2 is a "subsequent application in respect of the
same subject-matter as a previous first application"
P1, according to Art. 87(4), but fails to meet the proviso that P1 should be dead at the date of P2's filing.

The partial priority to P1 then makes up the deficiency.

Note that this doesn't require there to be a partial priority for "metal but not copper" from P2. Instead, it's a partial failure of the full priority to 'metal'.

You might say that amounts to the same thing. The difference is how you get to it. You are rightly pointing to the difficulty under G 2/98 of ascribing a partial priority for something that's not disclosed in P2.

But instead, I'm taking a different approach which does conform with G 2/98. It applies G 2/98's order twice, first to the 'same invention' in Art. 87(1), and then by analogy to the 'same subject-matter' in Art 87(4):

-- First, the claim says 'metal' and P2 says 'metal'. They are the same invention under Art. 87(1). So far the priority claim for 'metal' is allowable.
-- But to the extent the metal is copper, that's also the same subject-matter as P1 under Art 87(4). So the priority just allowed fails under Art. 87(4). But it only fails to that extent.

~~~~~~~~

Cat that Walks asks: "... would you see any difference between 'metal but not copper' and 'metal other than copper'?"

I think it's slightly more subtle than that, as explained above. P2 gives full priority for 'metal', except to the extent that the metal is copper. There's no separate priority for 'metal but not copper' or 'metal other than copper'.

Tim Jackson said...

Anon 15:11 writes: "Again, poisonous divisionals should be fixed, but let's leave G 2/98 alone."

How about if we leave the main Conclusion of G 2/98 alone, i.e. the requirement that the 'same invention' must be directly and unambiguously derivable from the priority document? How about if we just address point 6.7 as has been suggested?

There would then be no requirement for "a limited number of clearly defined alternative subject-matters". No need for you to try to ascribe a priority date to any alternative subject-matter. And no need for any alternative subject-matter to be directly and unambiguously derivable from the priority document.

Anonymous said...

The problem arises that under the EPC, the disclosure of P1 stating "copper" is different to P2 disclosing copper but referring to metal. The technical content is the same to all intents and purposes - the EPO though has the narrow approach to added subject matter. In the US, for example, I would have thought there would be no problem with adding a claim to metal in P1 on the grounds that copper is a metal. This is the area which needs to be worked on, not the "partial priority" aspect. When I was developing my thoughts on poisonous divisionals, it became clear that the narrow approach to priority and added subject matter opened a perfectly logical attack even if in the real world it should not fly.
EdT

Tim Jackson said...

Anon 16:41 writes: "Conceptually rewrite the claim as an OR-claim:"

If the EBA holds that point 6.7 is wrong and there's no need divide the claim up into alternative subject matters, why do we need to conceptually rewrite it?

The claim of the parent says 'metal'. It has a partial priority from P1 to the extent that the metal is copper. Having junked all the case law that was based on point 6.7, this partial priority doesn't depend on hypothetical considerations of alternative subject matters in rewritten claims.

This partial priority from P1 is now identical to the divisional's partial priority from P1. So that's the end of the matter.

Proof of the pudding said...

Anon 21 Aug 15:11 - sorry to be pedantic, but you have not pointed to any provision of the EPC (or a legislative intention behind such a provision). Whilst clearly important, a decision of the EBoA interpreting the legislation is really not the same thing - especially as it is perhaps possible that you might not be interpreting that decision correctly.

There was a typo in my earlier post and I should therefore clarify that my wish is for the EBoA to provide an interpretation that serves the intended purpose behind Articles 87 and 88 EPC and that avoids the absurdities that arise from "strict" interpretations of G2/98.

The purpose of Articles 87 and 88 EPC is clearly not to provide "bear traps" for unwary applicants (e.g. foreign applicants who cannot be expected to be intimately familiar with all of the intricacies of drafting practices under the very latest EPO case law). Thus, any interpretation of those Articles that produces such a result must be wrong.

Remember, the interpretation of Article 88 EPC really ought to be consistent with Article 4F of the Paris Convention, which states that:
"No country of the Union may refuse a priority or a patent application on the ground that the applicant claims multiple priorities, even if they originate in different countries, or on the ground that an application claiming one or more priorities contains one or more elements that were not included in the application or applications whose priority is claimed, provided that, in both cases, there is unity of invention within the meaning of the law of the country"

Hard to see how poisonous divisions are consistent with this legislative intent, is it not?

Anonymous said...

@Proof of the pudding:
sorry to be pedantic, but you have not pointed to any provision of the EPC (or a legislative intention behind such a provision). Whilst clearly important, a decision of the EBoA interpreting the legislation is really not the same thing - especially as it is perhaps possible that you might not be interpreting that decision correctly.

Did I ever claim there was such an explicit provision in the EPC? In my comment of 21 Aug 09:30 I already referred to the order of G 2/98. Did you read the order?

Hard to see how poisonous divisions are consistent with this legislative intent, is it not?
Not difficult at all. Art. 4F PC merely states that an application cannot be refused on the ground that it extends beyond the content of the priority document. That is pretty basic stuff nowadays and the EPC clearly complies with it. The EPC even goes further and does not require there to be unity between the application and the priority document.

More importantly, I thought I have been rather clear in stating that and explaining why the problem of poisonous divisionals needs a solution. But somehow there is this strange craze about partial priorities that is entirely unnecessary to achieve that aim.

Tim Jackson said...

Anon 00:38, you agree that the problem of poisonous divisionals needs a solution. What solution do you propose without involving partial priorities?

Proof of the pudding said...

Anon 22 Aug 00:38 - My mistake. When you quoted my question, I assumed that what followed immediately afterwards was your attempt at answering that question. I did not realise that you were instead discussing something different.

It looks like you have something against using priority to solve the issue of poisonous divisionals. However, given that the difference between a divisional and a 3rd party document is a claim to (partial) priority, I am not sure that it is worthwhile looking elsewhere for a workable solution. However, chacun à son gout, as they say.

With regard to the Paris Convention, I am not so sure that the legislative intent was as narrow as you imply.

A benefit of priority provided by Article 4F is that your application cannot be rejected solely upon the basis that the disclosure has been broadened relative to the disclosure of the priority document - provided that the disclosure retains unity of invention.

However, Article 4G(2) goes on to say that all divisional applications will "preserve... the benefit of the right of priority".

If a divisional retains the benefit of priority provided by Article 4F, it appears to me that an application that is divided cannot be rejected solely upon the basis that the disclosure has been broadened relative to the disclosure of the priority document.

However, is not the broadening of the disclosure relative to that of the priority document the entire basis of the "poison" in poisonous divisionals?

There are two other elements that need to be in place for the poison to kick in, namely retaining the original (narrow) disclosure in the later application and including a claim to the broader disclosure. However, I cannot see why either of these elements would make a blind bit of difference when it comes to assessing whether a divisional application can benefit from the combined effects of Articles 4F and 4G of the Paris Convention.

Tim Jackson said...

Anon 00:38, on re-reading the thread, I guess you might be the same Anon who was previously suggesting that the poisonous divisionals problem should be fixed by carefully interpreting Art. 54(3)?

OK, so I think you would leave partial priorities to be determined on the basis of point 6.7 of G 2/98, as at present? That is, only if it "gives rise to the claiming of a limited number of clearly defined alternative subject-matters".

But then I guess you would favour a 'no' to Question 5 of the referral, so that a divisional and its parent couldn't be cited against each other under Art. 54(3)?

However, I suspect that poisonous divisionals might not be the only problem derived from the requirement for a "limited number of clearly defined alternative subject-matters" if partial priority is to be granted.

Here's a 54(3) scenario which I think might still be a problem with your approach. Comments welcome: I'm trying to get this clear in my mind.

~~~~~~~

Scenario:

-- On 1 January 2015, a UK company Whitehat Ltd files a UK application GB1 which describes and claims a species X1.

-- They continue their work, and as a result they find that the invention works across a broader range. On 1 June 2015, they file a European application EP1, claiming priority from GB1. Thereafter, GB1 is abandoned and never published.

-- EP1 claims a genus X, which encompasses the species X1. EP1 describes embodiments X1, X2, X3 etc which provide support and enablement under Arts. 83 and 84 across the range for the genus X. Nevertheless, the genus is claimed generically; it would cover many other embodiments too.

-- Meanwhile, Blackhat Inc (Whitehat's US competitor) has also been working. They independently invent the species X1, but they are behind Whitehat. On 1 May 2015, they file a US patent application US1 which describes and claims X1.

-- But Blackhat never do any more work. They file a European application EP2 on 1 May 2016, which is identical to US1 and claims its priority. But it only describes and claims the species X1, not the genus X. It is published on 1 November 2016.

-- Subsequently, Blackhat's EP2 is cited against Whitehat's EP1 under Art. 54(3). What is the outcome?

I think the outcome ought to be that Whitehat's broad claim to the genus X should survive. Under Art. 54(3), their partial priority from GB1 (1 January 2015) is earlier than Blackhat's priority from US1 (1 May 2015). And Whitehat are the ones who did the work; Blackhat never even contemplated the broad genus X.

But when Whitehat's EP1 is examined (or reaches an EPO Board of Appeal, or the UK Patents Court), someone refers to point 6.7 of G 2/98, and the case law which it has spawned. Whitehat's broad claim to the genus X can only have partial priority from GB1 if this "gives rise to the claiming of a limited number of clearly defined alternative subject-matters".

Does the broad claim to the genus X meet that requirement? Under most of the case law, no. The broad claim covers many embodiments, not a limited number of clearly defined ones.

So under the case law, Whitehat's meritorious broad claim is invalid. True, they can have a narrow claim for X1 (and perhaps for X2 and X3). And as consolation, Blackhat's EP2 is also invalid. But inspired by Whitehat's original work on the genus X, the way is clear for Blackhat and other competitors to develop embodiments X4, X5 etc.

~~~~~~~

Can anyone see a flaw in this reasoning? If not, I think that re-interpreting Art 54(3), as proposed by Question 5, is never going to be the answer. The problem is broader than poisonous divisionals. It lies with how partial priority is interpreted under point 6.7 of G 2/98.

Anonymous said...

An interesting post, Tim. This illustrates for me how the concept of "partial priority" is doomed to failure. Whitehat has priority for X1 alone. Any subsequent application claiming broader than the individual solutions (x1, X2, etc.) will be anticipated by blackhat's filing. This would not be the case if the EPO were to allow a broader concept of the "same invention" but I fear this is so firmly entrenched, it is not going to hppen.
EdT

MaxDrei said...

Now you're talking, Tim and EdT. If you are still there, can you comment on what I write below, on a thought which occurred on reading your most recent exchange.

In licensing, we have the concept of "franked" goods, on which the licence fee has been paid. Can something like that be applied to Tim's scenario? What I mean is, looking at Whitehat's claim to genus X, part of its scope, that which corresponds to species X1, can be thought of as being "franked", back to 1st Jan 2015, by GB1. US1, directed to X1 but filed later, is disabled as a 54(3) attack by this "franking" notion. This is the Tufty Approach, is it not?

Tufty works with self-collision because there is never one iota or scintilla more of disclosure in the 54(3) reference than in the attacked case. It would also work for Tim's White/Black scenario if US1 contains not ons iota more disclosure about X than White's earlier GB1. But in real life, that's highly improbable. When two patent attorneys independently priority file on Discovery X1, their respective claims might be identical but surely their respective disclosures won't be. For Tim's hypo though, I guess we are to asume that the respective White and Black prio doc disclosures are verbatim the same?

The EBA has on occasion reversed itself. It handed down G2/98 before the argument over "partial priority" and its repercussions grew so intense. I would think there should be no shame or loss of face to the EBA if it has another crack at the problem, and replaces G2/98 with something more intellectually satisfying.

Mind you, then Whitehat and Blackhat both discover X1, both take that discovery and run with it, and both file their own stream of patent applications, it is inevitably going to be messy sorting out who gets what. We just need a high quality legal roadmap to guide us through the treacherous swamp of facts that real life cases exhibit.

Is "franking" a concept that one can do anything with, when it comes to 54(3) law and poisonous priorities and divisionals? It carries echoes of the UK Patents Act 1949, doesn't it? Section 52 was it? Couldn't the EBA reach back to 1973, look at Section 52 and say it was not a bad idea, which even now could inspire a solution to the problem of "self-collision"?

MaxDrei said...

Afterthought: It occurs to me that a very similar problem arises under the grace period provisions of the AIA in the USA. What is the scope of the "grace" given? Is it exactly what the inventor makes available to the public before the prio date? Or is it that plus all equivalents to it? Still not settled, I believe.

Anonymous said...

My preference for a test for a valid priority would be if the priority filing supports the claimed subject matter rather than whether the claimed subject matter is disclosed, such that the recital of a species (or perhaps better, two) in the priority filing entitles the applicant to priority for the genus if the latter would be considered to be supported by the disclosure.
EdT

Tim Jackson said...

Max, perhaps what you call "franking" is just the natural way in which novelty and priority ought to work with each other under the EPC.

Both rely on the same test: that subject matter should be "directly and unambiguously derivable" (at least implicitly). In the case of priority, this stems from the main Conclusion of G 2/98.

So if the EBA were to answer 'no' to Q1 of the referral, novelty and priority would counterbalance each other exactly, as follows:

For novelty, the disclosure of a species can destroy the novelty of a genus which is claimed generically. More precisely, that follows when the generic claim encompasses a species which is directly and unambiguously derivable from the prior art disclosure.

For full priority, the situation is different. Under G 2/98, a generic claim for a genus only gets full priority if the genus itself is directly and unambiguously derivable from the priority document. The disclosure of a species doesn't achieve that.

But then Art. 88(3) comes to the rescue with the Paris Convention notion of partial priority. A generic claim can have partial priority from a species. More precisely, that follows when the generic claim encompasses a species which is directly and unambiguously derivable from the priority disclosure.

Do you see what I did there? I copied and pasted the same phrase for novelty and partial priority. The partial priority test is the same as the novelty test.

So if you are considering the same species and the same genus, as in an Art. 54(3) poisonous priority situation, then the partial priority exactly counterbalances the novelty attack. This means that the novelty attack must fail. If you want to call that "franking", then fine.

So why have things gone wrong with poisonous priorities? It's because G 2/98 pronounced the "directly and unambiguously derivable" test for full priority, but didn't just leave it at that. It also added point 6.7, requiring "a limited number of clearly defined alternative subject-matters". That has been taken as an additional test for partial priority, leading to an imbalance between the tests for novelty and partial priority.

Interestingly, if you read point 6.7 it's actually talking about multiple priorities, not partial priorities. Maybe we need to look at the subsequent case law to find out how it came to be applied it to partial priorities.

By the way, this still doesn't go as far as s.52, PA 1949, since it's novelty only. As we've discussed previously, s.52 also "franked" against obviousness attacks.

THE US anon said...

MaxDrei,

An interesting question, but one that has a ready response: ordinary understanding.

Scope discussed follows what scope is - there is no NEW definition of "scope" that would supplant the established understanding under law.

Unless you can demonstrate a clear rationale for somehow wanting to have a different understanding of what the word in its ordinary sense carries (that would be the natural use of equivalents), you appear to want to hide an elephant in a mousehole.

MaxDrei said...

I have just read Tim Jackson's of today at 18:32 and it left me feeling like I want to mutter "Hallelujah". I'm thinking, wouldn't it be nice if the EBA were to read it too, and then adopt it for their own forthcoming Opinion.

EdT, I'm sceptical, about setting up one standard "supported by" for priority and a different one for novelty. I'm thinking "undisclosed intermediate generalsations", prosecution amendments, and Art 123(2) EPC. Do you want the 123(2) standard also to be loosened as far as "supported by", do you want it to stay strict, directly and unambiguously derivable, or do you want something in between? Are you advocating one standard for intermediate generalisations during the Paris year and a different one after the year is over?

Tim Jackson said...

EdT writes: "My preference for a test for a valid priority would be if the priority filing supports the claimed subject matter rather than whether the claimed subject matter is disclosed, such that the recital of a species (or perhaps better, two) in the priority filing entitles the applicant to priority for the genus if the latter would be considered to be supported by the disclosure."

If we're going down that route, I'd add that the priority filing should also enable the claimed subject matter across the range (Art. 83) as well as supporting it (Art. 84).

Realistically, of course, there's no chance. This was the question decided by G 2/98. The EBA is not going to overturn it's main Conclusion, even if they can be persuaded to revisit point 6.7.

Wouldn't it be nice if extension of the subject matter (Art. 123(2)) was also decided by whether the original disclosure supported and enabled the revised claim, rather than by whether it was "directly and unambiguously derivable"? I'm not sure, but maybe that's closer to the US view of added subject matter, and is why Americans can't understand the European approach?

But realistically that's not going to change, either.

Anonymous said...

Yes.

MaxDrei said...

So now, Tim Jackson, EdT, we are on to the issue whether claims are disclosures. If they are, and if the disclosure of a document ought to be the same for Art 54 as it is for Art 87 or 123(2) EPC, then you can't go for a "supported by" standard for Art 87 and 123(2) but a "d + u derivable" standard for Art 54. I don't fancy facing novelty attacks in which I am told that the disclosure of D1 is enough to "support" an objection under Art 54. That's one reason why I support a "d + u derivable" standard for the allowability of prosecution amendments, whether in the Paris year or afterwards. Another reason is that it strengthens the degree of confidence I have in my clearance opinions.

You can have a loose "supported" standard for Art 87 and 123(2) issues under First to Invent. I think it ain't right in First to File territory, where the only thing that determines priority between rival filers is what they disclosed to the PTO in their respective patent applications as filed. For that very reason, I expect to see the courts of appeal in the US, when they are first confronted by priority contests under the AIA, getting progressively stricter on "same invention" and "added matter" issues, until they arrive, in 40 years or so, at something like the EPO's "d + u derivable" strict standard.

Anonymous said...

Probably my point of view is tainted by the work I did one developing the "poisonous" divisional attack where the parent had a wider scope of protection than the original filing but with essentially the same technical disclosure, rather than the addition of new species together with a more generic claim.

It comes down to the correct interpretation of "same invention" Should the novelty/added subject matter test be applied or should some leeway be given to the applicant to redefine their invention, perhaps following the results of a search on the priority filing? The latter to me makes sense and hence the test is more in line with is the new claim supported by the original disclosure rather than actually disclosed.

EdT

MaxDrei said...

Thanks EdT. In what follows, I'm not necessarily disagreeing with you, just teasing out the issues.

UK law always used to give Applicant a one year grace period (following the priority date) for drafting claims that are "supported by" the disclosure of the priority document. The EPC gives you just as much time to file your claims. It is just that to meet the "supported by" standard you have to state your invention to the EPO in terms verbatim to your future claim, already at the priority date.

Inside the EPO, they are quick to put all the blame for lost rights on "poor drafting". They don't know how hard it is! But isn't it hard to deny, that you should be required to state what your invention is, at the widest level of generality you can imagine, already by the time you file on it, at the Patent Office. Once you have found that widest level of generality, it should be more or less routine, to step down in scope, through a nice set of dependent claims.

Why does a patent attorney of ordinary competence first need to see the results of the EPO search before she (or he) can draft a suitably wide claim 1?

Anonymous said...

Let's take an example of a case originating in the US.

The priority filing has a single independent claim and all dependent claims referring back, as is their practice, only to claim 1. Following a search, claim 1 is shown to be too narrow but on the information available, claim 2 looks fine. A PCT application is filed with the switched on attorney including multiple dependencies and also the original statements of invention.

On entering the regional phase, the new claim 1 is shown to be obvious so the applicant combines claims 1 and 2 (claims 2 and 3 from the priority filing). The application is allowed. A divisional is now filed combining claims 1 and 3 from the priority filing.

Claim 1 of the allowed EP is not entitled to priority since the claimed combination of features was not originally disclosed and the divisional now anticipates the allowed claim. Is this really fair to the applicant?

EdT

Anonymous said...

@Tim Jackson:
Anon 00:38, on re-reading the thread, I guess you might be the same Anon who was previously suggesting that the poisonous divisionals problem should be fixed by carefully interpreting Art. 54(3)?

Yes, the main idea being that Art. 54(3) should not be understood as allowing self-collision, i.e. subject-matter with a particular effective date disclosed in the EP itself cannot form prior art under Art. 54(3) for that EP. In other words: the fact that subject-matter with a particular effective date is disclosed in the EP itself with that same effective date should protect that EP against that subject-matter under Art. 54(3) (even if that subject-matter with that same date is disclosed in a divisional or parent or even some unrelated published European application).

This rule has no explicit basis in Art. 54(3), but I think it makes sense (self-collision being ridiculous) and I believe it precisely solves the problem without having unintended or unexpected effects in other areas of patent law. I would prefer it over a rule that only prevents parents and divisionals (and the EP itself) from being cited against an EP, as that would discriminate against applicants who file multiple applications having identical descriptions (for example claiming a different invention in each).

OK, so I think you would leave partial priorities to be determined on the basis of point 6.7 of G 2/98, as at present? That is, only if it "gives rise to the claiming of a limited number of clearly defined alternative subject-matters".

Yes. I understand an OR-claim to be a claim that could be rewritten as a number of separate claims within the limits of Art. 123(2). The formulation of point 6.7 might not be optimal:
- "clearly defined" seems to refer to Art. 84, but I think that is a different matter. I would prefer "individually disclosed".
- the subject-matters do not need to be strictly "alternative" in the sense of being disjunct. For example, a claim to "a composition comprising one of substances A and B" might be rewritten as two claims "a composition comprising substance A" and "a composition comprising substance B", both "alternative subject-matters" covering compositions comprising substances A and B. So the alternative subject-matters may overlap (unlike what happens in the "to the extent that" approaches).

(Continuation in the next post.)

Anonymous said...

-- Subsequently, Blackhat's EP2 is cited against Whitehat's EP1 under Art. 54(3). What is the outcome?

Whitehat should be allowed to disclaim X1 under G 1/03. Unfortunately, G 2/10 prevents that in this specific case, because X1 happens to be disclosed in EP1. I find it difficult to agree with G 2/10 on this point.

Whitehat is obviously allowed to claim X1 separately: EP1 discloses X1, and X1 has priority from GB1.

If Whitehat is allowed to have a claim to X minus X1 (ignoring G 2/10) and a claim to X1, then I see no reason to be formalistic and require 2 separate claims. Whitehat can have its claim to X. In fact this corresponds to my understanding of multiple priorities: the generic claim to "X" may be rewritten as two separate claims "X minus X1" and "X1", so both alternatives within "X" have their own priority.

(Of course, if the G 1/03 approach would be applicable to undisclosed disclaimers of disclosed subject-matter, then the poisonous divisional problem would mostly disappear by the same method. The further conditions of G 1/03 might still present complications, though.)

Let's change the scenario a bit: EP2 only discloses XX1, which is more specific than X1.
Now Whitehat can disclaim XX1 to get X minus XX1 (G 1/03). It can also claim X1. Both claims together cover X, so the claim to X should be allowed as it is possible to rewrite is as separate claims to X minus XX1 and to X1 (or a claim "X minus XX1" OR X1). That G 2/10 stops this from working if XX1 happens to be not narrower than X1, but identical to X1, seems unjustified.

If EP2 discloses not X1 or XX1 but X4, then Whitehat can get X minus X4 (G 1/03). Whitehat may want to add a separate claim to X1, or claim "X minus X4" OR X1, in case X1 and X4 overlap.

So I don't think a more liberal ("to the extent that") approach to partial priorities is necessary here, and it could not replace G 1/03 anyway.

MaxDrei said...

Fairness is all very well Ed, but let's be equally fair to everybody involved. Not only Applicant but also the 2nd to file, one day later, with a similar embodiment but a different statement of invention. Novelty law invokes the concept of "planting the flag". So in what sense does one "plant a flag" when one files a priority document? Is it enough to describe a specific prototype to the Patent Office? Or should "planting the flag" involve a requirement for a "statement of invention" in my patent application, which defines my inventive concept, at the level of generality I visualise for my invention? I had thought that the value a competent drafter adds is to define the metes and bounds of the inventive concept, as opposed to doing no more than announcing the "best mode" so far known to you.

Anonymous said...

@Tim Jackson:
But then Art. 88(3) comes to the rescue with the Paris Convention notion of partial priority. A generic claim can have partial priority from a species. More precisely, that follows when the generic claim encompasses a species which is directly and unambiguously derivable from the priority disclosure.

Do you see what I did there? I copied and pasted the same phrase for novelty and partial priority. The partial priority test is the same as the novelty test.


It seems to follow from G 2/98, point 6.2, last sentence, that "elements of the European patent application" should be understood as subject-matter disclosed in the application: Both an "element of the invention" and an "element of the European patent application" actually constitute subject-matter as defined in a claim of the European patent application.

Your "partial priority" seems to be a different notion than regular / full "priority". But there is nothing like that in the EPC. Something (i.e. particular subject-matter) has full priority or not at all. It has full priority if it is disclosed in the priority document. (In case of multiple priority documents, different subject-matters may have different priorities.)

See also point 5 of G 2/98: "the protection from novelty destroying disclosures during a period of twelve months from the date of filing of the first application is necessary only in case of the filing of a subsequent application relating to the same invention". If the subsequent application claims a more general invention, it does not claim the same invention, so there is no right to priority.

It may be possible to interpret Art. 88(3) differently than I do here, but that is not what the EBA did in G 2/98. I can find no trace in G 2/98 of an "to the extent that" approach that assigns priorities to "elements" that are fractions of disclosed subject-matters.

An example where your approach seems to go "wrong":
P1 discloses copper.
The application claims metal.
D1 is prior art with a date between the dates of P1 and the application.
D1 discloses metal and does not specifically mention that the metal may be a metal other than copper.

In your approach, the claim to metal has priority from P1 to the extent that it covers copper. It has as effective date the filing date of the application to the extent that it does not cover copper, i.e. for "metal but not copper".

Now D1 is only prior art for the claim for the part "metal but not copper". Since it does not disclose an embodiment excluding copper, the part "metal but not copper" is new over D1.

If D1 is prior art under Art. 54(2), then the claim will likely lack inventive step. But lack of novelty seems to be the more appropriate outcome here.

If D1 is prior art under Art. 54(3), then the claim will survive. That does not seem right.

With the strict approach the same situation occurs if the application happens to have a basis for "copper" and "metal but not copper". A claim to "copper" would have priority (and anyway be new over D1) and a claim to "metal but not copper" would be new over D1. But now novelty is achieved on the basis of a feature disclosed in the application, so that simply has to be accepted.

Anonymous said...

@Tim Jackson:
An improvement of my example.

P1 discloses copper.
The application is filed one day later and discloses and claims metal. It does not mention copper.
D1 is an EP filed before P1 and published after the filing date of the application. Its description, claims and drawings are identical to those of the application.

D1 clearly is prior art under Art. 54(3).

According to your approach, metal has a partial priority from P1 for the part "copper". The remaining part has the filing date.
As explained in the memorandum and in G 2/98, partial and multiple priorities are about OR-claims. An OR-claim does not need to be written out with an actual "or" in it, but it does have such a structure: A or B, and the claim is new precisely if both A and B are new.
In this case we have "copper" and "metal but not copper" (or however you want to word the remaining part). Both parts are new over D1. So the claim to metal is new over D1!!

This example shows that it is crucial that each "alternative" of an OR-claim is individually disclosed in the application itself. Otherwise, the "conceptual" rewriting of a claim into an OR-claim introduces features that have no basis in the application as filed. Those features can give the applicant an unwarranted advantage.

MaxDrei said...

Are we not in danger here of making things more complicated than they need to be?

Let us first remind ourselves that what a claim "covers" is not necessarily the same as what it "discloses".

What does the undisclosed disclaimer look like, that finesses out of a 54(3) novelty attack? Now applying that to anon's P1/D1 hypo above, one would have to disclaim all the disclosure of D1, leaving nothing left to claim.

But when the 54(3) cite is your own, why should you have to "disclaim" any part of what the claim covers, for the reality is that it's all yours and does not belong to some other party who got to the Patent Office before you did.

As Tufty tried to explain,a long time ago now, the "splitting" of the claim into two coverage zones, one entitled to priority, the other not, is not a "real" splitting, to be effected in writing, in real words, warts and all, but, rather, just a notional one, for which there is no call to express in words.

As I wrote above, sorting out between rival applicants "who gets what" might be fiendishly difficult. Moreover, that particular difficulty is unavoidable. Compared with that irreducible problem, self-collision can just be waved away by the EBA, if it chooses, with a mere thought experiment.

Anonymous said...

Oddly,

"Let us first remind ourselves that what a claim "covers" is not necessarily the same as what it "discloses"."

Sounds as if you are making things more complicated than they need to be.
It is the application as a whole that serves the function of "discloses." So while it is without question a matter of fact that the claims are part of the disclosure, it is entirely unclear why you are trying to make this something about the claims alone.

Anonymous said...

@MaxDrei:
I'm not entirely sure what point you're trying to make in relation to my P1/D1 example(s).

What does the undisclosed disclaimer look like, that finesses out of a 54(3) novelty attack? Now applying that to anon's P1/D1 hypo above, one would have to disclaim all the disclosure of D1, leaving nothing left to claim.

This is of course true. If D1 is 54(3) prior art and discloses what is claimed in its full generality, applying G 1/03 does not leave you with anything. But this is exactly what G 1/03 intended. The invention belongs to the first one who disclosed it, i.e. the one who filed D1.

But when the 54(3) cite is your own, why should you have to "disclaim" any part of what the claim covers, for the reality is that it's all yours and does not belong to some other party who got to the Patent Office before you did.

If D1 is your own but really filed earlier (so we're not looking at a poisonous divisional situation), there is no reason why D1 should not be citable against the application. It is not relevant for 54(3) that D1 is your own. The applicant can get the protection he deserves through D1. (If D1 is a poisonous divisional, then everybody seems to agree that a solution has to be found one way or the other. The main discussion is about the best way to achieve that aim.)

As Tufty tried to explain,a long time ago now, the "splitting" of the claim into two coverage zones, one entitled to priority, the other not, is not a "real" splitting, to be effected in writing, in real words, warts and all, but, rather, just a notional one, for which there is no call to express in words.

Even Tufty probably agrees that the notional splitting results in a notional OR-claim (see memorandum, see G 2/98) of the notional form A OR B, and that such a notional OR-claim is new precisely if both notional parts A and B are new. That is the whole point of the notional splitting: Tufty uses it to notionally get two parts that are both new and concludes that the original claim is therefore new. Without rewriting it.

What my examples show is that this may result in undesirable outcomes. Metal becomes new over metal, without rewriting it, by notionally splitting the claim in "copper" and "metal but not copper", or even "metal with melting point >= 1000 degrees celcius" and "metal with melting point < 1000 degrees celcius". Both notional parts are new over a disclosure of metal.

How will Tufty solve this? Also allow a notional splitting of prior art disclosures? Reason that a disclosure of metal notionally discloses copper? Then nothing will be new anymore.

The "notional" and the "to the extent that" approaches work reasonably well if the prior art discloses single embodiments, but they fall apart if the prior art discloses the genus. This seems to have been overlooked.

It may still be possible to save Tim's approach (or maybe not, I have not tried yet myself), but at the very least it will have to abandon the idea that partial and multiple priorities are about OR-claims. And that gets you pretty far from the memorandum and G 2/98.

MaxDrei said...

Sorry I was not clearer. I had in mind the situation where D1 is a patent family member. Self-collision with your own earlier filings, non-family, is just as problematic as when D1 belongs to somebody else, I agree.

As you say, everybody agrees that self-collision of family members is a nonsense. What's not yet delivered to us by the EBA is the killer argument to get rid of the nonsense.

slartibartfast said...

Anon 16:39 - can you explain what you mean that the "notional"/"to the extent that" approaches "fall apart if the prior art discloses the genus"?

I think that the "issue" is really (or, at any rate, should really be) a non-issue and there is an awful lot of overcomplication going on in the discussions both here and elsewhere. To my mind T1222/11 is surely the correct approach and in line with the memorandum in the travaux preparatoires. I have yet to see a convincing explanation as to why this should not be correct - or at any rate an explanation which does not rely on an overly literal interpretation of G2/98 point 6.7.

Sticking with the "copper"/"metal" example:

Application 1 (let's call it "P1") discloses copper.
Application 2 (let's call it "A1") claims metal.

A1 claims priority from P1. P1 publishes and is therefore comprised in the state of the art under Art. 54(3) EPC to the extent that A1 is not entitled to the priority date of P1.

In this instance, as pointed out by Tufty and many other commentators, we can notionally divide the claim into "copper" and "other metals". Copper is entitled to priority from P1's disclosure of copper, therefore P1 is not prior art for that portion of A1's claim. "Other metals" are only entitled to the filing date of A1, so P1 is Art. 54(3) prior art for that portion of A1's claim, but P1 only discloses copper and so cannot anticipate "Other metals".

Simple, no?

Now let's consider that P1 instead "discloses the genus" as mentioned by Anon 16:39. Here presumably we mean that P1 discloses "metal". But in that case there is absolutely no question that A1's claim to "metal" is fully entitled to P1's priority date. P1 cannot constitute prior art against A1's claim, even if P1 is allowed to publish. There is no problem!

What about the intermediate situation? Say that P1 discloses "d-block metals" (i.e. a genus) and A1 claims "metals" (a super-genus encompassing the genus of P1). Here I cannot see any difference from the first situation above, in principle. We can notionally divide the A1 claim into "d-block metals" and "non-d-block metals". As before, A1 is entitled to priority for d-block metals and so P1 is not prior art against that notional portion of the claim. A1 is not entitled to priority for non-d-block metals, but P1's disclosure of d-block metals cannot anticipate A1's claim to non-d-block metals. Again, there is no problem.

Perhaps I've misunderstood Anon 16:39's remark?

I also don't understand what you mean by "notionally splitting prior art disclosures" - this seems to be overcomplicating the issue needlessly. I can't imagine why you would need to envisage splitting the prior art when the issue is really all about the priority entitlement of the new application. Perhaps someone can explain?

Anonymous said...

@slartibartfast:
Anon 16:39 - can you explain what you mean that the "notional"/"to the extent that" approaches "fall apart if the prior art discloses the genus"?

I've already explained it, but I'll do it again.

First note why the T 1222/11 is so "wonderful". In your example it allows you to rewrite "metal" as "copper" OR "other metals". The first part has priority, the second part is new over "copper". Since both parts are new over "copper" disclosed in a divisional, "metal" is new. Just as one would like.

But now suppose there is some "real" 54(3) prior art disclosing "metal" (so not a poisonous divisional). This prior art disclosing "metal" SHOULD take away novelty of the claim to "metal".

But it does not. The claim to metal can still be (notionally or whatever) rewritten as "copper" OR "other metals". Copper is new over "metal". "Other metals" is new over "metal". So the claim to "metal" is new over "metal".

Such ridiculous outcomes can only be avoided by requiring that the (notional or whatever) rewriting of a claim into an OR-claim does not extend beyond the content of the application as filed, i.e. each alternative subject-matter should be individually disclosed in the application as filed.

I have already explained how the poisonous divisional problem should be solved without creating the mess that the "partial priority" approach would make.

I also don't understand what you mean by "notionally splitting prior art disclosures"

That was sort of a joke. In a desparate attempt to repair the ridiculous outcome that "metal" can be made novel over a disclosure of "metal" by conceptually/notionally/whatever rewriting metal as an OR-claim (e.g. copper OR other metals, or "metal with a melting point >= 1000 degrees" OR "metal with a melting pint < 1000 degrees"), one could argue that this notional splitting should also be applied to the prior art disclosure itself. But then things quickly get even beyond ridiculous, as "metal" would be disclosing everything that is encompassed by it. So this is not serious.

Anonymous said...

Most of the above comments relate to the example of copper/metal, similar to the chlorine/halogen situation. This is, however, only one, essentially chemistry based "species/genus" scenario.

The original "poisonous divisional" attack was however based on a description which was common to the priority, parent and divisional filings, with the claimed subject matter varying, in a simplified manner, narrow/broad/narrow respectively. This is somewhat similar to the example I gave above with the US based priority filing.

For such situations, I have not been able to see how the "partial priority" type arguments help. If the concept of poisonous divisionals is considered to be an undesired consequence of the EPO's case law on the entitlement to priority and on the filing of divisional applications, either the enlarged board has to come up with a artificial interpretation of the wording of the EPC (along the lines of the straw man opposition solution) or revisit the established case law as to the meaning of "same invention".

Perhaps the solution lies in amending the EPC.

EdT

Dave said...

The idea that the claim can be notionally split for the purpose of determining novelty is not something that Tufty (or I) would agree with. The notional splitting is only for the purpose of resolving priority issues. It would be stupid to apply it to assessing novelty.

MaxDrei said...

All this talk of claim splitting reminds me of Robin Jacob's cautionary words, to the effect that we can learn much from the Americans: we should observe closely what they do; and then make sure we don't repeat their mistakes.

Remember how to "provoke an interference" to resolve a priority contest? You copy their claim into your application. What a shambles: their chosen terminology in the claim, supported by a disclosure written in your chosen terminology. Perhaps the result doesn't trouble them much in bio/pharma, where terminology is universal, but imagine it in another field where priority contests are routine and everybody has their own individual terminology, namely, strut matrices in trans-luminal stents to keep you alive. For example, suppose you and the other bloke both file on the same stent matrix improvement. Your struts and also his enhance both stenting force and flexibility along the lumen. Quite some achievement! But you claim struts that are "serpentine" while in his claim they are "zig-zag". Is that the "same" invention? Does "serpentine" mean "bends not elbows" while "zig-zag" means "elbows not bends"? I think not but who knows for sure? In engineering performance, are his "zig-zag" struts the same or different from, or equivalent to your "serpentine" ones (especially when one of you forms your matrix by laser-cutting and the other by chemical etching).

So, here at the EPO, let's not go down that route. Let's confine ourselves to self-collision between patent family members, and even then, let's keep the claim splitting no more than notional, shall we?

Anonymous said...

@Dave:
The idea that the claim can be notionally split for the purpose of determining novelty is not something that Tufty (or I) would agree with. The notional splitting is only for the purpose of resolving priority issues. It would be stupid to apply it to assessing novelty.

Avoiding poisonous divisionals by notional splitting is all about assessing novelty, is it not?

You take out copper from metal, and suddenly what's left of metal becomes new over copper.
But take out copper from metal, and what's left of metal becomes new over (the genus) metal as well.

You introduce a limitation that confers novelty. That is fine if the limitation was originally disclosed, but problematic if it was not.

The solution is not to introduce limitations, not even notionally, that have no basis in the application as filed.

(And again, note that both G 2/98 and the memorandum start with OR-claims and only then say that the alternative parts may have different priority dates.)

Proof of the pudding said...

Who needs to take copper out of metal? Why, for the purposes of assessing priority, is it not possible to view "metal" as "the same invention" as "copper" to the extent that the genus and the species overlap? Is that not exactly the kind of assessment that is made with respect to novelty?

Proof of the pudding said...

Who needs to take copper out of metal? Why, for the purposes of assessing priority, is it not possible to view "metal" as being "the same invention" as "copper" to the extent that the genus and the species overlap? Is that not exactly the kind of assessment that is made with respect to novelty?

The Cat that Walks by Himself said...

Monday, 24 August 2015 at 20:06:00 BST

It's a very good comment.

Anonymous said...

@Proof of the pudding:
Who needs to take copper out of metal?

Both G 2/98 and the memorandum state that multiple and partial priorities are about OR-claims. So if you want to have multiple or partial priorities for a claim, viewing it is as an OR-claim is the first step to take.

The idea of an OR-claim is that it has a structure A OR B and that it is new (and inventive) exactly if both A and B are new (and inventive).

The proponents of partial priorities as a cure against poisonous divisionals argue that to form an OR-claim, it is sufficient that the claim encompasses A and B (and that together they cover the whole claim). They do not require that A and B are individually disclosed in the application.

The proposed cure works as follows. If the application and its divisional disclose copper with priority and metal without priority (so that copper arguably is novelty destroying for copper under 54(3)), then a claim to metal is viewed as "copper" OR "metal but not copper". Both alternative parts of "metal" are new over "copper", so in this way novelty of the claim to metal is established.

But this trick can be used to re-establish novelty of anything. If a claim to "metal" is confronted with 54(3) prior art disclosing metal, just split "metal" in "copper" OR "metal but not copper". Both parts are new over metal, so metal is new over metal.

Or if there is regular 54(3) prior art disclosing copper, just conceptually split "metal" in "metal having property X" OR "metal not having property X" where X is a property that is not mentioned in the 54(3) prior art. Now both parts are new over copper, so the claim to metal is new over copper. No need anymore to apply G 1/03 in order to disclaim copper.

So: flexible OR-claims are an essential ingredient of the "partial priority" cure to poisonous divisionals, but these flexible OR-claims lead to chaos.

Why, for the purposes of assessing priority, is it not possible to view "metal" as "the same invention" as "copper" to the extent that the genus and the species overlap? Is that not exactly the kind of assessment that is made with respect to novelty?

What exactly are you proposing, how would it work? Would you still have OR-claims? If so, what are the rules for forming an OR-claim? If you avoid OR-claims but still fiddle with partial priorities, how do you reconcile that with the memorandum and with G 2/98?

Anonymous said...

"But this trick can be used to re-establish novelty of anything. If a claim to "metal" is confronted with 54(3) prior art disclosing metal, just split "metal" in "copper" OR "metal but not copper". Both parts are new over metal, so metal is new over metal."

No, because priority and novelty are different even if we apply identical rules for interpretation of the claims. In order to determine whether a claim is entitled to priority, application of the partial priority rule you recite works to avoid the poisonous priority issue. I think you agree with that.

If we forget about the priority aspect (let's assume priority doesn't exist in law) then in the above case the earlier application disclosing metal would, under current rules, destroy novelty of the later claim to metal. I agree that, if the claim could be interpreted in the way suggested as an 'OR ' claim, there would be an issue. However, there is no legal basis for considering the later claim to metal as a two-part OR claim. So, it simply is not an option.

Meanwhile, the ability to claim priority does exist, and it is this legal provision which provides the basis for the 'OR' claim interpretation.

MaxDrei said...

As to the so-called "trick" I do not see it as a "trick". Tricks are when an Applicant amends after filing by inserting new matter, illegitimately to improve its position relative to its filing date. Here, the so-called "trick" is just to restore primacy of the date on which you file your "application for protection" at the Patent Office. All that the "trick" is doing is to restore that fundamental undisputed primacy of the overall notion of "Paris Convention priority". If you do add to your priority date matter M, some additional matter M+ during the Paris year, you get your priority date for M but only your filing date for M+. That much is not in dispute. There just has to be a complementary way of managing the validity of the set of claims that has been founded on that discloisure.

Lancing the poison in priorities and divisionals requires no shift in the way Art 54(3) references are dealt with under novelty law.

So I agree with that last anon, that:

".....priority and novelty are different even if we apply identical rules for interpretation of the claims."

Tim Jackson said...

"The idea of an OR-claim is that it has a structure A OR B..."

OK

"... and that it is new (and inventive) exactly if both A and B are new (and inventive)."

No, neither G2/98 nor the FICPI memo are suggesting that. They are talking about the priority of A and B, not the novelty of A and B. But here's how the priority of A and B affect the novelty of the claim as a whole.

Priority stems from the subject matter disclosed in the priority document. That's why you look to see whether and to what extent the claim of the subsequent application relates to that subject matter. Partial priority occurs when some notional part A of the claim relates to the same subject matter (using the G2/98 test of "same invention"), but another notional part B does not.

The novelty test just compares the whole claim to the prior art. However, in doing so, if the prior art has an intervening date, the priority can be used as a shield. This follows from Art. 89 EPC - the effect of the priority is that the relevant subject matter has an earlier date of filing for the purposes of Arts. 54(2) and (3). But it only shields for the subject matter actually disclosed in the priority document.

So if the whole claim says metal, its novelty can be destroyed by an intervening prior art disclosure of metal. Partial priority just for copper is not good enough to shield that.

If the whole claim says metal, its novelty could also be destroyed by an intervening prior art disclosure of copper. But this time, the partial priority for copper is an effective shield.

Nowhere does this analysis look at the novelty of B alone.

Proof of the pudding said...

It looks like my earlier comment did not get posted, so I shall try again.

Anon - The point that I was making is that there appear to be different assessments under novelty and priority when it comes to deciding how an (expanded) genus relates to a more specific disclosure that is encompassed by the genus.

To my mind it is THAT discrepancy which leads to the problems of self-collision, etc.

If you think about it from first principles, things really ought to be much simpler. That is, assessments of both novelty and priority ought to ask the same question:
To the extent that the claimed genus overlaps with the earlier disclosure, does it relate to "the same invention"?.

If the answer to this question is "yes", then your claim does not have novelty. However, if the earlier disclosure is in your priority document, you should have entitlement to priority for the invention disclosed in that document.

On the other hand, if the answer is "no", then there is novelty but no entitlement to priority.

What the whole concept of self-collision is based upon is the idea that a claim can be effectively "novel" for the purposes of assessing priority but yet still lacking in novelty under Article 54(3). Whilst most of us clearly struggle with the absurdity of this conclusion, it appears that few are prepared to contemplate that the FICPI memorandum and/or G2/98 might not have got things quite right when it comes to assessing partial priority.

Personally, I am not convinced that an "OR" claim is an essential prerequisite for assigning partial priorities. On the other hand, I can understand why there is confusion surrounding this point. This is because, if you add something extra to an "AND" claim, you are almost always narrowing the scope - which makes it much trickier to assess whether the resulting (more precisely defined) invention is "the same" as the broader disclosure in the priority document. However, just because things can sometimes get a bit tricky does not mean that the EPO should abandon common sense and end up with a situation where novelty and priority are assessed according to different standards.

MaxDrei said...

Pudding, you are confusing me, when you urge the EPO to hold to the same standard when assessing priority as when assessing novelty.

How can that be, when Paris is all about "same invention" but Art 54(3) and 123(2) are not. We did away with "prior claiming" back in 1973. Since then, we have had "whole contents" for Art 54(3) and 123(2).

Thus, for Paris Convention priority the Gold Standard is still "same invention" yet for novelty and new matter the concept of "same invention" has no relevance.

Can you say more?

Anonymous said...

Have a read of T1890/09. That sets out the differences between priority and novelty.

Proof of the pudding said...

Max - Article 54(1) EPC states that "An invention shall be considered to be new if it does not form part of the state of the art".

Thus, to take away novelty, the state of the art needs to disclose the claimed invention (i.e. the same invention as defined in the claims). This is all that I mean when I say that the assessment of novelty involves asking the question of whether the earlier disclosure relates to "the same invention".

I was not saying anything about Article 123(2), the wording of which clearly relates to "subject-matter" (and not to an "invention").

Nevertheless, I think that having equivalent (or corresponding) standards for assessing both novelty and priority is pretty much essential - as otherwise you end up with a mess like self-collision.

Tim Jackson said...

T1890/09 is in German and I've only scanned it quickly. However, as far as I can see it doesn't consider partial priority at all. So I don't think it is very useful in the present discussion of how partial priority affects (or should affect) novelty.

The decision just held that the claim didn't have full priority, because the priority document omitted a feature of the claim. The priority document then destroyed its novelty under Art. 54(3).

In my view, partial priority should have been considered. Whether it would have led to a different result would have depended which interpretation of "a limited number of clearly defined subject-matters" the Board then chose to follow. Or like Q1 and Q2 in the present referral, they too could have asked the Enlarged Board whether "a limited number of clearly defined subject-matters" is really the correct test.

From point 3.1 of the Reasons of T1890/09, however, it appears that the opponent only sprang the poisonous priority issue shortly before the oral proceedings in the appeal. It looks as if the proprietor was unprepared. He just argued for full priority, not partial priority. Then when that failed he argued that the poisonous priority attack was inadmissible because it was too late. That failed too.

MaxDrei said...

Pudding points out at 15:27 today that Art 54(1) provides that "an invention" shall be new if "it" does not form part of the art.

If I read him right, he is suggesting that Art 54(1) looks whether the "balloon" of claimed subject matter is "part of the state of the art". Fair enough, one would think that, wouldn't one, from the literal wording of Art 54(1) EPC? Except that, as we all know, you don't need the balloon in the art, just a novelty-destroying "needle" that will puncture the balloon, the claim. The degree of generality with which the needle is described in the document asserted against the novelty of the claim is irrelevant to the issue whether it does destroy novelty, whether it does puncture the balloon. In fact, when enablement is in issue, the more specific and detailed the prior art disclosure, the more penetrative will be the novelty attack.

So I still do not accept his point, that priority (same invention) and novelty under Art 54 EPC are the same thing legally. The EPO's "Gold Standard" has to do with a fact, what any given document discloses to its skilled reader. We need the same Gold Standard for the factual underpinning of all olegal issues arising under all of Art 54, 87 and 123(2) EPC, it goes without saying. But the questions of law, when to accord priority, when to allow a prosecution amendment, and when to find a lack of novelty, are not the same as each other.

That, anyway, is how I see it at the moment.

Anonymous said...

@Anonymous:
No, because priority and novelty are different even if we apply identical rules for interpretation of the claims. In order to determine whether a claim is entitled to priority, application of the partial priority rule you recite works to avoid the poisonous priority issue. I think you agree with that.

I'm only applying identical rules.

The starting point for the partial priority cure to poisonous divisionals is forming the OR-claim. Split the claim into two new parts (notionally or however you want it), and now the claim is new. That is how OR-claims work and that is what makes the cure work.

If this trick works, then why can I not split my claim to metal to establish novelty over metal? I don't even need priorities for this; an OR-claim with two parts, each narrower than metal, will do.

(Suppose the application actually discloses metal, copper and "metal but not copper". In that case a claim to "metal" can be rewritten as two separate claims to "copper" and "metal but not copper". Now each of these two claims really is new over metal. This might at first sight be somewhat surprising and even disturbing, but this really is how it is. One can still discuss whether it is allowable to keep the original wording "metal" or whether the applicant should switch to two separate claims, but for his final scope of protection that does not matter. I'd say he can keep "metal". As one can see from this example, being able to split for the purpose of novelty can be very useful and hence really should have a basis in the application as filed.)

If we forget about the priority aspect (let's assume priority doesn't exist in law) then in the above case the earlier application disclosing metal would, under current rules, destroy novelty of the later claim to metal. I agree that, if the claim could be interpreted in the way suggested as an 'OR ' claim, there would be an issue. However, there is no legal basis for considering the later claim to metal as a two-part OR claim. So, it simply is not an option.

I suppose you agree that if the application individually discloses the two parts A and B, there is nothing to stop the applicant from rewriting his claim into an OR-claim A OR B having identical scope to his original claim. I suppose you also agree that if both A and B are new, then A OR B is new. (If you do not agree, then consider two claims: 1. A; 2. B. Both are new.)

If you agree with me that OR-claims require both parts to be individually disclosed in the application, then you should also agree that partial priorities are no cure for poisonous divisionals (so the cure will just have to be found somewhere else! I'm not arguing in favour of poisonous divisionals).

On the other hand, if you are still of the view that partial priorities form a cure to poisonous divisionals, then I would like to know how that should work exactly. Do we agree (with the memorandum, G 2/98, T 1222/11 and T 577/13) that the first step is viewing the claim as an OR-claim? What precisely are the conditions for a claim to be (equivalent to) an OR-claim?

Those conditions should somehow be flexible enough to cure poisonous divisionals, but rigid enough to exclude ridiculous results. I'm pretty sure such conditions do not exist. (And of course they won't have a legal basis.)

Anonymous said...

@Tim Jackson:
"... and that it is new (and inventive) exactly if both A and B are new (and inventive)."

No, neither G2/98 nor the FICPI memo are suggesting that.


Are you really suggesting that a claim to "A OR B" is not new precisely when both A and B are new?

The memorandum is rather clear about an OR-claim with alternatives A and B being equivalent to two separate claims:
1. A
2. B

It is also widely accepted that a claim to "iron or copper" is new precisely if iron is new and copper is new.

Note that being able to assess novelty and inventive step separately for the alternative parts of an OR-claim is why it makes sense to assign different priority dates to those parts.

It seems to be your position that it is the other way around: because different priority dates are assigned, the alternatives are to be examined separately for novelty and inventive step.

Maybe you even only want them to be examined separately against documents that are prior art against one part and not against the other part. However, that gives a messy situation in case of a potential inventive step attack based on documents D1 and D2, one document being prior art for both parts and the other being prior art only for one part.

In any event, consider the following example:
P1 discloses copper.
The application discloses copper and metal and claims metal.
D1 is an independent EP filed between P1 and the application and discloses metal.

What are now the (notional) parts of the OR-claim? Copper and ...?
If the other part is "metal but not copper", then your approach leads to the claimed "metal" being new over "metal" in D1. An unacceptable outcome in my book.
If the other part is "metal", then you haven't fixed poisonous priorities: metal is not new over "copper" in a divisional of the application.

You will probably say that the other part is "metal" and that it is shielded against "copper" in the divisional by its "partial priority" for the part "copper".

My first observation is that you are not following the OR-claim approach of the memorandum, G 2/98, T 1222/11. In fact, you have no need for the "copper" alternative at all.

My second observation is that your approach might be unproblematic in case of 54(3) prior art that is only used for novelty, but creates chaos for 54(2) prior art and inventive step. If D2 discloses copper and is published between the filing dates of P1 and the application, how are you going to examine inventive step of "metal"?
If metal is really protected against copper, then D2 is not citable against the claim. That seems to be an unacceptable outcome (and is in conflict with G 3/93 (*)).
If D2 is citable against the claim, then how is metal protected against disclosures of copper? Do we have to examine "metal but not copper" for inventive step over D2? Then you are effectively looking at the alternative "metal but not copper" after all. We should then also examine "metal but not copper" for novelty against D1, leading to the unacceptable conclusion that that part is novel and the claim to "metal" is new over "metal" in D1.

(*) It follows from G 3/93 that whether a document is prior art does not depend on the subject-matter disclosed in that document.

Anonymous said...

@Proof of the pudding:
To my mind it is THAT discrepancy which leads to the problems of self-collision, etc.

Note that what I care about is that the "partial priority" approach to solving the problems of self-collision and poisonous divisionals either leads to contradictions or to ridiculous outcomes in carefully constructed examples. I am not worried about the self-collision and poisonous divisional problems themselves, as they can be solved in other ways without creating a mess.

Give me a detailed and complete (non-handwaving) recipe for applying partial priorities, and I will give an example where that recipe fails miserably in the sense that it gives an outcome that you will call ridiculous (or I will show that it does not solve poisonous divisionals).

That is, assessments of both novelty and priority ought to ask the same question:
To the extent that the claimed genus overlaps with the earlier disclosure, does it relate to "the same invention"?


Novelty only asks whether the claim encompasses subject-matter in an earlier disclosure.

For priority, see G 2/98 and G 3/93. First you determine the subject-matter that you are going to examine for novelty and inventive step (might be the whole claim, might be (each) one of several alternatives). Then you determine its priority date (essentially by determining the earliest priority document that discloses that same subject-matter, G 2/98). When you have its priority date, you know which publications are citable against it (no need to look at the content of the publications here, only their date is important!, G 3/93). Then you do the examination of novelty (for both 54(2) and 54(3) documents) and of inventive step (only for 54(2) documents).

I don't understand what you are suggesting (I am sorry, but for me it is not a recipe that I could apply to a concrete case), but I don't think it complies with the scheme I have just outlined and which is firmly based on G 3/93 and G 2/98.

it appears that few are prepared to contemplate that the FICPI memorandum and/or G2/98 might not have got things quite right when it comes to assessing partial priority.

This shows that I am having some success in pointing out the deficiencies of the partial priority "cure". That cure has its origin in an interpretation of G 2/98 according to which the literal examples of the memorandum are to be taken as gospel, and now you are saying that the memorandum might not have gotten things quite right.

Before you start working out a whole new (and preferably consistent!) theory of disclosure and priority, let me ask a question: why? It would be trivial for the EBA to wipe out the problem of poisonous divisionals by stating that members of a family of divisionals cannot be cited against each other. It would also be easy to solve the same problem for simultaneously filed unrelated (but essentially identical) applications, if the EBA deems that desirable. No need for a whole new theory just to resolve those issues.

Proof of the pudding said...

@Anon 29 Aug 2015 01:26
"Novelty only asks whether the claim encompasses subject-matter in an earlier disclosure".

And what is the problem with applying the same standard to partial priority?

Max - I hope that this explains to you the concept that I have proposed does not change the assessment of novelty (i.e. it would still be destroyed by the "needle" as well as by the "balloon").

MaxDrei said...

Anonymous writes:

"It would be trivial for the EBA to wipe out the problem of poisonous divisionals by stating that members of a family of divisionals cannot be cited against each other."

but I don't think it is "trivial" so to "state". To my mind, you can't just "state" what you want the law to be, by fiat. Rather, you have to get to it, from the statute, by reasoning. That's the non-trivial part.

Pudding, you ask "what is the problem". I'm surprised that you ask. I had supposed the problem (nonsensical self-collision) to be as plain as a pikestaff. To repeat though, I'm here as a dilettante, an interested bystander. I will never carry the burden of writing the text of a decision coming out of DG3. What I care about is that DG3 caselaw gets ever more respect from all over the world, rather than increasing frustration, incomprehension, impatience and ridicule.

Come on EBA. The Supreme Courts of the Member States, and in the wider world, have an excuse for blather, muzzy thinking and obfuscation, that they are not specialist patents courts. But the EBA is. It did alright with G3/14 on Art 84 clarity, with its excellent sentence about being pragmatic rather than overly theoretical. Now let's have another sensible working roadmap please, if necessary dumping ideas expressed in earlier G Decisions, thereby to re-set the compass, going forward, and purge the caselaw body of all this self-collision nonsense.

The Cat that Walks by Himself said...

For me in this subject there's only one question left.
Namely, whether it's desirable to allow a lower patentability threshold of art. 54 (3), i.e. novelty only, for patentee's own further/divisional applications.

I have an impression that it's a way to extension of patentability period and patenting on novelty only. Can it lead to a chain of "generalisation applications" ?

Anonymous said...

"If this trick works, then why can I not split my claim to metal to establish novelty over metal?"

Absolutely you can. Provided you do so when drafting your application. Either that, or get all nations back to Paris to agree a new convention on entitlement to novelty.

MaxDrei said...

Can CWH say more clearly please what precisely is the "only one question left"? For self-collision in Europe, we are NOT talking about everything filed by the same Applicant or Inventor. Rather, we are discussing the extent to which different members of one single patent family, all declaring the same priority date(s), can "collide" with each other under Art 54(3) EPC.

Specifically, I cannot gather what CWH means by:

"...a lower patentability threshold of art. 54 (3), i.e. novelty only, for patentee's own further/divisional applications."

Perhaps he means that my priority document, which under Art 54(3) EPC is effective to destroy the novelty of your later-filed application for the same subject matter, shall be deemed not to destroy the novelty of any claim of my own patent application. The thing is though, my application, unlike yours, does contain a declaration of my own priority date. That declaration is the "shield" that Tim Jackson so helpfully mentions. I've got one, and you haven't. That's all there is to it.

When the USA switched away from First to Invent to First Inventor to File, there were a large number of "Chicken Little" commentators running round telling us all that "The Sky is Falling". Well, it hasn't fallen yet. Likewise, here, there seem to be commentators also telling us that if the EBA sweeps away the nonsense of self-collision between family members the sky will fall down. I still don't see it.

Anonymous said...

@Anonymous:
"If this trick works, then why can I not split my claim to metal to establish novelty over metal?"

Absolutely you can. Provided you do so when drafting your application. Either that, or get all nations back to Paris to agree a new convention on entitlement to novelty.


We agree on this. So the trick only works if the application as filed discloses the alternative subject-matters into which the claim is to be split. That means the trick is of limited use for working around poisonous divisionals. (Which is not a big deal, as the EBA can choose to address Art. 54(3) directly.)


@MaxDrei:
"It would be trivial for the EBA to wipe out the problem of poisonous divisionals by stating that members of a family of divisionals cannot be cited against each other."

but I don't think it is "trivial" so to "state". To my mind, you can't just "state" what you want the law to be, by fiat. Rather, you have to get to it, from the statute, by reasoning. That's the non-trivial part.


In principle I agree, but in practice the EBA has considerable freedom. G 5/83 and G 1/03 are two examples where the EBA created a rule that definitely was not in the statute. Of course the EBA still needs to give some justification for any new rule.

Now let's have another sensible working roadmap please, if necessary dumping ideas expressed in earlier G Decisions, thereby to re-set the compass, going forward, and purge the caselaw body of all this self-collision nonsense.

I don't see this as a problem that the case law has created, but as a problem that is in the EPC and that the case law has not yet solved. It is caused by the interaction of Art. 54(3) with Art. 89. But we'll see what the EBA will do.

Anonymous said...

Max, possibly it will all seem simpler when (if) the EBA simply resolve the issue. Too may attorneys find this an over complicated issue to deal with for my liking. I'm concerned.

Anonymous said...

@Proof of the pudding:
"Novelty only asks whether the claim encompasses subject-matter in an earlier disclosure".

And what is the problem with applying the same standard to partial priority?


The main problem is that "applying the same standard to partial priority" does not have a clear meaning. The EPC only refers to "priority". Art. 88(2) and (3) only state that certain parts of an application may have one priority date and other parts another priority date or none at all. The term partial priority is commonly used to refer to the latter case: one part has regular priority, another part has no priority at all. There is no concept in the EPC of particular subject-matter having "partial priority".

One could say that a claim to metal may have "partial priority" to copper which somehow "shields" the claim against certain intermediate documents (in particular documents that disclose copper) but not against others (such as documents that disclose metal). But that is simply too vague a notion: we need to pin down the precise details of what this means.

If it means that a document with a date between P1 and the application does not count as prior art if it discloses copper, but does count as prior art if it discloses metal, then that is in direct conflict with G 3/93. That decision explains that whether a disclosure is prior art only depends on the date of the disclosure and not on its contents. (Note that the date may depend on the contents in case of 54(3) prior art that claims a priority, but that is a different matter.)

If, on the other hand, it means that we do not examine "metal" for novelty (and inventive step) against intermediate documents, but "metal but not copper", then we get the unacceptable result that the claim to "metal" is new over "metal" (without there being any basis in the application as filed for actually splitting the claim to metal into two or more parts that are new over metal).

Such conflicts, inconsistencies and contradictions remain hidden as long as we keep things vague: a "partial priority" that "shields". Keeping things vague simply does not help. Instead, we need to consider in detail what rules and tests to apply. The full recipe has to be spelled out.

MaxDrei said...

The anonymous commenter of earlier today is worried that the "shield" idea is "keeping things vague" and so will not help.

I don't follow.

Early in his post he envisages the partial priority shield as one that protects against "certain intermediate documents" that are asserted as Art 54(3) novelty attacks. That is not how I envisage the "shield" at all. For me there is no shield other than the imaginary one that wards off Art 54(3) novelty attacks that come from your own priority document. For all other Art 54(3) attacks, it is "as you were". No change in the law of Art 54(3) EPC whatsoever. I fail to see how the shield idea keeps things vague. It seems to me to be, on the contrary, a crystal clear way to eliminate the burgeoning vagueness, whether or not your very own priority document can be a fatal novelty attack.

I wonder, Tim Jackson, do you see the shield different from me?

Anonymous said...

@MaxDrei:
The anonymous commenter of earlier today is worried that the "shield" idea is "keeping things vague" and so will not help.

I don't follow.


As long as one spells out what it really means, "shielding" is not vague.

If one does not spell it out, it tends to mean different things depending on the criticism it is exposed to.

For me there is no shield other than the imaginary one that wards off Art 54(3) novelty attacks that come from your own priority document. For all other Art 54(3) attacks, it is "as you were". No change in the law of Art 54(3) EPC whatsoever.

That is a good solution and similar to what I have proposed (novelty attacks cannot be based on 54(3) prior art that is disclosed with the same date in the EP itself). Your solution (novelty attacks cannot be based on 54(3) prior art that is disclosed with the same date in a priority document) might be preferable. (With mine, a divisional that claims the genus and does not disclose the species would still be killed by the parent that discloses the species.)

In my view your solution does change the law of Art. 54(3), as it removes certain disclosures from the prior art under Art. 54(3).

I do not consider your solution to be based on "partial priorities". You're not assigning different dates to different parts.

MaxDrei said...

Thanks for that. Pointing out the difference between the Tufty Approach and the one you advocate is important.

There is no mention in the EPC or the Imp Regs of anything called a "shield". If the EBA is now going to create one, to stop the nonsense of self-collision under Art 54(3) EPC between patent family members that all declare the same filing and priority dates, it has to get there from the EPC. One way is to perform a "notional" splitting of the claim, into a part that takes priority from the prio doc and another part that doesn't. Another way is to excise patent family members from the 54(3) prior art universe.

Given that there is nothing new in the concept of assigning different dates to different "parts" of a single claim, I prefer to purge the poison using Tufty's notional claim splitting rather than an EBA re-write of Art 54(3) EPC.

I'm glad to think I'm not assigning different dates to different parts of the claim. It seems to me that I am, but I don't really mind which of us is right on that because, in the case of colliding family members, it is all notional anyway.

Of course, in the case of colliding cases that are not members of the same family, the claim splitting is real and has somehow to be accomplished. Often tricky, frequently not possible. But, free of any requirement to find words available and effective, a splitting that is no more than notional is always possible.

Anonymous said...

@MaxDrei:
There is no mention in the EPC or the Imp Regs of anything called a "shield". If the EBA is now going to create one, to stop the nonsense of self-collision under Art 54(3) EPC between patent family members that all declare the same filing and priority dates, it has to get there from the EPC. One way is to perform a "notional" splitting of the claim, into a part that takes priority from the prio doc and another part that doesn't. Another way is to excise patent family members from the 54(3) prior art universe.

But if the EBA reasons via a partial priority approach to eliminate poisonous divisionals, then that same approach will also be applicable in other situations, and in particular it will lead to "metal" being found new over "metal" as in my example.

I don't see how the EBA could ever argue that the partial priority approach may be applied where it gives a result that is "desirable" (no poisonous divisionals) and not where it gives a result that is ridiculous.

In my opinion, the clean solution is to excise exactly those disclosures from the 54(3) prior art universe that hurt. Those are the disclosures of subject-matter having a particular date that is disclosed in a priority document with the same date (or, alternatively, that is disclosed in the application as filed with the same date). This also takes care of simultaneously filed identical applications that are not family members (because they disclose subject-matter that has the same date as identical subject-matter in the priority document or application as filed).

As justification for creating the rule the EBA could argue that Art. 54(3) clearly was not intended to let the priority document and/or the application itself be detrimental for novelty of the application.

This solution involves an element of law making, but the EBA does not do anything else in G 1/03.

Anonymous said...

@MaxDrei (still same anonymous here):
I'm glad to think I'm not assigning different dates to different parts of the claim. It seems to me that I am, ...

Different priority dates for different parts would mean that certain disclosures would be prior art for one part (because of their date), but not for the other (again because of their date); that is how priorities work (G 3/93). This is not the same as excising certain disclosures from the 54(3) prior art for the whole claim.

In my view, different priority dates for different parts also necessarily implies that the two parts are examined for novelty and inventive step independently (the whole claim being new and inventive precisely if each part is new and inventive). This is because it makes no sense to examine inventive step of the claims as a whole in view of, for example, a combination of documents D1 and D2 where D1 is prior art for some parts of the claim and D2 is prior art for other parts of the claim. So the "notional" splitting is in fact a very real splitting when it concerns the examination of novelty and inventive step (and should therefore have a 123(2)-basis in the application as filed).

MaxDrei said...

Thanks for those last two mails. I regret my haste yesterday in posting before reading very carefully what the auto-correct had written for me. I intended "I'm glad you think....." but what I posted was "I'm glad to think....". Most unfortunate and confusing. Sorry about that.

The way I see it, self-collision is a 54(3) issue, so we are spared wrestling with dire Art 56 consequences if the EBA takes up the "shield" idea.

You envisage zealous attorneys using the shield in non-family situations. But I imagine an EBA opinion which creates a narrow shield defence against self-collision with wording that effectively shuts off its spread to situations where the sword is a non-family Art 54(3) document.

There seems to be little between us. Whenever a TBA decides against me, I leave the room unhappy, and keen to read in the written decision where the Board went wrong. When I get the decision in writing, however, I sometimes have to concede that the decision is not bad. So, here, I look foirward to an EBA opinion which I can agree is not bad. We both just want an elegant solution that works, I think. One that increases legal certainty and doesn't harm it.

The Cat that Walks by Himself said...

Re MaxDrei: Perhaps he means that my priority document, which under Art 54(3) EPC is effective to destroy the novelty of your later-filed application for the same subject matter, shall be deemed not to destroy the novelty of any claim of my own patent application. The thing is though, my application, unlike yours, does contain a declaration of my own priority date. That declaration is the "shield" that Tim Jackson so helpfully mentions. I've got one, and you haven't. That's all there is to it.


Indeed, I had in the mind a similar situation. Let's say:

Applicant A: copper at P1. Let's say Applicant A does not have an enabling disclosure for metal as genus at P1.
Applicant B: metal, in particular copper at P2.
Applicant A: metal, in particular copper at P3.

P3 later than P2 later than P1.

If I would follow your "shield" reasoning, I would have to give a patent for 'metal' to Applicant A, who is not a first inventor to file 'metal'.

It is like declaring as a winner in the 4×100m relay the team that has won first 100m...




MaxDrei said...

Cat, I am grateful for your hypothetical but do not agree with your conclusion. It seems that we have different ideas what constitutes the shield. What I say is: no divisional, no shield.

This is late in the evening, and written in haste, but here goes.

A was first to file (on date P1) with an enabling disclosure of copper. He is entitled to a claim to copper.

B's Art 54(3) enabling first disclosure of metal was filed (P2) before A's (later, on date P3). It therefore destroys the novelty of A's claim to metal.

For me then, B can have his claim to metal, provided he disclaims copper.

Now imagine that there is no B. Instead, A files a divisional, with claims that are not entitled to date P1 but with a scope that embraces copper. Absent the "shield" proposed in these columns, the P1 enabling disclosure of copper destroys the novelty of any such claim. I think that's wrong. I think that the validity of the claim should be judged as of its date P3 and if there is any intermediate filing by B, such as the one you suggest, that first enabling filing on metal, by B, is definitely novelty-destroying for A. One invokes the "shield" only against the fatal "auto-immune" disease brought on by the A-publication of your very own priority document.

I do not see your reasoning. Will you give us more?

Anonymous said...

If the applicant were to withdraw the priority doc before publication it would not be 54(3) prior art as was standard practice in the UK.
A rebel

Proof of the pudding said...

It seems that there is a concern that approaching novelty under Article 54(3) using the "Tufty" approach to partial priorities (notional separation of the claim into two alternatives). For example, as others have pointed out, "metal" (as notionally separated into "copper" and "metal, but not copper") might then be deemed novel over an earlier disclosure of "metal".

However, for at least two reasons, such a concern should not rule out a common sense approach to partial priorities. Due to length restrictions, I will deal with these two reasons in separate comments.

The first reason is that the salami-slicing approach to achieving novelty over a "generic" disclosure can produce equally absurd results when the separation into alternatives is real (as opposed to just notional).

For example, consider the following situation in which A1 is Article 54(3) prior art against A2:

A1 discloses "metal"
A2 discloses both "copper" and "metal, but not copper".

Can A2 legitimately claim both of the alternatives that it discloses? I presume that many of those commentators who have raised concerns about "notional" splitting of claims would think so. If so, is the root cause of the concern expressed by those commentators really to do with partial priorities? I think not.

As I have previously stated, novelty of selection is another beast entirely from partial priorities - and, in my view, is a problem that also needs to be tackled by the Enlarged Board.

But I digress. The most important point will follow in my next comment, which explains how I believe that novelty can be assessed in a situation of partial priority.

Proof of the pudding said...

The concept that I am envisaging for assessing novelty in a situation of partial priority can perhaps be explained by an analogy involving battleships.

Let's imagine that:
- the disclosures of each document together represent a battleship that is plotted on a two-dimensional graph;
- all battleships are aligned parallel to the x-axis and are spaced apart (on the y-axis) in date order;
- each battleship has a length and position (on the x-axis) that is determined by the (breadth of) the subject matter disclosed;
- all battleships have an array of guns spread out along their entire length, which guns can fire only in lines parallel to the y-axis and then only in one direction (i.e. towards ships that are "later" on the y-axis);
- all ships fire their guns in strict date order;
- guns fired from ships representing priority documents only destroy guns on an "enemy" ship;
- for the purpose of assessing novelty, any gun fired from an Article 54(2) or (3) document that hits an "enemy" ship will sink that ship.

Now, with this analogy in mind, what happens with the situation where:
- an application claims "metal" but has priority only for copper; and
- there is an intervening disclosure of "metal"?

The answer is that the priority ship (which fires first) knocks out the guns of the intervening disclosure, but only along the portion of that ship that relates to "copper".

The intervening disclosure ship then has guns left that it can fire, relating to the subject matter that has not been knocked out (i.e. the portion of "metal" that does not relate to "copper"). When it fires its guns, it sinks the application ship (though not a life raft on the ship that represents a fall back claim to "copper").

Thus, in this way, it is possible to view the claim to "metal" as an uninterrupted whole whilst still assigning multiple priorities to different embodiments covered by the claim.

The analogy can be pushed further too. When it comes to assessing inventive step, one could imagine a fresh set of guns being able to fire at various angles relative to the y-axis (depending upon the angles that those on the intervening disclosure ship would think of setting). Thus, an intervening disclosure of "copper" might not sink a "metal" application ship when firing for novelty, but could when firing for inventive step (if the intervening disclosure is Article 54(2) prior art for "metal").

Proof of the pudding said...

The approach to partial priority that I have outlined (i.e. assigning priority to some, but not all, embodiments covered by a claim) appears to me to be completely consistent with the Paris Convention. For example, as discussed in G2/98:
"The possibility of claiming multiple priorities was introduced into the Paris Convention in order to avoid improvements of the original invention having to be prosecuted in applications for patents of addition. This makes it clear that ”element” was not understood as a feature but as an embodiment"

The trouble that I have with G2/98 is that the phrase "a limited number of clearly defined alternative subject-matters" appears to have no basis in the EPC, the Paris Convention or the legislative intent behind either.

I have previously argued that the intent behind the Paris Convention rules out refusal solely upon the grounds of self-collision (between divisionals). Why, therefore did the Enlarged Board chose language that leads to that absurd result? And what does "clearly defined" mean? Defined where? In the application or the priority document? Or is it OK if the alternatives only emerge by comparison of the two?

Would we not be better off if the Enlarged Board had instead indicated that partial priority can be awarded for disclosures of a priority document that represent embodiments of the later claim?

This would seem to deal with the tricky issue behind G2/98 (i.e. because it would not represent an embodiment of a broader "AND" claim in the priority document, a more narrowly-defined "AND" claim would not be entitled to the priority date). It would also seem to avoid the problem of self-collision (e.g. because the later, narrow claim could be afforded priority for those embodiments from the priority document that still represent embodiments of the invention as more narrowly defined).

I can only hope that the Enlarged Board is brave enough to step out of this particular corner that they have painted themselves into.

Anonymous said...

@MaxDrei:
We both just want an elegant solution that works, I think. One that increases legal certainty and doesn't harm it.

We seem to agree on the end result. To get there, I think the EBA needs question 5.

I see two ways:
- declare that certain (poisonous) disclosures don't count as prior art under 54(3), or
- reason that claim splitting would be allowable via a G 1/03-disclaimer (this is problematic in view of point 4.7 of G 2/10, but that could be overruled). If the claim can be split to get two valid claims that precisely cover the original claim, there should be no need to actually perform the split.

@The Cat that Walks by Himself:
If I would follow your "shield" reasoning, I would have to give a patent for 'metal' to Applicant A, who is not a first inventor to file 'metal'.

In my view the "shield" should only remove P1 (copper) from the 54(3) prior art, because the application (or, alternatively, the priority document) discloses copper with the same date as P1. (So I agree with MaxDrei.)

@Proof of the pudding:
The first reason is that the salami-slicing approach to achieving novelty over a "generic" disclosure can produce equally absurd results when the separation into alternatives is real (as opposed to just notional).

But these are not absurd results. If the split is real (i.e. has a basis in the application as filed), the applicant can actually formulate separate claims. Separate claims indisputably have to be examined separately for novelty and inventive step. So if both claims are new and inventive, the applicant can have both. What is happening here is that the applicant benefits from a limitation that is disclosed in the application as filed, which is perfectly fine. What would be absurd is if the applicant could have the same benefit from a notional limitation not having any basis in the application as filed.

If your proposal somehow "fixes" the "absurd" results of real splits, then it will likely make it impossible to render a claim novel by introducing a further limitation. Do you realise that?

I have previously argued that the intent behind the Paris Convention rules out refusal solely upon the grounds of self-collision (between divisionals).

Not a chance! The drafters of the PC had no knowledge of Art. 54(3).
The best your argument will do is explain why Art. 54(3) should be interpreted as ruling out self-collision.

Why, therefore did the Enlarged Board chose language that leads to that absurd result?
Self-collision is caused by an unforeseen interaction between Art. 54(3) and 89 EPC, not by G 2/98. That a particular approach to priority (or to Art. 123(2)) possibly might have masked it is a different matter.

Would we not be better off if the Enlarged Board had instead indicated that partial priority can be awarded for disclosures of a priority document that represent embodiments of the later claim?

I assume that by "embodiment" you mean some kind of atomic particle covered by a claim. In chemistry such claims and corresponding embodiments might exist, but in general they do not. More importantly, prior art disclosures often disclose "clouds" of embodiments instead of atomic particles. This is why your partial priority approach goes wrong when confronted with prior art disclosing "metal". The EBA in G 2/98 got this completely right.

I can only hope that the Enlarged Board is brave enough to step out of this particular corner that they have painted themselves into.

You do not agree that, as has been pointed out above, there are simple solutions to poisonous divisionals that do not require a complete rewrite of the case law on novelty, the disclosure test, and priorities (a rewrite that likely is impossible to get both consistent and sensible anyway)?

Proof of the pudding said...

Anon 2 Sep 01:06 - if you do not see arbitrary salami-slicing (and effective double patenting) as an absurd result, then you and I will have to agree to differ.

The patent system is intended to reward innovation. There is nothing "innovative" about arbitrarily dividing up a genus into different parts. I therefore see no reason why a second patent - providing up to an additional 18 months' protection - should be awarded in such circumstances.

Anyway, I really don't think that you have read my previous comments correctly. I was not proposing a solution to the problems with the novelty of selection - just merely highlighting that the concerns raised over notional splitting of claims were a red herring and should not prevent a solution to self-collision being sought in the interpretation of partial priorities.

MaxDrei said...

I'm impressed. If I knew what name by which to salute that last anonymous, posting very early this morning, I would do so. Well said, and even before that, well thought through, madam or sir! Thank you.

By the way, are we not now in times when contributers should cough up a pseudonym?

MaxDrei said...

This one is prompted by the comment from Pudding earlier this morning.
I see now what it is that troubles him/her: the spectre of 18 months of "double patenting".

I never did understand why so many people inside the EPO get their knickers in a twist over this supposed evil. Does it trouble the patent judges, the litigants, thie legal advisers? Not much, if at all.

Art 54(3) is in the EPC for quite another reason, namely to adjudicate priority between rival filers, each of which is inventive over the 54(2) state of the art and sincerely believes himselöf to be the very "First to File" and therefore entitled to a patent. In return for all the official fees they have paid, you could simply dole out patents on the same invention to each of these 18 month span petitioners. That really would be "double patenting" or triple or quadruple or whatever. There is a well-known patent attorney retired from service in an oil company, who gives training courses and puts up a slide to show how nine oil companies all hit on the same invention within 18 months of each other and then filed on it. All of them! Each thought it was going to get a pre-emptive monopoly. In the monumental scrap over who would get what, many of they filed cascades of divs. Even before self-collision of patent family members was invented, a huge task, to settle what claims each rival Applicant can take to grant.

Hence 54(3), to adjudicate between these rival filers. Compared to what's obvious, what is novel is relatively clean-cut, b&W issue, thank goodness.

I am OK with EPO law on 54(3) apart from the nonsense of self collision. I am OK with EPO law on "novelty of selection" and OK with "distinctly different invention" as the litmus test for divisionals.

What's not to like, in EPO Art 54 jurisprudence? Only 54(3) self-collision of patent family members.

Robin Jacob again. Observe the Americans, learn from them, but don't copy their mistakes. In their new statute, the AIA, 54(3) art is good for obviousness attacks. Talk about turbo-charging First to File. Not content with harmonising with the Rest of the World to First to File, the Congress has to make it a typically American "Winner Takes it All" second filer gets zilch system. Is that what we want (specially when we are acting for a party other than the one with the most elderly filing date)? I think not.

Pudding, do I do you wrong? Would you like to respond?

Proof of the pudding said...

@ Anon 2 Sep 01:06
You do not agree that, as has been pointed out above, there are simple solutions to poisonous divisionals that do not require a complete rewrite of the case law on novelty, the disclosure test, and priorities (a rewrite that likely is impossible to get both consistent and sensible anyway)?

Not really, no. I can see why "exempting" divisionals from Article 54(3) might help, but I cannot see any legislative basis for providing such an exemption in the law of novelty. Also, would you extend the exemption to priority documents that are published EP applications (or PCT applications that have entered the EP regional phase)? If not, then what you present as a "cure" for self-collision is only a partial cure at best.

On the other hand, partial priorities may well provide a complete cure. The basic concept would be to award partial priority to any disclosure of the priority document that represents an embodiment of the invention as defined in the claim of the application that is under consideration.

(Just to clarify, by "embodiment" I simply mean anything that, relative to the terms of the claim under consideration, is more precisely defined - i.e. anything from an "atom" to a (narrower) "cloud".)

I have seen nothing written so far that persuades me that it would not be possible to approach partial priority in this way. To my mind, if a disclosure of a document is capable of destroying novelty it is surely equally capable of giving rise to a right of (partial) priority.

Applying partial priorities in this way requires nothing other than jettisoning (or re-interpreting) some rather unfortunate language from G2/98 ("a limited number of number of clearly defined alternative subject-matters"). This would appear to be the simplest and cleanest solution possible - as not only does the problematic language appear to have no clear basis in the legislation, but a more generous interpretation of partial priorities arguably has solid basis in the (legislative intent behind) the Paris Convention.

With partial priorities potentially providing such a clean, compete and effective solution (which solution does not require the invention of some kind of special exception), you can perhaps understand why I am not tempted to look to Article 54(3) for a cure to self-collision.

Proof of the pudding said...

Max - I just believe that the bargain with the public that underpins the patent system demands a meaningful, technical contribution to the art before a patent can be awarded. The effective re-claiming of the same whole that has been arbitrarily divided into pieces just does not seem to me to meet that demand.

In any event, this is all a diversion from my main point about partial priorities. This is not only because is it unlikely that the kind of effective double patenting that I am envisaging will leave the realm of the hypothetical, but also because I do not believe that it is necessary to notionally split a claim into different parts in order to assess novelty in a situation of partial priorities. I refer you to my analogy above involving battleships.

Anonymous said...

Pudding - There is no "evergreening" - additional 18 months - if the later item must draw to the earlier item, for the clock even for the later item runs not with its filing, but from the earlier filing.

Max - The "choose a pseudonym" timeline is September 8 - as has been mentioned at least twice now.

MaxDrei said...

Pudding, you exhort me to get into your notional array of battleships. So I went back to it and tried again. I still don't "get" it. Sorry.

You tell me that each battleship has a length along the X axis that is determined by the "breadth" of the subject matter it discloses. I don't like equating scope of claims with enabling disclosure. Very often, applicants Claim wider than they enable. I assume then, that you mean "disclosure that is enabled". I suppose that each functioning gun on deck is an enabling disclosure (what I called above a balloon-puncturing "needle"). So might some ships have only one gun, or even no gun at all?

But then I read that even when a gun is functioning, it is capable of hitting only enemy ships. You seem to have come up with that rule out of whole cloth. The military have an expression "friendly fire". Our problem is that, on the face of it, Art 54(3) allows friendly fire.

So how do we stop it? Not just by simply announcing that friendly fire is forbidden. The military has tried that already, and it doesn't always work. Let's instead invent a bullet-proof vest that deflects friendly fire but is no defence against the guns of other parties.

That way, we can leave all other aspects of present 54(3) jurisprudence undisturbed. Agreed?



Proof of the pudding said...

Max - Yes, the "breadth" along the x-axis is determined by the enabled disclosure (i.e. a disclosure that is capable of destroying novelty - or giving rise to a right of priority).

However, I did not say that guns could not fire at "friendly" ships. It is just that the illustration that I gave did not involve "friendly fire" (as it was supposed to merely illustrate why "metal" with a partial priority for "copper" would not be novel over an intervening disclosure of "metal").

For divisionals, the analogy needs to include a further assumption. That is, we would need to imagine that the "priority" boats fire first and can hit the guns of an identically positioned "prior art" boat.

If the "priority" boats truly are identical to the "prior art" boats (which they should be in the case of divisional applications), then the latter would be left with no guns left to fire. Which means no Article 54(3) issue for the "application" ships. Problem solved, n'est-ce pas?

As you can see, this should work perfectly when the (enabled) disclosures that count as Article 54(3) prior art are identical to the disclosures that give rise to a right of partial priority.

This is because any priority-entitled disclosure in the application (i.e. a disclosure that has the potential to destroy the novelty of your claim) is "neutralised" by the fact that your claim is entitled to partial priority for precisely the same subject matter.

You can advocate your special "exemption" for divisionals under Article 54(3), if you wish. However, as that only does half a job, I would much rather stick with my alternative that seems to get the whole job done.

Anonymous said...

@Proof of the pudding:
Anon 2 Sep 01:06 - if you do not see arbitrary salami-slicing (and effective double patenting) as an absurd result, then you and I will have to agree to differ.

Suppose the application discloses:
- a composition comprising X,
- a composition comprising X and Y,
- a composition comprising X but not Y.
The application originally claims a composition comprising X.
A regular (non-poisonous) Art. 54(3) document discloses a composition comprising X and makes no mention of Y.

What do you do if the applicant replaces his single claim with two claims:
1. a composition comprising X and Y.
2. a composition comprising X but not Y.

Both claims 1 and 2 are now new over the document disclosing "a composition comprising X". Together these claims give protection for "a composition comprising X". I see no way around it, and I see no problem with it. The applicant benefits from having disclosures of the two alternatives in the application as filed.

The patent system is intended to reward innovation. There is nothing "innovative" about arbitrarily dividing up a genus into different parts.

Art. 54(3) does not require innovation but only novelty. The split is not arbitrary, as it has a basis in the application as filed.

Art. 54(3) is intended to avoid double protection, but it is an imperfect tool. If the earlier application claims metal and the later application claims copper, Art. 54(3) does not prevent a patent on the claim to copper, even though that claim protects many embodiments already protected by the earlier claim on metal (if granted).

Anonymous said...

@Proof of the pudding:
I can see why "exempting" divisionals from Article 54(3) might help, but I cannot see any legislative basis for providing such an exemption in the law of novelty. Also, would you extend the exemption to priority documents that are published EP applications (or PCT applications that have entered the EP regional phase)?

I don't propose to exempt divisionals. I propose to exempt any 54(3) disclosure of subject-matter with a particular date that is identical to subject-matter with the same date in the application itself. Allowing such subject-matter to be cited under Art. 54(3) would make a European application be novelty destroying for itself, which is so ridiculous that a legal basis for not allowing it is hardly needed.

So: if the application discloses copper with a certain priority date, then a 54(3) disclosure of copper with that date cannot be cited against the application, no matter whether you find that disclosure in the application, in a divisional, some other kind of family member, or even a completely unrelated EP.

There are just two small steps to take: (1) recognise that actual self-collision is not allowed under Art. 54(3), this identifies the subject-matter / date combinations that are "poisonous"; (2) extrapolate to arbitrary documents disclosing such subject-matter with the same date.

Another justification for this surgery can be found in G 1/03. In that decision, the EBA looked at Art. 54(3) and argued that if an earlier application only partially covers a later application, the rights to what is not covered belong to the applicant of the later application. To achieve that, the EBA allowed the introduction of a disclaimer (without any legal basis btw). The solution is not complete, as the disclaimer is not allowable if it "becomes relevant for the assessment of inventive step or sufficiency of disclosure". This is to prevent the applicant from benefiting from the disclaimer, which after all has no basis in the application.

In a case of self-collision, the rights to what is not covered by the earlier disclosure (i.e. metal minus copper) should still belong to the applicant. If we allow a disclaimer, the applicant can get both copper (provided it is disclosed in the application!) and metal minus copper (thanks to the disclaimer). The limitations of the disclaimer solution are now still present, but we can easily remove those by allowing the applicant to join the two parts in one claim to "metal" having the filing date as effective date. There is then no way the applicant can benefit from the undisclosed disclaimer and there is consequently no need to make the solution dependent on the absence of certain prior art.

If we now look back at what this (virtual) disclaimer approach achieves, we see that it nicely removes the disclosures from the prior art under 54(3) that I have outlined above. Or put another way, having a disclosure of "copper" in the application with a particular priority date protects against any 54(3) disclosure of copper with the same date (in the sense that those disclosures are not citable under 54(3)). This approach requires "copper" to be disclosed in the application itself (and not only in the priority document), which is why I prefer my approach over MaxDrei's (who proposes to let disclosures in the priority document act as shields).

Hmmm, it seems this G 1/03 approach could also protect against 54(3) disclosures of copper having a later date than the priority date of copper. If we accept that (I'm undecided), it becomes similar to a partial priority approach, EXCEPT that it does not disturb the examination of novelty and inventive step with respect to other 54(2) and 54(3) documents. That is exactly what we should want. With partial priorities, you change the prior art that can be cited against various parts of the claims, which gets messy (think of inventive step attacks based on a combination of documents and try to avoid an advantage from the undisclosed notional limitation).

The Cat the Walks by Himself said...

@Proof of the pudding:

You've reminded me school times when we played Battleship game:
https://en.wikipedia.org/wiki/Battleship_(game)

Patent attorneys, supporters of XY approach, can train their partial priority skills here:
http://en.battleship-game.org/

Coming back to the subject, the more I think about the partial priority issue the more I like the idea that a divisional application for the purpose of 54(3) EPC test should be treated as a third party application.

In other words, I think that "shielding by own priority application" would lead to double patenting and other uncertainties related to the first to file principle.

MaxDrei said...

Cat's last posting suggests that there are people out there who really do disapprove of divisionals. Perhaps they lament the cancellation of the 24 month cap on the time available for filing a divisional.

I can understand that view. There are applicants who file divisionals, one generation after another, so as always to have "something pending" at the EPO right up to the end of the 20 year patent term. This strikes many as an abuse. And probably it is!

But if it is an abuse, where's the harm? Given a strict approach under Art 123(2) EPC (and it certainly is a given), what progress can an Applicant make with his cascade of ever-pending divs.

Take the Medinol stent case, for example. Multiple generations of divs, DG1 allowing them every time, despite the prohibition on divs to inventions that are not distinctly different. But every time the owner asserted them, the involved court threw the complaint out, telling the patent owner to come back only when he has a patent that has survived EPO opposition proceedings under Art 100(c) EPC added matter).

Priority documents poisoning their own one year later filed family members really is a mischief. By contrast, divisionals as such, and directed to distinctly different inventions that were fully disclosed already in the PCT as filed, are not an abuse but only a fair way to deliver to a deserving Inventor, with reasonable legal certainty, a scope of protection that is commensurate with their contribution to the art.

The Cat that Walks by Himself said...

@ MaxDrei
Priority documents poisoning their own one year later filed family members really is a mischief

Allright, how about the following test:

Applicant A: a specific priority embodiment at P1.
Hypothetical Applicant B: genus embodiment at P2.
Applicant A: genus embodiment at P2.

Test: if B would be allowed to disclaim, then A is entitled to partial priority.

Anonymous said...

@Cat that Walks by Himself:
In other words, I think that "shielding by own priority application" would lead to double patenting and other uncertainties related to the first to file principle.

The solution I propose would let subject-matter with a particular date in the application itself "shield" the application against 54(3) disclosures of that same subject-matter with the same date (in the precise sense that those disclosures cannot be cited against the application under Art. 54(3)). Alternatively, one could let subject-matter in the priority document shield the application in a similar manner. (This is not a "partial priority" solution, as a priority-based solution would necessarily affect novelty under both 54(2) and (3) and inventive step.)

How would this lead to forms of "double patenting" that are not already possible now? Could you give an example?

MaxDrei said...

Cat I'm not understanding your point. What I see is that both A and B file on one and the same date P2 an enabling disclosure of a Genus.

However, A declares an earlier filing date P1, for a species within the genus. So A is entitled to a patent on the Genus including the species whereas B can have a patent for the Genus only if its claims disclaim A's species with earlier date P1.

It works out different though, if A files a divisional and allows it to go to A publication. Because then (thanks to 54(3) self-collision) A can no longer have a claim to the Genus, in either the parent or the divisional. This is because any such claim is entitled only to P2 and the other of A's A publications (with its P1 enabling disclosure of the species) destroys its novelty of the claim to the Genus. Why should the mere step of filing a divisional and taking it to A publication change what A can substantively claim?

Given all that, not in dispute, what are you driving at, please, in your last line?

The Cat that Walks by Himself said...

@MaxDrei
Cat I'm not understanding your point. What I see is that both A and B file on one and the same date P2 an enabling disclosure of a Genus.

Applicant B is hypothetical, only for the purpose of the test.

I compare

Situation I:
Applicant A: a specific priority embodiment at P1.
Applicant A: genus embodiment at P2.

with

Situation II:
Applicant A: a specific priority embodiment at P1.
Hypothetical Applicant B: genus embodiment at P2.

The test that I suggest treats all applications with priority date P2 as third party applications.

If in Situation II, Applicant B would be allowed to disclaim a specific embodiment of P1, then, in Situation I, Applicant A is entitled to partial priority for the specific embodiment of P1.

The idea behind is that patentability of all applications with later date P2 is evaluated in the same way agains an earlier specific embodiment of P1.

MaxDrei said...

Be that as it may, Cat, I want now to say that I work in engineering cases rather than chemical ones, so all this Genus/Species terminology is not fitting, and leaves me feeling uncomfortable.

Take my typical situation. Inventor tells the story, I draft and we file. Inventor continues to work on the invention.

Time to draft the PCT. We look at what the inventor brings new to the table, since we filed the priority document, and we map it on to the priority document. In describing the funnelling means, for example, we should better have used the word "receive" rather than "contain" and "tapering" rather than "conical".

In general, and sadly, I often find that, despite my best endeavours on behalf on the inventor, I could have phrased more felicitously, this that or the other passage in the priority document, the better to express more accurately, precisely and clearly the contribution the inventor's invention has given to the art.

BIG PROBLEM: do I change the wording of the priority document or not? If I do, the public benefits (and that after all is the point of it all, isn't it). But if I change the wording, might my clientrend up with no protection at all, purely because of Art 54(3) EPC and self-collision.

As you know, more or less ANY change of wording, after PCT filing, is greeted with a 123(2) objection, that subject matter has been added. Any change before PCT filing is just as scary because then my claims are deemed to be unworthy of Paris priority.

The likelihood of 54(3) art from another party is low. I can factor that into the risk assessment. If it happens, we have a priority competition and for that contest the rules have been well-understood since 1978. It's difficult, but it cannot be made any easier. At least it is less exhausting than declaring an interference between the two rival Applicants. Fortunately, it doesn't happen that often. It is a bit like the attitude expressed in G3/14 on the subject of poor clarity after issue. It's going to happen, from time to time. Tough though. We have to live with it. The courts can cope. I like that pragmatism.

But back to self-collision. Filing divisionals is an everyday matter, more than likely to happen than a priority contest between rival filers. After the EPO gives me the results of its first class search it could very likely be that my most ambitious claims are no longer tenable. Then, with self-collision, I find all my claims, in both parent and divisional, and however narrow their scope, all blown away, and this simply because in the PCT I made the big mistake of defining the invention in words more accurate and clear than those I used in the priority document. What has Genus and Species terminology to say to us, in this situation?

It is all very well those inside the EPO pontificating, that they have no sympathy, that all Applicant difficulties are self-imposed by incompetent drafting. Just learn to draft, they say, and all will be well. What do they know though, about the real world of drafting in the engineering arts, where it's all about choosing the words that are the most apt and fit for purpose. They should try it sometime.

Self-collision! It ain't fair. It ain't right.

Are you listening, EBA? Something should be done about it!

Anonymous said...

The EBA, in point 4 of G2/98, determined, through contrary interpretation of Article 4H of the Paris Convention, that the priority for a claim can be refused pursu¬ant to the provision of Article 4H of the Paris Convention if “elements” of the “invention” for which priority is claimed are not disclosed in the priority applica¬tion. On the other hand, according to the provision of Article 4F of the Paris Con¬vention, it is not possible to refuse a priority on the ground that an application contains “elements” that are not included in the priority application. In order to concurrently satisfy these two provisions that are conflicting at first glance, it is necessary to understand “element” not as a “feature” but as an “embodiment.”
However, this logic ignores an important prerequisite prescribed in the text of the Paris Convention. The EBA states, at the beginning of its reasoning, that “a claim, i.e. an ‘element of the invention’ within the meaning of Article 4H of the Paris Convention. . .” therefore, the logic is developed based on the prem¬ise that “claim” and “element of the invention” are the same concept. As it is understandable from the statement “priority . . . is to be acknowledged, if the sub¬ject-matter of the claim is specifically disclosed . . .,” the EBA argues about the validity of priority only for a claim. In addition, the subject matter of a priority is, as a given, limited to an “invention,” as it is stated in the middle of the reasoning that “according to Article 4H of the Paris Convention, an invention for which prior¬ity is claimed need not be defined in a claim of the application whose priority is claimed. . . .” However, in the text of Article 4H of the Paris Convention, is the subject matter of a priority necessarily limited to an “invention”?
A key to solving this question is hid¬den in a linguistic problem. The official text of the Paris Convention is in French, and according to Article 29 of the Paris Convention, in case of differences of opinion on the interpretation of the vari¬ous texts, the French text shall prevail. Then, it is obvious in terms of grammar that, in “certains éléments de l’invention pour lesquels on revendique la priorité” in French text of Article 4H of the Paris Convention, the antecedent, which re¬sponds to relative pronoun “lesquels” in relative clause “pour lesquels on re¬vendique la priorité,” is not “l’invention” but “éléments.”
Seen from this perspective, EBA’s limited interpretation of “element” as “embodiment” is a misunderstanding that was made based on the English text, in which it is ambiguous whether the an¬tecedent of relative pronoun “which” in relative clause “for which priority is claimed” in the provision of Article 4H of the Paris Convention is “invention” or “elements.” Therefore, it is significantly questionable to recognize such inter¬preta¬tion as justifi¬able (there may be cases where “element” is one of the choices of invention such as “embodiment,” but ele¬ment is not limited to “embodiment,” and it is natural to ordi-narily understand “element” as “feature” of the invention). Article 4H of the Paris Convention pro¬vides that “element” of the invention may become subject to priority, and it is a provision that forms a prerequisite for the multiple priority sys¬tem and the partial priority system. How¬ever, the EBA inter¬preted the provision in a twisted way.

Anonymous said...

EBoA Opinion G3/93, for the first time, argued about the validity of priority only for a claim. The EBoA instructed as follows:
“The corresponding Article 4, Section A(1), Paris Convention, makes no mention of the subject-matter of the subsequent application. It is generally held that the subsequent filing must concern the same subjectmatter as the first filing on which the right of priority is based [cf. R. Wieczorek, Die Unionspriorität im Patentrecht, Koln, Berlin, Bonn, München 1975, p. 149; G.H.C. Bodenhausen, Guide to the Application of the Paris Convention for the Protection of Industrial Property as Revised at Stockholm in 1967, Geneva 1968, at Article 4, Section A(1), sub (i)].”

Anonymous said...

But, at footnote 10 of p.156,Wieczorek says “Enthält die Nachanmeldung gegenüber der Erstanmeldung einen erfinderischen Überschuß, so liegt Erfindungsidentität dennoch vor, wenn dieser Überschuß für jeden Fachmann naheliegend war, also keine eigene Erfindungshöhe besitzt”.

Furthermore, at p.70, Wieczorek says “Klarzustellen bleibt noch, daß Teilprioritäten immer nur innerhalb einer Anmeldung, nicht aber innerhalb einzelner Ansprüche einer Anmeldung vorkommen können, denen als kleinster Einheit der geoffenbarten Erfindungsgedanken immer nur eine einheitliche Priorität zuerkannt werden kann”.

Anonymous said...

With regard to G.H.C. Bodenhausen, Guide to the Application of the Paris Convention for the Protection of Industrial Property, the actual statement at Article 4, Section A(1) is as follows:

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, utility models or inventor’s certificates it must concern the same invention or innovation. . . . Moreover, with regard to patents, special rules concerning the identity of the subject are given in Sections F, G and H of Article 4. These provisions will be commented upon below.

Furthermore, Bodenhausen explains multiple priorities and partial priority as stipulated in Article 4F of said Convention as follows:

It frequently happens that an invention is not immediately complete, so that, even after a patent application has been filed for it, improvements or additions are found which are made the subject of other patent applications. The Convention makes it possible to claim, in one and the same later application in other countries of the Union, separate (multiple) priorities for the different parts of the invention, based on the various first applications made in respect of each of those parts, provided, of course, that these various applications are filed within the term of priority counted from the first application. ……
It frequently happens that, after a first application for a patent has been filed, subsequent applications in respect of the same invention for which the priority of the first application is claimed contain elements of the invention which were not present in the first application and which either no separate patent application has been filed in time to claim multiple priorities or no separate patent application will be filed at all (for example, because the added elements do not, in themselves, have an inventive character). Under the Convention, such additions in later applications will not prevent priority from being recognized for those other elements of the invention which were already present in the first application.

Subscribe to the IPKat's posts by email here

Just pop your email address into the box and click 'Subscribe':