In the EPO gold standard test for added matter, what is deleted can be just as problematic as what is added. The unforgiving gold standard test for added matter requires that any amendment, including the deletion of a feature, must be directly and unambiguously derivable from the original application. The Board of Appeal in the recent decision of T 0422/23 rejected the Opposition Division's finding that a deleted feature was not essential (the "three point essentiality test"). Following the general direction of EPO case law, all that mattered for the Board of Appeal was whether the invention without the feature was directly and unambiguously disclosed in the original application. For the Board of Appeal, this did not require applying the three point essentiality test, because the outcome should be the same as the gold standard test. On the basis of this decision, the EPO's gold standard test seems to be only getting stricter with fewer opportunities for nuance.
Legal background: The 'gold standard' for added matter
At the EPO, the bar for amending an application or patent is infamously high. A European patent application or patent may not be amended in such a way that it adds subject matter which extends beyond the content of the application as filed (Article 123(2) EPC). The established case law for assessing added subject matter takes the form of the gold standard test, which asks whether the subject matter of the amendment can be derived directly and unambiguously by a skilled person from the application as filed as a whole. For divisional applications, the same principle applies with respect to the earlier application as filed (Article 76(1) EPC). Whilst most jurisdictions have similar provisions, the EPO approach to added matter is extraordinarily strict. Indeed, added matter may result from as little as a misplaced comma in the claims, as in the catastrophic comma loss cases (IPKat, IPKat).
 |
| Golden |
Importantly, the test for novelty at the EPO forms the other side of the coin of the added matter test. In particular, if a disclosure, selection or amendment is novel, then it is also added matter (and vice versa). The case law on novelty and added matter at the EPO is therefore inextricably linked, and the application of the gold standard test is relevant to both. In recent years, the Boards of Appeal have repeatedly emphasised that the gold standard test supersedes any other "special" tests that have previously been suggested for assessing novelty and added matter for certain types of inventions, including selection from lists (IPKat) and inventions relating to the increased purity of a known compound (IPKat).
Case background: The case of the disappearing battery
The patent in question in the present case (EP3571944), owned by Joyetech (Changzhou) Electronics Ltd., related to an intelligent controller for an electronic cigarette. The invention aimed to provide a controller that could detect the electrical condition of the atomizer's heating wire, for example, to determine if there was a short-circuit or open-circuit. The Opponents were anonymous behind strawperson oppositions.
At Opposition, the Opposition Division (OD) decided to maintain the patent on the basis of an auxiliary request. The independent claim of this request defined an electronic cigarette comprising an intelligent controller with various modules: a switch, a voltage acquisition module, a control module, and a display module. Importantly, the original application (of which the case in question was a divisional) described and claimed an intelligent controller that also comprised a battery, which was defined as being electrically connected to the other modules to supply them with power. In the version maintained by the Opposition Division, the feature of the battery had been removed from the definition of the intelligent controller.
The Opponents appealed the OD decision, arguing inter alia that the omission of the battery feature constituted added subject matter. The OD had concluded that the battery could be omitted without violating Article 76(1) EPC because the language relating to the battery was not "functionally or linguistically associated with the remaining features". The OD specifically applied the so-called "three point essentiality test" with respect to the removal of features. According to this test, an amendment removing a feature from a claim does not add matter if:
(i) the replaced or removed feature was not explained as essential in the originally filed disclosure;
(ii) the skilled person would directly and unambiguously recognise that the feature is not, as such, indispensable for the function of the invention in the light of the technical problem the invention serves to solve; and
(iii) the skilled person would recognise that the replacement or removal requires no modification of one or more features to compensate for the change (it does not in itself alter the invention).
Applying this test, the OD found that the removal of the battery feature did not add matter on the basis that the presence of a battery would not be recognised as an essential feature by a skilled person, and that removing this feature from the claim would not require modification of any of the remaining features. The OD further noted that, applying also the gold standard test, it would be evident to a skilled person that the language of the claim in the original application relating to the battery was "not functionally or linguistically associated with the remaining features of the claim", and so could be omitted without adding matter.
Always the gold standard
On appeal, the question before the Board of Appeal was whether the OD had properly applied the test for added matter with relation to the omission of the battery feature. Unfortunately for the Patentee, the Board of Appeal did not follow the OD reasoning. The Board of Appeal particularly rejected the argument that the battery was a "severable" feature. Instead, in the view of the Board of Appeal, the battery was "without a doubt originally defined to be a part of the electronic controller" The Board of Appeal also found the OD's conclusion on the functional connection hard to follow, stating: "the battery is defined there as being connected to the switch module, the voltage acquisition module, the control module and the display module and moreover as being configured to supply working voltage power to these elements, which is without doubt its function. Therefore, the battery is originally disclosed functionally [as well as structurally and indeed linguistically] associated with those elements and presented as being an integral part of the intelligent controller."
For the Board of Appeal, it was therefore clear that the omission of the battery feature added matter. The original application disclosed an intelligent controller comprising a battery connected and functioning in a specific way. As such, an intelligent controller without such a battery could not be said to be disclosed, directly or unambiguously. For the Board of Appeal therefore, the claims thus contained added subject matter, and the main request on appeal was not allowable.
Final thoughts
The EPO's "gold standard" for added matter is notoriously unforgiving, and often far stricter than similar tests applied in (any?) other jurisdictions (IPKat). The present case highlights how the gold standard may be applied with equal force to the removal of features as it does to their addition or combination. In line with the recent case law rejecting special tests that add nuance to the gold standard for different subject matter or inventions, the Board of Appeal relied solely on the gold standard test instead of following the OD approach of evaluation according to the three point essentiality test. For this Kat, T 0422/23 is perhaps a warning that the EPO may consider most features present in an original independent claim as "essential", or at the very least that arguing for the non-essentiality of a feature will be an uphill battle. From the drafting perspective, therefore, the requirements of the EPO teach us to be ruthless with deleting inessential features from the independent claims before the application is filed.
Further reading
- Boeing's comma drama: Commas and taking the description into account when construing a claim (T 1127/16) (Apr 2021)
- Gold Standard test for novelty reigns supreme, even for subranges (T 1688/20) (Jan 2023)
- Another case of catastrophic comma loss (T 1473/19): Interpreting the claims in view of the description (Jan 2023)
- Defining what the invention is not can be as important as defining what the invention is (T 0273/22, Chimeric antibodies/REGENERON) (13 Nov 2023)
- Unambiguous disclosure without patent profanity (T 2171/21) (20 Dec 2023)
- Pitfalls of cell therapy manufacturing IP - A case study (T 0868/23) (7 Jul 2025)
Reviewed by Dr Rose Hughes
on
Sunday, September 28, 2025
Rating:
Follow the IPKat on LinkedIn 
![[Book Review] Decolonizing Intellectual Property Law – An Afrocenteric Approach](https://blogger.googleusercontent.com/img/a/AVvXsEiLxxU1FAFmN4Kz-lkqEmoEQPwj8Yi6OcAENC94bZ9A2-MObgI1bqc9y5AOtyDN89b8TXfzDXdMI03idZYOrThD3xgbQXv9nz9k9x_0bdrWHwp4tAKvHqzas8RgQxmdGuLqqJ9pel3t5XPuRPndnxaiGsfcC8YiRtiV5UlUhNFM9DC8Mtlcm526Ew=s72-w244-c-h320)
Posts
The EPO prides itself on being "real world" but is this decision in such a world, I wonder.
ReplyDeleteI'm OK with the Gold Standard for all of i) added matter ii) novelty and iii) priority, so long as its "derivable" is interpreted fairly.
So, suppose, here, the document under scrutiny for what it discloses was instead of the application as filed a prior publication cited under Art 54 EPC and would destroy novelty if it discloses to the skilled reader a system as claimed ie one not having a battery. To me it seems a tad unworldly that the skilled reader is incapable of "deriving" (directly and unambiguously imagining?) the system undergoing a change of battery. Just because the definition of the subject matter sought to be protected includes a battery does not unassailably prove that the disclosure in the document is no more than that. Documents routinely disclose more than they claim. A publications are a resource from which skilled readers derive as much as they can imagine.
I think we need more cases, so the Boards can tell us what it means for a skilled reader, deemed hungry for technical teaching, and with a "mind willing to understand" to "derive" something from a printed document. After all, the Boards have told us already that the skilled reader is more than a logician, a mathematician or a mere wordsmith. How much more, I should like to know.
Max Drei, Just in case you are still looking at comments on this blog, what would you say to a proposal that an EPO Examining Division should include an economist, who would evaluate the claim scope in respect of how it impacts the economy? I'm very curious to hear your view. I believe having an economist as part of the Examining Division would be very helpful (in an ideal world). Thanks
DeleteYou're pulling my leg aren't you Santa? Ideal world? What's that then? The economist member would have even less work to do than the legal member.
DeleteThank you Max Drei. I think I will put this particular personal project on hold until there is more support for the proposal.
DeleteThe stricter the EPO gets on added matter, the more protection there is for third parties who need to operate in that same space. At the moment the system is rigged to be pro-patentee because they are the people paying to maintain it, and it is fair they derive most benefits, but we do need to consider also what can third parties rely on to be sure of what their freedom to operate is. The assumption that the independent claims on filing represent the maximum degree of protection and the minimal essential features seems to be a fair one, and that is what the EPO is tending to informally use as a starting point, it seem to me. That may be the way forward. Thank you for a great writeup Rose.
ReplyDeleteEquality of arms, between patent applicants and their competitors? You make a good point, Santa, very persuasively. A "starting point" that is "fair" strikes me as very important, to preserve and nurture confidence in the ability of a patent system to stimulate technical innovation and thereby to "promote the progress" of society. The stringency of the EPO's Gold Standard plays a valuable role in curbing abuses in the filing of divisionals, generation upon generation.
DeleteThe EPO has, from the get go, expected higher drafting standards from its applicant community. Perhaps this is because, at least until now, it has been free from supervision by courts of appeal operating in the civil litigation space free from input from the specialist patent practitioner world, with judges ignorant of what standards can reasonably be expected from those who draft patent applications, and in consequence easily led astray by zealous and plausible litigator advocates.
Thanks Max Drei. Certainly the sheer volume of cases the EPO gets through, and the conscious desire to grant and maintain valid cases means there is a lot which the EPO does right compared to other patent offices. It undergoes a constant evolution in case law and patent practice which is responsive to changing needs and new technologies. However it still has its faults. Some Examining Divisions are more strict than others. Some Opposition Divisions are too patentee friendly. There is clearly injustice on individual cases, and opposition oral proceedings can be too unpredictable with decisions being taken on advocacy rather than merit of the arguments. In response to your point about curbing abuses, the big weakness here is that 'abuse' has not be codified into the EPC. Here the large corporations still have too much power, and the EPC does not give legislative support to creating case law as to what 'abuse' is and how it will impact the EPO's actions. No one is even talking about this (as SME's don't have a voice), which means it won't be addressed soon. Thank you again for your thoughts
DeleteI have not read the patent but typically batteries are replaceable, and when something comprising a battery is disclosed, you typically automatically conceive its existence also without the battery, for example, as someone pointed out above, when time has come to exchange the battery. In this sense, a device without a battery would appear to be at least implicitly disclosed. Or not?
ReplyDelete@ Max Drei and Magnus Stiebe,
ReplyDeleteI have commented the present decision in my blog.
Deletion of features always rings the bell of the three point test, which is still in the Guidelines H-V, 3.1 so that first instance divisions are still allowed to use it. Applied properly, it leads to the same result as the “golden standard” defined in G 2/10.
You do however seem to forget one essential aspect: without a battery, or in general a power supply, an electronic cigarette is useless. In the present case, the disclosure in the parent application mentioned a battery.
Extrapolating to a device without a battery, when one with a battery was originally disclosed, represents added matter, whether you like it or not.
It is understandable that the manufacturer could sell its electronic cigarette, without a power supply. But in the present case this was not originally disclosed. And here is the problem.
It is obvious that the electronic cigarette can be sold without a battery or a power supply, or be without a power supply when not in use, but this relates to obviousness and not to novelty.
A piece of prior art with the same features as claimed, but without a battery would not be novelty destroying, but the claim would manifestly lack inventive step. The criterion for assessing added matter is novelty, not inventive step.
Added matter and novelty are the two sides of the same coin. What is missed on side, is gained on the other, but you cannot have both at the same time.
See for instance T 583/09: the term "implicit disclosure" should not be construed to mean matter that does not belong to the content of the technical information provided by a document but may be rendered obvious on the basis of the content of the disclosure.
As one commentator said on my blog, it was poor drafting, but this is what often happens. Once filed the applicant/proprietor is stuck. My experience is that often applicants/proprietors want to save costs when filing. This can later have dear consequences.
Without disagreeing with anything you said, I think 'implicit disclosure' remains a deep mystery at the EPO, and is no longer used by Examining and Opposition Divisions, in my experience. I do not believe any definition of it has ever been given in EPO case law.
DeleteMy problem is that the word derivable strikes me as having a wider meaning than implicit. But if the Boards tell us that it means the same that's OK with me.
DeleteAdded subject matter you say. Well, looking at the figures, we have Fig. 1 with no battery and we have various figures showing circuitry, with each indicating a battery connection but no battery. The EPO has interpreted the gold standard to have the meaning "explicitly disclosed" and this is the wrong standard. The standard is, or should be, what the skilled person would understand, but then it is difficult for the members of the boards to put themselves in the shoes of a skilled person.
DeleteOf course, every case has to be judged on its unique matrix of fact, and of course the skilled reader already has a good understanding of the meaning of "implicit" and "derivable". But how about this hypothetical:
Delete1) I file A1 with drawings of my hand-held battery-powered device, showing an empty battery compartment to which is given a reference number that the description identifies as "battery compartment".
2) A few months later I file A2 in which the only difference is that I draw a battery inside the compartment, give it a reference number and claim the device including the battery. Does my claim survive Art 54(3) attack based on A1? Is the battery of A2 "implicit" in A1? Is it "derivable" from A1?
Perhaps the English doctrine of "inevitable result" (of practising what A1 discloses) can help us to define what is "implicit"?
When looking at the original application, the battery is a feature of claim 1. I would say this creates at least a presumption that it is an « essential feature ». The question is : is the presumption rebuttable in the light of the description ?
ReplyDeleteAs a matter of fact, the description refers many times to the battery. While it does not say explicitly that it is « essential », the battery is present in all the disclosed embodiments, so that it looks like a component which is always necessarily present. And what is very significant, the description never mentions or suggests that the battery could be dispensed with.
Mr Hagel, your comment brings to mind my two mobile phones. The old one has a battery which I can remove, the new one not.
Delete@ Santa,
DeleteI do not think that 'implicit disclosure' is a deep mystery at the EPO. Implicit features are dealt with in the Guidelines G-VI, 2 (the rubber ball) under novelty, and G-VI, 5.1 under novelty and parameters. On the 367 decisions of the board found under the keyword “implicit”, there are quite a few decisions in which the boards have acknowledged novelty in denying that the prior art contains implicit features. There are also decisions in which novelty has been denied due to implicit features.
I have devised the following definitions of what is explicit/implicit:
Explicit disclosure = same terminology in the prior art and in the claim. For an explicit feature bibliographic data is enough.
Implicit disclosure = different terminology, but identical feature for the skilled person in view of its general knowledge. For an implicit feature, bibliographic data is not enough, an explanation has to be given to show the identity of features.
Identical does not mean equivalent.
@ Max Drei,
A lot might derivable from a disclosure, for examples equivalents. When it comes to added matter, it has to be “directly and unambiguously derivable”. This is a limitative definition.
In your example, A2 will not survive an Art 54(3) attack based on A1. A1 and A2 disclose the same device and the essence of a battery compartment is to receive a battery. Through the designation in A1 of a battery compartment, disclosing a battery in A2 is, imho, "implicit" in A1, and hence directly and unambiguously "derivable" from A1. The "inevitable result" of a battery compartment is actually to receive a battery.
@ Kant
The EPO has NOT interpreted the gold standard to have the meaning "explicitly disclosed" and this is the NOT wrong standard. The gold standard also includes what is “implicitly disclosed”, i.e. identical for the skilled person. See above.
@ Mr Hagel,
Delete@ Mr Hagel
Essential is, for good reasons, a word carefully avoided in descriptions.
A feature F might nevertheless be essential without being declared so. It may result from the circumstances, in that, for instance, said feature F is found in all embodiments disclosed. This feature is then implicitly essential. Any independent claim without this feature might thus represent an intermediate generalisation.
It's beginning to look as if adding a battery is implicit but the converse, removing one, might be patentably novel. Does that make sense? It will depend upon the facts.
Delete@ Max Drei,
DeletePlease don’t confuse the issues. In the present case, the battery was disclosed in the description and mentioned in the independent claim.
In this constellation, there is no other possibility than to consider the deletion of the battery as resulting in an intermediate generalisation.
The parent application was simply badly drafted, but this is not the fault of the EPO.
It seems to me that often added subject-matter problems arise because often the EPO adopts a check box box approach to examination (aka the so-called gold standard). In cases like this, they start from the reference to essential features of the invention that appears in Rule 43 EPC. However, does the skilled person reading a patent specification really base their determination of the essential features of an invention on this technical requirement relating to claim drafting, or a real world understanding of the essential features of the invention proper?
DeleteIt goes without saying that a power source for the claimed intelligent controller is essential to work the device, but is the form of the power source an essential feature of the invention proper as it would be understood by the skilled person? I have not read more than can be found on this webpage, but on the face of what I have read, it looks as though the approach taken by the OD produced a result that is likely to be in line with what the skilled person would have derived from the application. By way of an analogy, if an invention relates to the features of a cam belt system in an internal combustion engine, on a macro level a fuel supply is clearly an essential feature since without it the engine, and by extension, the invention will not 'work'. However, in terms of the actual invention as it would be understood by the skilled person looking that what the invention really does/sets out to do, the fuel supply is an irrelevance. In this case, if the independent claim specified an internal engine comprising a cam belt system etc, but made no mention of a fuel supply system, I cannot see an EPO Examiner relying on Rule 43 to require that a fuel supply system is added to the claim on the basis it is an essential feature. Equally, in the present case, in the absence of any disclosure that a battery or particular form of battery is an essential element of the invention proper and in the absence of a specific disclosure of power sources not originally disclosed resulting from the amendment, it seems to me that the three step investigation adopted by the OD provides a result likely to be far closer to how the skilled person would view things that adopted by the BOA.
Agreed, Daniel, that the thing was badly drafted. Agreed that to delete the battery is to create an "intermediate generalisation". But that alone doesn't end the debate, does it?
DeleteAs I see it, the decisive question is whether, in making that creation, one has added matter to the disclosure, in other words,whether it is one of those "intermediate generalisations" that was "undisclosed" in the original filing. Every dependent claim is directed to an "intermediate" generalisation. Not every amendment to a dependent claim, not every newly-formulated "intermediate generalisation", adds matter, does it?
When I draw a distinction between adding a battery and subtracting a battery I think I am not confusing the issue but, rather, forcing us all to think what it is that makes adding a battery to the claim OK but subtracting one not OK.
I revert to my mobile phone example. In my old phone there is a battery compartment and I can take the battery out whenever I please. In my new one, the battery is "built-in" and the end user can't remove it. I see that as an illustration of the importance of the unique facts of any given case, that will determine when matter is added and when it is not.
For what it's worth, I am now at the point where I do not disagree with the Board's decision here. You, and others in this thread, have convinced me that it is in order.
Dear Max Drei,
DeleteI have some problems with your understanding of dependent claims.
In my opinion, dependent claims are not intermediate generalisations, but optional features to be added to the independent claim. This is why it is important when the subject-matter of a dependent claim is added to an independent claim in order to limit it, that the description is amended and the optional character of the corresponding features in the description to be deleted. This is also what is required in G 1/24 and all decisions applying it.
However, I do fully agree with you that it is very important to look at the specific case before judging what is directly and unambiguously derivable from the original disclosure.
As far as battery of a mobile phone is concerned, nowadays end users cannot remove them, put as service point can do this easily. With the proper equipment, even an end user can do it, so your example is not fully convincing.
I have seen during my active time so many cases in which the original drafting was sub-optimal. But this is often due to the fact that applicants are saving at the wrong place. That’s an old wisdom, you get what you paid for.
@Anonymous,
DeleteI am very weary with hypothetic cases. They are often over simplifying and do not reflect the proper situation.
When the invention is about a wind screen wiper for a car, nobody would care whether the car has three, four of five doors, three or four wheels, an electric or a thermic engine. Even the windscreen is not essential if the skilled person sees that the wind screen wiper works on any kind of windscreen, front or back.
What is essential is that all the features allowing the windscreen to solve the objective technical problem and to produce the corresponding effect..
In this respect, in your example of a cam belt system, it would indeed never come to the mind of an examiner to request the fuel system to be added to the claim for a cam belt system.
In the present case, the OD might have thought it is obvious for the battery to be dispensed with. But this is not what the gold standard requires. In the present case, the electronic cigarette was only ever disclosed with a battery.
In the present case, it would have been enough to mention a battery to power the controller. The type of battery, rechargeable or not, is then irrelevant. This was missed when the application was originally filed.
If a battery was mentioned in the priority document, deleting, or leaving the battery open, would not amount to added matter, but would result in the priority not being valid.
You are weary of hypotheticals, but then choose to add one of your own! However, you actually get to the point I was making in your comment on the windscreen wiper. In view of the disclosure of the application as a whole, would the skilled person consider the battery an essential feature of the claimed invention that was actually required in order to to solve the objective technical problem? On the face of it, the OD approach seems to address that issue rather better than the check box approach so often employed at the EPO.
DeleteAs to whether a dependent claim can fairly be described as an "intermediate generalisation" I was thinking of the drafting process. The inventor shows you the individual embodiment (combination of technical features) that has solved their technical problem. I, as drafter, climb what some Americans call the "ladder of abstraction", trying to reach the highest level of abstraction that nevertheless still solves the problem, by systematically removing technical details found in the embodiment. The rungs of that ladder are the basis for the dependent claims. The highest point on the ladder is claim 1. It is in that sense that I see the dependent claims each as a generalisation(from the inventor's individual embodiment) each at an "intermediate" level of generalisation.
ReplyDeleteSo I hesitate to accept your opinion that dependent claims are not intermediate generalisations.
I share your view that hypotheticals are often unhelpful. But sometimes a carefully formulated hypo can help. I guess that every patent attorney exam question is a hypothetical and helps to illuminate whether it is safe to award the candidate a certficate to practise as a patent attorney.
Dear Max Drei,
DeleteI will try to explain my vision and try to show that dependent claims cannot correspond to intermediate generalisations.
I agree that the highest level of abstraction is that of the independent claim.
Let’s say you start with a claim with lots of features and you remove systematically features which are not essential and which still solve the problem the inventor is trying to solve and are novel and inventive.
This original claim corresponding to the embodiment comprises the features ABCDEFG.
Once the features which are not strictly necessary for solving the problem, but still offering novelty and inventive step are deleted, we end up with ABCD in the independent claim, and claim 2/1 comprising EF and claim 3/2or1 comprising G.
If EF are inextricably linked, then only claiming E or F in claim 2/1 represents an intermediate generalisation of the features EF.
I understand the term intermediate generalisation if, in a claim WXYZ, Z is not kept, but is indispensable to obtain a given effect. Claim WXY represents an intermediate generalisation of the claim WXYZ.
Coming back to the example above, the claim ABCD can be held to represent an intermediate generalisation of the claim with all bells and whistles ABCDEFG.
Dependent claims 2/1 comprising EF and claim 3/2or1 comprising G cannot thus not be considered as an intermediate generalisation, as they are representing optional features allowing to limit the independent claim.
Could I convince you now?
Sorry to say, Daniel, I am not bright enough to follow your thinking. I will therefore try again, to explain my thinking. Perhaps your reply to this fresh attempt will help me to see the light.
DeleteHaving trained in the UK, I see a claim as defining the boundary between what subject matter is protected by the patent and what is outside the scope of protection. A "boundary fence" if you like.
Or the membrane of a balloon that defines the protected space.
I see the optimal embodiment, what the Americans call the "best mode", as a diamond inside the balloon. I see the dependent claims as successive concentric balloon membranes, inside the claim 1 balloon. Or, in a medieval castle, successive concentric rings (wall or moat) inside the outermost boundary wall. These concentric rings represent intermediate protective layers. In that they protect intermediately large volumes of space they are intermediate generalisations of the inventive concept most perefectly embodied as that diamond in the middle of the protected space.
Now I should turn to the reply from Francis Hagel.
Dear Max Drei,
DeleteThe present reply is quite long, but I hope it will help clarifying the situation.
Here is part 1
There is no need to be bright or not in order to follow my line of reasoning. I think that we are speaking about the same process, simply our definitions do not tie up.
I can follow you with your idea of the balloons and the diamond being in the smallest balloon inside the claim 1 balloon. The balloons can also be compared with the famous Russian dolls in which the smallest one is dug deep inside the broadest one, with plenty of others in-between.
You are talking about intermediate layers of protection, and I can perfectly follow you there. It is manifest that starting from the most inbound layer, any broader layer represents a generalisation of the most inbound layer or from the preceding inbound layer, in order to arrive at the broadest or most general protection layer, i.e. that of the independent claim.
At the EPO this is not an what is meant by an intermediate generalisation. In the EPO jargon, an intermediate generalisation occurs when a feature necessary to obtain a certain effect and which is inextricably linked with the other features, has been deleted.
I hope that the following example will allow you to grasp what is understood by an intermediate generalisation at the EPO.
The description discloses a perforated layer in which the holes have to achieve the following effect: let air circulate from one side of the layer to the other one, whereas water cannot flow in the opposite direction.
In the description the size of the holes is defined in functional terms: the holes are large enough to let air pass in one direction, but small enough for water not to pass in the opposite direction. It is actually the definition of a Gore-TexTM membrane.
The claim does however only mention a perforated layer. This represents an intermediate generalisation of the teaching disclosed in the description, and hence is to be queried under Art 123(2). The teaching in the claim is broader, i.e. generalised over the teaching in the description. This is the definition of an intermediate generalisation at the EPO.
In order for the perforated layer to achieve the effect disclosed in the description, the holes of the perforated layer have to have a size which fulfils the double condition, large enough to allow air to pass through, but small enough for avoiding water ingress. It is only if this double condition is introduced in the claim, that the objection under Art 123(2) can be lifted.
In other words, the applicant/proprietor does not sit in an inescapable trap according to G 1/93 in case of an intermediate generalisation. The incomplete claim can be limited by adding the features found in the original disclosure.
Reply to Max Drei, part 2,
DeleteGeneralisation allowable
In the absence of any clearly recognisable functional or structural relationship among the features of a specific combination, see T 2787/19, or if the extracted feature is not inextricably linked with the omitted features, see T 352/19, then the „golden standard“ of G 2/10 is satisfied. In such a situation there is no intermediate generalisation.
Not allowable intermediate generalisation
In T 3069/19 it was held that, according to established case law, cf. T 219/09, Reasons 3.1, or T 1944/10, Reasons 3.2, it is normally not allowable to base an amended claim on the extraction of isolated features from a set of features originally disclosed only in combination, e.g. a specific embodiment in the description or drawings of the original application.
Amended subject-matter that amounts to a generalisation of a particular embodiment disclosed in the original application but is still more specific than the original definition of the invention in general terms (claim 1) is called an "intermediate generalisation", cf. T 191/04, T 2311/10, and even sometimes an "intermediate restriction“, cf. T 461/05, T 879/09.
Application of this principle to claiming a priority
For the same token, if in the priority document the description discloses the functional features of the size and the corresponding effect, then a claim limited to a perforated layer in the subsequent application will not benefit from the priority. See also T 59/11 or T 330/14.
There is no added matter or intermediate generalisation in the meaning of Art 123(2) between the priority application and the subsequent application, but what is claimed in the subsequent application is not the same invention as that disclosed in the priority application under G 1/98.
General considerations
All what I have explained here is also fully in accordance with G 1/24 and the decisions applying it. If the applicant/proprietor wants the interpretation given in the description to be taken into account, he must bring it in the independent claim. This works not only if the description contains a limiting interpretation, but applies as well if the description contains a broadening interpretation.
Priority, novelty, inventive step, added matter all follow the same rule. One starts with the claim and checks whether the definition the skilled person derives from the claimed features ties up with the definition in the description. If this is not the case, the only way for the applicant/proprietor to remedy the defect is to amend the claim. As made clear in T 2027/23, this avoids interpretative somersaults, and there is no need for the patent to be its own dictionary.
To come back to our original problem: if the description systematically discloses a battery in combination with the electronic cigarette, deleting the battery in a claim represents an intermediate generalisation. It is irrelevant whether it is obvious that the battery could be deleted. The deletion of the battery is not directly and unambiguously derivable from the original disclosure.
@MaxDrei
ReplyDeleteAs a corporate practitioner, I see the claim drafting exercise as involving other dimensions than the level of abstraction related to the terms of the claim. This is because the preliminary parameters of the claimed subject matter i.e. the field of the invention and the category of claim (product/apparatus/process) give the opportunity for strategic decisions.
A first key issue is the definition of the field of the invention as reflected in the claims. Compare a broad definition only based on a category of devices, such as a category of components e.g. a capacitor, or a pressure sensor, with a more specific definition including a field of use (low voltage vs hign voltage capacitor, or a pressure sensor for use in a car, or in an oil well, or for space applications, etc). An automatic reliance on the broad definition may be at odds with the applicant’s objectives focused on the field of use of real interest and invites trouble during prosecution, esp. if the applicant has not included suitable fallback positions in the application.
The second issue is the choice of the claim category. Depending on the nature of the invention and the business context, it may be well advised to write claims which match the activities along the supply chain. For example, you may claim both an apparatus and a subsystem featuring the same inventive concept so as to target suppliers’ activities. Suppliers’ infringing activities are thus direct infringement, not just contributory infringement. Or you may target distributors or users by claiming a method of using the apparatus.
Indeed. Can't quarrel with any of that, Francis Hagel. Would you care to get involved in the discussion between D X Thomas and I, as to whether a dependent claim can ever fairly be described as an "intermediate generalisation"?
DeleteDear Max Drei,
DeleteMr Hagel’s considerations are very interesting and useful when it comes to define a protection strategy, but do actually not reply to the question of what is an intermediate generalisation.
I hope that my long reply has clarified the picture. What you rightly consider a generalisation when you come from the most limited claim to a higher ranking claim, is not an intermediate generalisation under EPO practice. An intermediate generalisation occurs when inextricably linked features are ripped apart.
I hope I could now convince you.
Well, Daniel, I now have a better grasp of your line of thinking. Thank you. Just one question though: what (at the EPO) is the difference in meaning (if any) between a) an intermediate generalisation and b) an undisclosed intermediate generalisation? In other words, is that extra word "undisclosed" 100% redundant (at the EPO)?
DeleteDear Max Drei,
DeleteAn intermediate generalisation is in essence "not directly and unambiguously derivable" from the original disclosure. Adding the term “undisclosed” to “intermediate generalisation” is therefore absolutely redundant.
What I have explained in my long reply, is what I have been teaching in lots of venues when it comes to added matter and intermediate generalisation, for instance when training candidates for the EQE. It is not my invention, but results directly and unambiguously from the case law of the EBA and of the BAs.
T 164/21 and T 1906/11
In T 164/21, it was held that in order to determine whether an amendment conforms with the requirements of Art 123(2), the only relevant question is whether a skilled person faced with the amended version of the application, as compared to a skilled person having seen only the version originally disclosed, would derive from that amended version any additional technically relevant information.
This position is in line with decision T 1906/11, in which it was held that, whether an amendment constitutes an intermediate generalisation is irrelevant, in order to decide whether it is allowable or not. In presence of additional technically relevant information a violation of the provisions of Art 123(2) occurs.
An intermediate generalisation boils down to adding technical information to an original disclosure, like for any other infringement of the "gold standard". .
Intermediate generalisation or broadening of features?
It is disputable whether broadening of features and intermediate generalisations are different in nature. In both cases, there is an isolation of certain features which are then generalised.
Case law seems to conclude more often to an intermediate generalisation than to a broadening of features.
See also T 88/10, T 582/08, T 1887/10, as examples of not allowable intermediate generalisations.
See also T 1906/11, T 116/20, as examples of allowable intermediate generalisations.
All the decisions I have cited in my replies to you, stem from an extensive PPt presentation on added matter. This presentation is continuously updated.
Nothing in the EPC prevents the proprietor for amending the claims to incorporate an intermediate generalisation. What A. 123(2) does prevent is the incorporation of an undisclosed intermediate generalisation. As MaxDrei has succinctly pointed out, dependent claims are intermediate generalisations in that they add particular features to the claimed subject matter of the higher claim but not the totality of the disclosure of the embodiment.
DeleteSigh. Kant it is as if you have not taken the time and trouble to read the thread of comments. The term "intermediate generalisation" means one thing to you but, according to Daniel Thomas, something different within the EPO.
DeleteAccordingly, it might help everybody if the EPO's people when writing their Decisions, stick to "Gold Standard" terminology and eschew, from now on, use of the term "intermediate generalisation".
That, or instead use the term "undisclosed intermediate generalisation"! Then, everybody would be on the same page!
But, apart from that, thanks for your support. I thought I must be the odd mann out and I'm happy that I'm not.
@MaxDrei
ReplyDeleteComing after Mr Thomas’ answers, I am not sure this will be of any help.
It seems your debate with Mr Thomas starts from a misunderstanding. Mr Thomas uses « intermediate generalisation » in accordance with EPO parlance i.e. including the implicit addition that it is inadmissible new matter, whereas for you an intermediate generalisation may be perfectly admissible, as you point out when you link intermediate generalisations to dependent claims positioned at different abstraction levels.
But you refer to different contexts : you refer to the drafting process i.e. the context of filing an application, and there can be no new matter issue for dependent claims of an original application, while Mr Thomas refers to post-filing amendments in the form of intermediate generalisations which may indeed be inadmissible new matter.
That being said, I tend to agree with Mr Thomas’ comments as to what is or not inadmissible according to the case law.
As to the general question of broadening claims after filing, this is not prohibited before grant, but it may be an uphill battle, as there is the Art 123(2) issue but in addition there has to be in the original application technical information relevant to the broadened claim to support inventive step arguments.
Good. Thank you, both. It is clearer now. If I now see it right, at the EPO, the word "intermediate" seems to mean "a level of generalisation located somewhere between the various levels of generalisation (lying between the broadest claim and the best mode disclosed embodiment) already disclosed in the patent application as filed". Thus, every time the EPO uses the term "intermediate" generalisation it is tantamount to a declaration of an offence against Art 123(2) EPC.
ReplyDeleteI see that, as long ago as T1906/11, the Board simply confined its Art 123(2) enquiry to the binary question: does the amendment that was made in prosecution breach the Gold Standard Y/N? Good so! Daniel, you mention "intermediate generalisation" here on 2 October at 19:11. The thread might have been shorter if you had not used that term. But then I would have remained under an unfortunate misapprehension. Thanks for all the time you have spent, in this long thread, putting my understanding straight.