For the half-year to 30 June 2014, the IPKat's regular team is supplemented by contributions from guest bloggers Alberto Bellan, Darren Meale and Nadia Zegze.

Two of our regular Kats are currently on blogging sabbaticals. They are David Brophy and Catherine Lee.

Tuesday, 3 April 2007

Some exclusions are more equal than others


The IPKat's friend Edward Tomlinson of Frohwitter has pointed out a thought-provoking recent decision from the EPO Technical Board of Appeal (T 0471/05).

The decision relates to whether a method of designing an optical system could be entitled to patent protection. The claimed method in application EP98932446.2 included designing the optical system according to particular mathematical conditions specified by an equation in an independent claim. The claim did not, however, specify that the optical system needed to be made, nor how the design was carried out.

The application was refused by the examining division, because the subject matter for which protection was sought was the mere design of an optical system and encompassed "purely abstract and conceptual implementations excluded from patent protection pursuant to Articles 52(1), (2) and (3) EPC" (reasons for the decision 2.1).

The applicant argued that the method involved "technical considerations" and, according to T 619/02, had technical character.

The Board disagreed with this, because there was a presupposition in T 619/02 that the subject matter for which protection is sought should relate, at least implicitly, to "a physical entity or a physical activity" (reasons, 2.2). Although the claimed invention could be carried out using some physical means or resulting in some physical manufacture of the design, the claimed method did not include any of this and was therefore not restricted to a non-excluded activity.

The applicant tried to get round this by submitting an amended claimed that specifically disclaimed excluded subject matter. Even if such a disclaimer were to be allowed without introducing subject matter not present in the application as-filed, the Board considered that the main objection still held, i.e. that the claim was not confined to physical technical implementations.

A second request, which added the step of "using an optics design program" in the Board's view,

"defines an activity involving inherently and necessarily the use of such hardware fed with the optics design program and the claimed design conditions, i.e. defines a physical, technical activity" (reasons 4.1).

This request therefore overcame the Board's objections under Articles 52(1), (2) and (3) EPC, and such a claim would be entitled to patent protection.

The Board then considered the novelty and inventive step of the claim, finding both to be satisfied, and therefore decided that a patent should be granted on the request with the computer program step.

The IPKat thinks that this decision illustrates very well the odd-seeming logic that occassionally prevails at the EPO Board level. Although the decision is on the face of it in line with previous similar decisions (notably T 0453/91, enthusiastically followed by Pumfrey J in Halliburton and Cappellini/Bloomberg), it does seem to take the reasoning to its extreme. A computer program is excluded from patentability as such, as we all know. However, running a computer program inevitably involves "technical" means, i.e. a computer (following at least the Hitachi decision, T 0258/03). This can therefore, according to the Board's decision in this case, 'ground' an otherwise excluded invention in a technical area and make it patentable.

Presumably this same reasoning would not apply to the other named exclusions under Article 52. Or would it? Could it also be argued that the presentation of information, for example, necessarily involves technical means, given that pen and paper is sufficient (according to the Hitachi decision) to provide "technicality"? How about doing business? Is it possible to do business without any technical means?

Merpel says she is worried about the EPO Board being in danger of getting run over on zebra crossings.

36 comments:

Pieter Hintjens said...

What confuses me, the poor ignorant layman, is why the EPO goes through all these hoops to justify its practices.

Why does the EPO not just admit that it allows patents on software and business methods?

Is it perhaps ashamed to be involved in such a dirty business? Every patent is one more monopoly, and the EPO has been granting monopolies like state-run economies never went out of fashion. Does the economic cost to society never come into consideration? Does it just come down to angels on the head of a pin and a bunch of theologians who dissect the meaning of a word even though language is by definition an approximation?

Let's admit the truth: the law on patents lets one claim anything at all, if one can hack the terminology into shape. All pretense of "just cause" for patents is lost. Patent law is now a religion, where dogma and The Word rules over science, where the high priests hand out blessings to the rich, and where the Church and State are deeply intertwined.

Luke Ueda-Sarson said...

Presumably the EPO does not "admit" it gives out patents on software, because in its eyes, it doesn't - it gives out patents on software implementation, such as the one discussed here.

It seems to me that the problem is that patents are not be given out on programs "as such". But what exactly does that mean?

I fail to see why the, eg., UK practice of not giving out patents to software, even when implemented in a machine (and therefore surely no longer programs "as such") is any better than the EPO approach.

The UK approach has been that giving patents out on programs embedded in hardware is exalting "form over substance" - but isn't this precisely the sort of arbitrary red line approach that the EPC text calls for?

It seems to me, if you want to stop the EPO giving out patents on on "software" and business methods, etc, you need to tighten up the wording of the EPC.



Regards, Luke

Gerontius said...

Confused laymen should read the case law more carefully - and I think the IPKat should have too before commenting about how the same reasoning might apply to other exclusions - this was the same mistake made in Aerotel/Macrossan when the Microsoft decision logic was applied to music on a CD.

The claims were originally deemed not patentable because they consisted of nothing more than a design process. True, a device made according to the design would be a better device, but the mere design was a mental act or mathematical algorithm lacking any technical character. However, by adding the optics design program, an implicit technical character was added to the claim so the exclusion "as such" was avoided.

The IPKat essentially ends the story there, but that is misleading. Unlike the UK Patent Office, the EPO have upgraded the inventive step requirements in light of this downgrading of the exclusions.

Turning to inventive step, it was decided that the design method was novel and inventive because the optical system thus designed was a better optical system in a technical sense. This line of reasoning is independent of the fact that the design method could be carried out on a computer.

Consequently, such a line of reasoning could not be used to infer that business methods or presentations of information implemented on a computer are patentable - it has long been held that such things are not excluded as such, but the fact that a computer is involved does not assist in overcoming the inventive step barrier that would be thrown up against the non-technical business method and presentation of information.

We're still in line with Vicom here - computer programs are patentable when they implement a technical process or where the implementation of that process results in a technical effect. Although the reasoning is different, I still think the end result is the same which is the ery reason for the recent decision not to refer such matters to the Enlarged Board just yet.

The UK patent office (sorry, IP office) would probably have rejected the claims involving the computer program (since the design process itself is still non-technical and adding the computer program can't help), but I would think the device claims would still be fine in light of Halliburton. This difference arises only because of the current UK position that implementation on a computer does not provide a relevant technical effect.

It's interesting to note that that one word "relevant" could be the whole cause of the current divergence between the UKIPO and the EPO

Gerontius said...

Although it's just occurred to me that there's now a way to poke a hole in the EPO case law.

Performing the design method with a computer program was deemed sufficiently technical. However, in light of Hitachi, so would writing the design out with a pen on a piece of paper (possibly). Can't quite see how such a claim could be rejected, but I have a strong feeling that the Boards of Appeal would want to.

So, anyone out there who wants to take a case to the Enlarged Board - invent a process that is, itself, non-technical, but which has a technical effect when performed, and then claim that technical process written out on paper or the act of writing it out on a piece of paper.

Hmmm... Shame I don't have any clients who I think would be willing to try that or, in fact, any pending applications which would have basis for adding such a step to the claims.

David said...

Steady on, G. This Kat did think about it before commenting on the applicability of the reasoning. We are, of course, talking about EPO decisions, which are often contradictory in this area. I leave it to others to decide for themselves how far the argument can go before it becomes reductio ad absurdum. The argument about music on a CD is one way of doing this, but I think it is misleading, since the "relevant technical effect" argument deals with it quite well.

Gerontius said...

My point was that you have to go a lot further than the post implied to start poking holes through the case law and, as evidenced by the first reply post on this topic, over-simplification of the issue can lead to confusion among those who see software patents as a dirty business.

David said...

So, having taken your point, is my original comment about pen and paper re Hitachi not a potential hole?

Anonymous said...

In the Hitachi decision, it was stated that

"In accordance with the principles set out in decision T 641/00 (OJ EPO 2003,352; cf headnote I), the invention will
be assessed with respect to the requirement of inventive step by taking account of only those features which contribute
to a technical character. The features that make a technical contribution therefore need to be determined."

Since the only feature of "technical character" was the use of a computer program, following Hitach, I would have thought that this is the feature which would have to be considered for inventive step. Since implementing a given method using a computer would generally not be inventive, I have difficulty seeing how this Philips decision is in agreement with Hitachi.

EdT

David said...

I would have to disagree with you there, EdT. The claimed features of the invention in this case were seen as contributing to the technical character, in that they "allow[ed] for a relatively large volume in image space where abberations stay relatively low" (reasons, 4.3). The key feature is what contribution the features make, not the nature of the features themselves (which were purely mathematical). The problem with the application originally was that the scope covered any way of carrying it out. Keeping it to an implementation in a computer, in the Board's view, limited this scope to a technical field.
The decision does not therefore appear to be inconsistent with Hitachi.

Anonymous said...

David,

The Board, however, indicated that claim 1 of the main request (i.e. the method not specifying the computer program) did not meet the requirements for technical character. Technical character was only provided by the implementation using a computer program. Following the logic (?) of Hitachi, I would have expected the optical aspects to have been disregarded since these were expressly indicated to be non-technical.

EdT

Gerontius said...

David, my only problem with your post was the failure to discuss the inventive step implications. The reason I have a problem with such an omission is because the anti-software patent brigade do exactly the same when arguing that because a "computer-implemented thing" is "an invention", you can get patents on anything involving software when the fact is you can't.

Lobbysists conveniently forget that "invention" is only the first hurdle to get over, and when they don't forget it they say that the other hurdles are so low as to be meaningless. Actually, I've long felt that those other hurdles are far more useful because, as pointed out by Luke Ueda-Sarson, the invention hurdle is the least clearly defined (and probably least capable of being defined) one of the lot.

I just don't like to see the IPKat adding fuel to that particular fire.

David said...

All perfectly valid points, of course. However, on reading the decision I noticed much emphasis on whether it was possible to make the claim technical, and less on whether it was inventive. The 'first hurdle' was in fact the last thing to be discussed, and seemed to be almost an afterthought. This seems to be the consequence of taking the Hitachi line of reasoning. I don't see that my post adds any ammunition for the anti-patent lobby that was not there already in the decision itself. Any well-informed person can read the decision for themselves, and ought not to be swayed by any implicit bias in our arguments.

Gerontius said...

Leaving aside any concerns I have with the post, which David has carefully addressed, I've been thinking about EdT's comment.

In my recent experience before the EPO examining divisions, I think that they would agree with you, Ed: if something is non-technical it is non-technical and cannot contribute to an inventive step under any circumstances.

However, I think the BoA case law can be interpreted slightly more broadly (particularly in light of the warning in Konami) that a feature that may seem non-technical may, in fact, prove to be technical if it provides a technical effect.

It's all a matter of intepretation, and it will be interesting to see which side of this particular boundary the EPO finally end up falling on.

Hartmut Pilch said...

Hi Gerontius, the invention test (in short: does this claim teach us any new empirical knowledge in applied natural science?) is one of the most straightforward and simple tests that have ever been applied by patent courts. Certainly simpler than those strange creatures involving "inventive step" that the TBA use. And there is an even simpler test: that of whether the claim describes anything more than a program for computers, a.k.a. system and method using conventional computing equipmennt, a.k.a. computer-implemented (non-)invention.

This is btw the position that was adopted by a large majority of the European Parliament at 2 occasions . Why do you want to avoid adding fire to it? Do you have a vested interest in software patents? Or has the EPO logic twisted your mind to the extent that you can't see the neat simplicity of the EP's (and original EPC's) position anymore?

Hartmut Pilch said...

Gerontius, the tests which the "anti-patent crowd", including the European Parliament and the German patent courts of the 1980s, wish to see applied, are among the simplest and most easy to apply that have ever been used in the patent system:
(1) does the claimed object teach us anything new in the field of applied natural science?
(2) does the claimed object consist of more than a rule for operating a conventional general-purpose computing system?
Rule 2 can be applied mechanically for sorting out program claims and almost so for sorting out "system and method" claims. Rule 1 may be needed in some difficult cases and requires some human thinking, but still leads to more predictable results than anything I've read from the EPO about "inventive step" so far.
Why do you want to avoid reinforcing this simple message? What is there for you to fear?

Gerontius said...

Helmut, the falsity of your suggestion that it is easy to determine whether an invention teches something in the field of applied natural sciences is perfectly illustrated by this very decision we are talking about which is ultimately concerned with designing an improved optical apparatus.

Other "twisted thinking" in your reply:

1. Your position is one that was adopted by a large majority of the European Parliament at 2 occasions

Absolutely not. There was never much consensus at any level of parliament. Hundreds of contradictory, and often self-contradictory amendments to the directive were submitted by all quarters. Some of which may have supported your interpretation but most of which, including the final proposed text, were quite vague.

2. The EPC supports your position.

I'd be interested to know where the "applied natural science test" is provided by the EPC any more than the technical contribution test. At least the word "technical" was mentioned in one of the rules to the EPC

3. That your interpretation is the only valid one.

That shows more arrogance than anyone else entering this discussion has shown. At least everyone else recognises the inherent difficulties in interpreting the EPC and the subsequent case law. Personally, I'm not interested in trying to interpret the EPC - not only can we never know what the framers intended, but technology has changed so radically since 1976 that their intentions are irrelevant.

Instead, my job is to interpret the current case law. That's my vested interest: to be able to advise my clients in a helpful and responsible manner and thus keep my clients. What's yours?

Anonymous said...

OK people are getting "technical effects","technical problems" and “technical ....“ mixed up. It's not helped by the EPO using the phrase "technical character" inconsistently but the main test is actually very simple. I suggest people look to Comvik which established the new approach and is well written, rather than Hitachi which confirmed the test.

Contrary to Gerontius's otherwise excellent posts the EPO did not “upgrade” inventive step (though it did change most people's perception of it).

The test has two completely independent steps.

1. In order to be an "invention" the claim cannot be completely abstract and must contain one technical feature. Use of a computer will do.

2.In order to be inventive the claims must pass the problem and solution approach, which is:
a. what is the closest prior art
b. what are the differences between the claim and the cpa.
c. what are the technical effects of those differences
d what technical problem has been solved in light of those effects
e,. would it be obvious to modify the cpa to solve the technical problem by incorporating the differences identified in b.


That is the same test that it has always been only people didn't used to take as much notice of the two mentions of “technical”.

The case law says that when doing step e you ignore any differences which don't make any contribution towards solving the technical problem of step d. This isn't new and was stated in several cases (mostly chemical) before Comvik as the reason to ignore arbitrary parameters that don't do anything useful.

Note the case law does not say “ignore all excluded subject matter” as Jacob and Pumfrey wrongly said, nor does it say “ignore non technical features” (as gerontius implies and even a EPO examiner has to me), nor does step 1, and which features were or weren't used to overcome that hurdle, have any relevance at all to step 2 (as EdT suggests). An algorithm is non technical but frequently contributes towards solving a technical problem.

[There have bee a few new embellishments to inventive step like use of claimed features not in the prior art to formulate the t problem (Comvik) and solving not obviating the problem (Hitachi) but they don't apply most the time]

Anonymous said...

Three cheers for that last contribution. PSA is supremely elegant but nevertheless sold short, all over the world, because no patent litigator deems it worthy of his serious study. I had been wondering how to reconcile it with CII problems. Now I know.

Anonymous said...

H Pilch can fool himself that his "applied natural science" test works because it is not the basis of real decisions emerging from courts in Europe today. But you only have to read Greg Aharonian's Patnews blog to see what would happen if ever it were. People like Mr Aharonian have no difficulty whatsoever getting any business method claim they want, under that particular test. The EPO's problem and solution approach invokes "technical" (which is itself a red rag to Mr Aharonian) but we see from Anonymous immediately above that PSA when correctly carried through does actually do the filtering business effectively and in line with common sense. It therefore safeguards respect for the patent system from non-specialist business people - a non-trivial achievement in these days of popular but ill-informed attacks on the patenrt system. Sad for PSA is that no litigator needs to bother to try to understand it. Since most of the learned commentary written for publication comes from litigators, what's published is almost invariably ignorant or even derogatory of PSA. No court can dictate to the EPO what tool it shall use to determine what is obvious. Thank goodness. Otherwise we would never have been given the stunningly elegant gift of EPO PSA which, to this day, goes unappreciated by the judges and the litigators. Even in THIS column, we STILL need anonymous to work it through for readers.

Hartmut Pilch said...

Dear editor, why are you holding back my posts while letting Anonymous's stuff, that refers to me, get through? To answer Anonymous once more: the test that I'm advocating is a test that has been proven by a lot of caselaw in the 80s, and the patent applications on which it was used are no different today than at the time. And Greg Aharonian's reasoning is certainly no challenge for them. If you look at http://eupat.ffii.org/players/aharonian/, you'll find pointers to some pages in discussion forums where Aharonian's bluffs have been dissected by many people. If you take Aharonian's rants about "modern science" in which "information has become a force of nature" and the like seriously, then you just discredit yourself.

Anonymous said...

Anonymous, the PSA may be intellectually appealing in some ways, but
(1) it results in the systematic granting of lots of trivial software and business method patents
(2) it makes the written law (EPC) irrelevant and unintelligible, whereas the methods of the 1980s that FFII and the European Parliament advocate make sense of the written law.

David said...

Apologies if your posts are not getting through straight away. The IPKat does not censor or hold back anybody's views. He has, however, had a problem with spam, so recent comments have needed to be approved by a Kat before being posted. This sometimes results in a bit of a delay, particularly if all the Kats are out playing and forget to check the blog frequently. Moderation has been turned off for now, so your posts will appear immediately. The IPKat reserves the right to turn it back on, if spam starts appearing again.

Rodrigo Calvo said...

I see several of the above contributors (the Usual Suspects, I'd call them) making once and again the point that the European Parliament somehow endorsed their point of view.

Let me be clear about this: AFAIK, when the CII Directive was discussed, the European Parliament ultimately failed to endorse ANY point of view. The CII Directive could have been amended in the EP to be as restrictive as Mssrs. Pilch and Hintjens would have liked it to be, and then they could have reasonably claimed a triumph. Instead, the final amended compromise proposal put to vote was so unsatisfactory for both the pro and anti camps that both sides rejected it, and the best possibility to draw a clear line in the sand between patentable and unpatentable subject matter was missed. Considering their role in scuppering the best chance there was to clarify the legislation, I would think that FFII could do better than to criticise the EPO Boards of Appeal for their decisions when they have to base them on the unclear provisions of the EPC.

Hartmut Pilch said...

Mr. Calvo may not know the history of the Software patent directive and the rules of the EU's Codecision procedure very well.

In June 2005 the European Parliament did not have the option of amending the directive. It merely had the option of entering in another round of negotiations with a Council of Ministers, staffed by the usual suspects from the EPO's administrative council,
who had already signalled many times their unwillingness to even discuss anything that deviated from the "established practise" of the EPO, i.e. the doctrines of the BOA that we are criticising here.

At the same time, the Codecision procedure makes the hurdles for the Parliament high. Thus, although there was a solid majority, around 60%, behind the 21 Amendments tabled by M. Rocard and backed by the whole left and half of the right wing, the risk of missing the required 50% of *all* MEPs (with the absent ones counting in favor of the Council) was still considerable. Thus, when the pro-software-patent camp suddenly, around 2005-07-05, proposed rejection of the directive project, it was not reasonable for majority to suddenly become a champion of this project.

The story about the MEPs rejecting the directive out of confusion, which Mr. Calvo is spreading here, seems to originate from MEPs Toine Manders and Sharon Bowles. Manders was the first to spread this version in his press release, and Bowles is sticking to it now. It may be true from their point of view. This is because to them, the EPO has always been the only source of clarity, and when their own party group, ALDE, broke away from their guidance, that could of course only mean that ALDE was plunging into confusion. Manders and Bowles were among those who would never allow even the possibility that any alternative to the EPO's practise could ever be admitted into the discussion. And thanks to the power of the EPO people in the EU's Council of Ministers, they were able to uphold that view through 2005-07-07 until this day.

David said...

It has been said here before, but it is worth making the point again in this context. If people above such as Peter Hintjens and Hartmut Pilch(who are, incidentally, president and treasurer respectively of the FFII) are to have any chance of properly addressing what they view as the problems of the European patent system, they should come up with a workable solution rather than simply raising conspiracy theories and negative propaganda, neither of which gets us anywhere. I have not yet seen a viable solution that is better than the current EPO problem & solution approach, but perhaps someone could point one out to me.

Hartmut Pilch said...

David, what approach is more positive than that of the "21 Cross-Partisan Amendments" of the European Parliament, or the "10 Core Clarifications" and other explanations of the simple and practical rules of patentability of Art 52 EPC on http://eupat.ffii.org/ that explain them. If we had come up only with "negative propaganda and conspiracy theories", we wouldn't have been able to have an impact on the EU's legislative process regarding software patents. As I explained in several postings that you didn't allow onto your forum, the classical interpretation of Art 52 EPC, as practised in the early days of the EPO, was clear and adequate and still would be today, as can be verified by applying it to sample patent applications. The EPO has not mandate from any legitimated legislator to do anything else than apply those simple and adequate rules. And it's not too late now. Ms Brimelow could refer the case to the EBA if she wants.

David said...

Reading the 21 Cross-Partisan Amendments, they don't seem to be very workable or viable to me. Others can easily see that this is the case without me having to go through them in detail.

As for negative propaganda and conspiracy theories, one only has to read what your president says above (see the first comment) to get a flavour of this.

To repeat myself (see above), the IPKat does not censor comments. Your comments have never been blocked. The more likely cause is a fault with either Blogger or your own software. I would be very interested to see what you view as being the "classical interpretation" of Article 52 EPC, and certainly would not want to prevent other readers from having the benefit of it.

Hartmut Pilch said...

Art 52 EPC is a concise and elegant way to state what could be elaborated in the following 2 slightly more explicit rules. The first was elaborated in detail by the EPO's first Examination Guidelines (of 1978), the 2nd by the German courts:

1. Claims to "program, characterised by that it ..." or equivalent claims like "system and method for processing data on a general-purpose computer, characterised by that ..." can be rejected outright, without looking at the content.
2. When it comes to compound claims , such as "computer-aided car braking system, characterised by ...", you can't mechanically reject but have to see whether this claim discloses a new teaching a field of applied natural science (technical teaching). The new part must be "technical" in this sense, and the "technical" part must be new.

Rule 2 is implicit in Art 52 EPC, because the listing of 52.2 is not exhaustive and thus needs an auxiliary theory to support it, and the concept of "technical character" is a kind of customary law, well grounded in the philosophy of science, that has always been used to this end. It is quite elegant of the lawmaker to not explicitly state the theory.

The 21 Cross-Partisan Amendments are more verbose. You normally wouldn't write a law like that. But the EP was forced to express itself in this way, because it's only way of influencing the process was by amending a given text, which was itself quite bloated and hardly a masterpiece of elegant law writing. I think under the given circumstances the EP produced astonishingly coherent and workable counter-proposals, even in 2003-09-24. Incidentally I've talked about them with some judges and patent law scholars who said that they wouldn't have any trouble applying what came from the EP, it was fairly clear.

Btw I'm attending the EPO's EPF event today and tomorrow, are you there? After seeing today's "4 future scenarios" performance I can't help feeling that the future of the EPO would look much brighter if they (i.e. the EU Council WG) had seized the opportunity and helped the EP work out a nice clean text along the lines of the EP's intentions, which is to restore the clean lines of 1978. The EPO might have up to 40% fewer patents to grant then, but it would also have much less trouble and instability. It was apparently misled by the fear that patents would be marginalised in a "post-industrial" "knowledge society". An exaggerated fear, and marginalisation is better than dysfunctionality and collapse. The Dispositionsprogramm decision of 1976 expresses that wisdom very nicely already: The patent system is not intended to be a reception basin for all emerging unprotected ideas but a specialist system for a special area, that of applied natural science, maybe not as big in the future as in the past, but also not small even today.

David said...

Many thanks for the clarifications. The only difference that I can see with how your 2 'rules' differ from how the EPO works now is that computer program claims can, under certain circumstances, be allowed (following T 935/97 and T 1173/97). The second of your rules is indistinguishable from how the EPO problem-solution approach works now, as far as I can see.

Could it perhaps all come down to whether the law should state that computer program claims should not be allowed (or perhaps should not be enforceable)?

Hartmut Pilch said...

David, indeed the EPO has not completely abandoned rule 2, and the biggest rupture came, as you say, with the IBM 1 and 2 decisions of 1998. However, the EPO has been stretching rule 2 so much that they had, for the sake of consistency with their stretching, break rule 1. Which means that their stretching violated rule 1 in the first place, even before IBM 1 and 2, starting with Vicom in 1986.

The methodology of Pension Benefits and Hitachi, which is the subject of this thread, has in fact added a third rupture by saying that anything computer-implemented is an invention in the sense of the law. When it became apparent that this became politically hazardous (business method patents were largely welcomed by the patent community until 2000), what used to be the invention test of rule 2 crept back in, as part of the "inventive step" test, but in a rather non-commital, highly politically oriented way. The Pension Benefits decision was, like the IBM 1&2 decisions, closely connected to the ongoing legislative efforts of the time. But writing rules of this kind into the law was very inelegant, disliked even by some well-known pro-software-patent lawyers as far as they were still concerned about the integrity and elegance of the legal framework. It was quite easy for us to build a big coalition against this kind of lawmaking, much easier than obtaining support for the 21 Cross-Partisan Amendments.

The "problem-solution approach" itself is not a problem. Indeed "rule 2" describes a problem-solution approach. It is always the solution, i.e. the new part, that should be called the "invention" and that must conform to the requirements of patentability. This has traditionally been the common sense of the patent system, and most people still thought like that even during the directive debate. The EPO's reasoning, which really consists in giving up the problem-solution approach and then claiming to be transposing it, in some hard-to-understand way, into the inventive step test, posed great communication problems to the patent lobby, and it is still posing those problems today, as this thread witnesses. Deservedly so, I'd say. You do well to stimulate a discussion on your blog where they face the whole backlash of the awkward contradictions which they themselves created, hoping that a directive would quickly clean them up for them.

Gerontius said...

I think I understand why Helmut thinks this "new teaching in applied natural science test" is an easy one. He is equating "new" with "novelty" as defined in Art 54 EPC. I don't know about Germany, but in the UK, the history behind such a test is far from straightforward.

To sum up. It all started with Vicom. There, the EPO said that what mattered was the technical contribution. This was taken up by the UK courts with lots of references as to what was "new" about the invention and that this must not be excluded - this is effectively the test that Helmut is proposing.

It wasn't explicitly stated, but a careful reading of the early UK judgments shows that they didn't mean "new" as in "novel" (I'll come back to this point when we reach Aerotel).

Peter Prescott in CFPH overlooked this subtle distinction between new and novel when he started talking about the UK test involving looking at what is novel and that this wasn't enough - you also had to look at what was inventive. He thus defined the "advance" and said that this advance must be non-excluded. This approach, however, was equivalent to "Falconer" reasoning (according to the UKIPO) where the excluded bits of a claim could just be ignored. But, according to the Court of Appeal, such an approach would result in many good and valuable inventions being denied a patent. You have to look beyond the wording of the claim at the "substance" of the invention.

Consequently, when the court of appeal considered the issue again in Aerotel, they dispensed with the advance and re-established the distinction between "newness" and "novelty" by saying that what matters is the "contribution". This contribution being a vague concept about what the inventor has contributed to society as a matter of substance. One way of looking at this is in terms of the "effect" that an invention has. The effect of the invention is part of the substance of the invention and if processing data on a standard computer has a real-world technical effect, then good for the patentee!

So, even if you ignore the question of what is meant by "applied natural science", the problem with your test, Helmut, is that working out what is "new" is an extremely difficult past-time. The EPO have therefore rightly moved away from it, in my view since it can never provide people with certainty over where the boundary lies. You may not like the current boundary set by the EPO, but it's a clearer boundary than the one defined by your test as evidenced by a decade of variable UK practice since Fujitsu.

I just want to make one other point about my saying earlier that the EPO "upgraded" their inventive step test, which has been disputed. I meant this in relation to the UK where business method features can be novel and inventive. So, in the history of the EPO there may not have been an upgrade (to be honest, I'm not sure of the history on that point) but there is certainly a clear distinction between UK and EPO practice on this point.

Hartmut Pilch said...

Gerontius, by saying "the new part is the invention", i didn't mean to refer to any legal criterial of novelty. As soon as words are mapped to a particular examination procedure, a process of formalisation starts that devoids those words from meaning and in the end leads to the discussion about "substance" which you mention.

No matter how you translate "new" into an examination mechanism, what I meant to say is that the "invention" is not the same as the claimed object. It is not enough it the claimed object contains some technical features or technical effects, let alone technical problems. What is important is the solultion, the contribution, or, in more colloquial terms, the new part. There is no formalism by which you can conclude from a claim wording what the invention is. "Rule 2" (of the 2 rules in which I summarised Art 52 EPC above) is not a formalistic one. Although I wouldn't say it's difficult to use, it certainly requires more analytical capabilities on the part of the examiner than rule 1. The question to be asked is not what kind of features the claim lists and whether these fall under "technical", but, what is the knowledge, the teaching, that this claimed invention contributes to the state of the art. This teaching should be in the field of applied natural science, or, in one famous, very consistent and very strict interpretation that one support of the EP in the 2003-09-24 vote, should contain new insights into cause-effect relations of controllable forces of nature.

I think there has never been much of a problem of lack of clarity of this understanding of the law. The problem was, rather that, as you quote the UK Court of Appeal saying, "such an approach would result in many good and valuable inventions being denied a patent", i.e. in the view of a large part of the patent community this approach was consistent and understandable but not adequate. We of course say that it was both consistent and adequate. Adequate because it denied broad monopoly protection to those kinds of advances that don't require laboratory experimentation, that people mostly come up with simply by logical thinking, and that tend to lead to abstract functional claims, where a problem is claimed and a huge array of solutions is forbidden, as is typically the case in business method patents. Thus the strict version of 'rule 2' is optimal, regarding both clarity and adequacy, but there are many less strict versions of it that are still compatible with 'rule 1'. Vicom is probably not compatible with 'rule 1' anymore, and Sohei certainly isn't. Which is why the EPO's TBA in 1998, 2 years before the anticipated legalisation of software patents, thought it couldn't wait anymore to officially break rule 1.

Gerontius said...

I think now I understand where you're coming from, Helmut, but I'm not sure you're quite getting what I'm saying because you appear to be contradicting me even when we're saying the same thing.

First, I now realise that you are intending "new" to mean essentially what the Aerotel CofA decision means by new. Therefore, current UK pratice appears to be almost exactly in line with your two rules and your interpretation of them.

The Falconer approach that I mentioned, and which had the effect of denying too many good inventions a patent, is the "formalistic" approach, that we both agree is wrong but is, in essence, very simple to understand and apply and therefore appealing to practitioners. This is the approach where you only look at the language of the claim to determine what is "new". In fact, the formalistic approach was said to be wrong for exactly the opposite reason that you suggest: it denied protection to those kinds of advances that DID require laboratory experimentation, such as inventions resulting from biotech discoveries.

So, the basic point of contention between us is whether it is simple to apply the "new" test, when "new" is used to mean the contribution provided by the claimed invention, without a formalistic application of the word "new". I say it is not easy and the CofA in Aerotel seem to agree with me. You think it is easy. We'll have to agree to disagree there.

The EPO agreed with me that it was fraught with difficulty, too, and therefore developed their new test which, from my perspective, is far easier to apply and I have a much easier time advising my clients in Europe as to whether or not their inventions are patentable. Please do not misinterpret this to mean that it is easy because I can ALWAYS say to my clients that their invention IS patentable, because I regularly have to give them bad news. But I can give them that bad news (or, conversely, good news) with a greater degree of certainty than I can in the UK.

In seeking to find the similarities between the different approaches, I think the EPO's test can be seen as a more practical and rigorous approach for determining the substance of the invention - although, as the EPO case law has developed, I think they have lost sight of that on occasion. As an aside, that is why I think a good EBA decision on the matter might be useful at this time (NB I never said the EPO approach was perfect, just better than your proposal or the UKIPO's practice). In developing their test, the EPO abandoned your rule 1 because they saw that, if you're looking at the "substance" of the invention, it is illogical to consider the formalism of whether the invention is expressed in terms of a computer program, a machine readable medium, a computer, or an apparatus.

Of course, a degree of formalism when it comes to claim language will always be important and necessary, but finding the right level is important. I think Pumfrey's recent "tethering" comments might have something to them on that front, if only they could be expressed more clearly, but your rule 1 is too high a level, in my view. The UKIPO currently appears to agree with you, but the courts have been pretty silent on the matter so far.

It will be interesting to see where this goes over the coming months as there are still several UK High Court judgments outstanding that address these issues.

Andre said...

Merpel says she is worried about the EPO Board being in danger of getting run over on zebra crossings.

It happened before and that's the anarchy of the BOA and its amateur "technical" judges: nothing happens. They decide cases and prejudice future patent law and no parliament can ever stop them. They brought us CII - without any economic evidence that patent law was appropriate or wanted for this sector. Of course BOA just decide single cases and take their freedom to obstruct patent law.

They brought us "further technical effects" and no one is able to revise their decisions. So the solution is checks&balances. BOA do not qualify to set case law or decide substantive matters. Examination guidelines need to be approved by Parliament. As procedures for Europarl review of patent policy are missing, the upcoming EU core constitution nseed to address the matter of democratic patent policy and put the EPO under political control. If Switzerland was a problem, then Switzerland and Turkey need to get phased out.

It's time for accountable patent policy.

David said...

Merpel tells me that she is a fan of the late Douglas Adams' writings. She fails to see what points you are making otherwise, Andre. Were there any?

Anonymous said...

As an observer to this discussion may I state that Gerontius could do well to start his points by writing name of the person with whom he is debating correctly.

Subscribe to the IPKat's posts by email here

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