McDonnell Boehnen Hulbert & Berghoff LLP

« Make a Backup Copy in Your Own Home: Allowed Only in France | Main | Nipper on Patent Reform For Small Entities »

May 13, 2005


Feed You can follow this conversation by subscribing to the comment feed for this post.

These professors apparently haven't had to deal with examiners arguing that it's obvious "because I say so and I find the suggestion to combine in the general knowledge in the field." And, this after you have just explained to them something like how a transistor works or the difference between a compound and a mixture.

It's about time the Supreme Court took a look at how horribly the CAFC is misconstruing 35 USC 103. The way it stands according to the CAFC's interpretation every slight variation of an invention needs to be documented in the prior art and explained on a grade school level, removing all meaning from the phrase "ordinary skill in the art". CAFC's mishandling of this statute has trickled down to the Board of Patent Appeals and Interferences and to the Examiner level resulting in the ability of Patent Examiners to use any common knowledge of the art being stripped away. This has resulted in an increasing number of patents relying on trivial features often being originally presented as one of many dependent claims having no apparent additional benefit or advantage over the invention covered by the independent claim. As a former Patent Examiner of over 4 years who also previously worked as a technician in the field that I was examining I can tell you first hand that the obviousness standard applied by the Patent office is a far cry from that held by one of ordinary skill.

Agreed with the above. At least with the "teaching/motivation/suggestion" criterion there is some semblance of a standard for examiners to follow in assessing obviousness. Removing that simply makes an obviousness determination even more examiner-dependent, and it is clear in my experience that there is little, if any, consistency between examiners in making obviousness rejections. If it is the law that an examiner must state specific reasons for an obviousness rejection, then a simple, hand-waving reference to what one of skill in the art would find obvious, alone, should not suffice.

There are really two things that must be considered from a policy point of view.

First is having some sort of reasonably objective standard instead of the frequent "It's obvious because I say it is" that examiners use. At least the current standard is more objective than subjective, but the professors seem to want to change that.

Second is the notion of inventive concept. Just because an inventor sees two somewhat related pieces of prior art and goes, "Aha! If I take idea #1 from A and idea #2 from B and combine them, I'll have solved my problem!" doesn't mean that an inventive concept didn't come into being. Why shouldn't that insight be patentable? Especially if no one else had that insight before.

Plus, the examples in the professors' brief are frankly embarrassing. The idea that you could combine a new storage medium with an old microphone (assuming you're not the inventor of the new storage medium) is already handled just fine by the existing standard. If "CD + microphone" is known, then "DVD is an improvement over the CD" (which is not exactly difficult to find when a new technology is introduced) is sufficient as a motivation to modify the prior art.

If these things are so widely known, they shouldn't be that hard to document. The professors suggest that "patents and academic publications" are the only sources of art available to the USPTO. First, I don't think that's true, and second, why not give the USPTO better access to the art, instead of watering down the standards for invention?

The problem with the non-obviousness standard as enforced by the CAFC is that it raises products and processes which are mostly matters of design choice variations from what is found in the prior art up to the standard of patentable inventions. While I agree that there should be a clear standards in determining patentability and effort should be made to assure that Patent Examiners apply these standards with uniformity, this does not mean the standard should be set so low. The patent system seems to me a delicate balance between public and private interests. The low standards currently set by the CAFC (and BPAI and Patent Office by extension) only hurt the public and devalue the truly innovative patents.

The professors are apparently unaware there is an elephant in the room. The point made above about the professors not having dealt with PTO Examiners is exactly right. Patent practitioners are keenly aware of the variability of Examiner office action responses and the lack of PHOSITAs at the PTO or anywhere else. Why is this so, besides PHOSITAs theoretical nature? As for the PTO, here’s my guess, for which I have little evidence, but certainly more evidence than the professors put forth quantifying the harm and expense to our patent system of having a “low” objective obviousness standard as compared to abandoning the “now-mature suggestion test.” In the art areas for which I prosecute patents, persons whom I deal with approximating true PHOSITAs receive more than twice the compensation as PTO Examiners. It is unfortunate for our patent system that the most knowledgeable technologists are not likely to be employed with the PTO. Unlike the professors, the Federal Circuit judges must have read a lot of office actions and responses in the records that come before them. If so, they certainly understand that an objective obviousness standard is more important to everyone’s dockets than allowing PTO Examiners to subjectively channel theoretical PHOSITAs. I’m not saying there is not a problem with obviousness in granted patents, but it does not follow that blowing away the “suggestion test” will result in fewer problems in the system.

And one more thing—the professors say: “Thus, it would appear that that the Federal Circuit’s “suggestion test” would have led to the opposite result in Graham itself.” Anyone tempted to accept their statement should read Graham at least from 383 U.S. 30 to the end. The Supreme Court says, “The Scoggin invention, as limited by the Patent Office and accepted by Scoggin, rests upon exceedingly small and quite nontechnical mechanical differences in a device which was old in the art. […] To us, the limited claims of the Scoggin patent are clearly evident from the prior art as it stood at the time of the invention.” [383 U.S. 1, 37] But don’t take my selective citations over the professors', read the whole thing. The Supreme Court did not accept the appellate court’s finding that there was “nothing in the prior art suggesting Scoggin's unique combination of these old features.” Just because an appellate court made an incorrect finding doesn’t negate the standard they attempted to apply. The result in Graham is completely consistent a “teaching, suggestion or motivation” test.

I really want to know if any of these professors has ever filed any patent application at all. If any one of them did, he/she would surely know the pain to fight with an Examiner when "obviousness" was used as an execuse to decline an application. Every examiner has every incentive to deny a application, "obviousness" is an effective weapon and is so easy to abuse. Without a objective standard, "obviousness" is a small inventors nightmare that is very costly both in terms of legal fees and energy. The teaching/motivation/suggestion is at least an objective standard the examiner must follow.
I urge these professors while object to CAFC's position, come up with a solution that provides "a objective standard". Or at least come up with an idea, appling for a patent and try to deal with PTO examiners.

I'm sure many inventors and patent attorneys are really frustrated by these obviousness "excuses" that Patent Examiners use to reject their patent applications. I'm sure they also find those pesky anticipation "excuses" troublesome as well.

All joking aside I agree that the US Patent Examiners are often not ones of ordinary skill in the art, resulting in incorrect rejections often being made. As a former Patent Examiner myself, I was surprised to find out in my first few weeks on the job that I was the only one in my art unit who had previously worked in the industry I was examining. But this is an altogether different problem from the low obviousness standard held by the CAFC and two wrongs don't make a right. As for Patent Examiners having incentives to reject applications, this is actually quite the opposite of reality. Anyone who is aware of the way Examiners are credited with their work will be aware that the system is set up to promote early allowance. The average Primary Examiner usually needs to allow 1-3 applications a week to make production and no credit is given to the time taken in writing final rejections. Based on my experience usually 60-80% of applications get allowed (this rate is probally >90% if you count RCEs).

In my experience the explicit suggestion test used by the CAFC does not meet the reality of obviousness to one of ordinary skill because everyday technicians (I used to be one in the printing industry) will often swap out parts or rearrange elements for often very trivial reasons (such as one part needs to be specially ordered while an equivalent part is readily available). These types of modifications are common and not really documented in patent literature that Examiners employ.

Perhaps a better standard than a mandatory suggestion test would be to balance the suggestion of the prior art versus the purported merits asserted by the applicant. I will illustrate with the following example:

An inventor presents the following three claims involving elements A,B,and C:
The inventor explains in the specification that Element A provides some advantage. The inventor further explains that Element A combined with Element B provides some further advantage. The inventor is silent as to any additional advantages provided by A+C.

An Examiner finds three references separately disclosing elements A, B, C in a common art. It is apparent that the elements are easily combinable however there is no explicit suggestion for such combination found.
1.A (anticipated)
2.A+B (nonobvious+inventive)
3.A+C (nonobvious+noninventive)

Under the explicit suggestion test both claims 2 and 3 would have to be allowed. However, does the inventor truly deserve exclusive rights to claim 3? Lacking an explanation on why this combination adds value to the art does the inventor really deserve such rights? What are the costs to the public for awarding such rights?

The patent system exists to benefit the public and not just patent holders. In order for the system to work there has to be a balance between the rights of these two groups. Hopefully future legislative and judicial efforts will help restore such balance.

The people who think obviousness standards are too high haven't looked at the software industry. Here, total trivialities are routinely the subject of patents.

This is to a great extent due to the patent examiners' not understanding the subject. It is also largely due to the examiners' failing to do proper research into prior art (in programming, the prior art is usually not published in dead tree journals or patented).

It is also due to the fact that software patent language is allowed to be frighteningly overbroad, including things which weren't really invented, but are merely conceived of. (Restoring the old requirement of providing a working model would help some here.)

But it is also due to allowing patents on A + B unless there is an explicit "suggestion" of combining them in the literature -- the mistaken rule under attack here. In computer programming, virtually anything can be combined with anything else, and routinely is. This is not recognized by the patent system.

Here's an attempted analogy: Inventor invents a hammer and describes the process of using it to hammer in a nail. He sees no need to patent it and does not.

Then Mr. Crook decides to patent the process of using a hammer to drive in a dowel. Because there was no written documentation of the combination of the two, the patent examiner would grant this. However, it's actually totally obvious to anyone who's done a little woodworking, and certainly shouldn't be patentable. (Next Mr. Crook patents hitting people on the head with a hammer. This is not a combination of anything, but is also totally obvious.)

This sort of patent gets granted all the time in software. See the various patents on using a stack to handle this, that, and the other thing -- once you've invented the stack (which was done, without patent, before electronic computers), they're *all* obvious.

In software, any existing algorithm or data structure can be applied to any existing situation with no real creative step -- it's obvious to any programmer -- and the current standards do not recognize this. (Whether the application works well or not is another matter, of course.) This is not true of most traditional physical problems. (Though it is of some of them; "a wooden cabinet with drawers to store tools in" is obvious once "a wooden cabinet with drawers to store clothes in" has been invented.)

(The existing incentives to write software indicate that software should not be patentable, on utilitarian grounds. But even if software is to be patentable: only a genuine "industrial process", i.e. an algorithm (process), which has a direct functional purpose (industrial), should be patentable; the algorithm should have to be genuinely non-obvious, not merely an application of an existing algorithm, or a nonfunctional tweak to it; and only the specific algorithm should be patentable, not "any method of doing X", which is the way most software patents are actually written these days. *All* of these conditions are currently routinely failed by patents granted.)

Contrary to suggestions above, actually using the standard of a "person of skill in the art" would *not* be an arbitrary excuse for the patent examiner to do whatever he wanted to -- at least not if implemented correctly. The patent examiner often *isn't* such a person, but the patent examiner can and should consult such a person (or perhaps several of them). Perhaps opinion letters from such people could be required in order to declare obviousness; in truly obvious cases this would be really easy for the patent examiner to get, and in truly nonobvious cases it wouldn't.

The comments above about software patents are right on. It is very clear that a standard tactic in the industry is to patent anything and everything you can think of, in the hopes that somehow someone at some future point will actually produce a practical implementation which can be construed as infringing upon an existing patent, and thus allow the company in question to "cash in" on someone else's geniune innovation in *implementation*. These patents create a "moral hazard" to innovators, and provide no geniune value to society.

Several writers above complain about the arbitraryness of examiners pretending some invention be obvious "because I say it is". These writers seem to me to resemble people who complain about the variable outcome of measurements using a yardstick. If instead all lengths were declared to be exactly pi meters (i.e. pi=3.1415...), then all measurements would agree exactly no matter who did the measurement, to an infinite number of decimals.

Even if determining non-obviousness is nontrivial and likely to be fraught with variable outcomes, the sugestion requirement translates to a zero-level universal obviousness declaration. Suggested combinations are simply non-novel, anticipated by the suggestion. Motivated combinations are hardly better.

The only question I have about changing the standard is how does one go about this and still maintain an objective standard? The key to the current standard is that it is evidentiary. If one tries to import some kind of "common sense" notion in to the determination, we are back to a subjective standard, something the 1952 reform was trying to get rid of.

The 24 professors write in their amicus brief "Inventive processes typically involve judgment, experience, and common sense, capable of connecting some dots. The suggestion test, rigidly applied, assumes away PHOSITA’s typical levels of creativity and insight and supports findings of non-obviousness even when only a modicum of additional insight is needed."

The suggestion here is that the court, sitting in the PHOSITA's position, should apply "some modicum of insight" above and beyond the documentary record to decide what is obvious, in effect connecting the dots with "insight" that the professors claim is many times lacking in the documentary record.

This is simply a return to the "flash of genius" test, in disguise, that section 103 (a) was designed to get rid of. Just how much "insight" does a PHOSITA have? The professor's suggestion for a fix is not good.

It seems like the real issue here is whether or not the suggestion test permits aggregations to be patented. If so, then there is something wrong. However, section 103(a) refers to whether "subject matter as a whole would have been obvious." If the court were to place more emphasis on the phase "as a whole" then aggregations would be ruled out.

That is what the Sakraisa decision is all about. The combination united old elements with no change in their respective functions - it was not a combination but an aggregation - and thus not patentable.

I believe that the wording of the statute includes this test - it just has to be articulated more clearly by the CAFC.

These comments are offered by an interested Englishman.

There was only one way that the decision could have gone in Graham v John Deere. The patentee's expert admitted that the difference relied on was unimportant to the working of the device. The decision quotes in one of the footnotes four lines of cross-examination that are a little gem - the surprise is that with such a damning admission on the trial record, the case went anywhere near the Supreme Court. Furthermore the patent in issue gave no indication that there was anything important about the reversal of components which was dredged up at trial. If you look at the Andersons-Black and Sakraida decisions you will see that they also involve similar uninspiring disclosures that do not say anything positive in support of the features relied on at trial. The real reason why these decisions went against the patentees was that the features relied on were not put forward in the application as filed as being of an inventive character, and when the patentee's lawyers subsequently tried to argue that thet were, the courts discelieved them. All the high-faluting rhetoric about Thomas Jefferson and about the onerous task of establishing invention in a claim made up of old elements is mere judicial hot air.

The present case is not like that. The point of the invention is to produce a simpler and more compact design with a lesser parts inventory. Many worthy patents are supportable on the basis of a new function of that kind. Furthermore, there is a conflict of expert testimony which as the Federal Circuit points out should not be resolved at the preliminary issue stage. So on the facts, which should matter even to 24 distinguished law professors, I am for the patentee, at least until the matter has gone to trial and the evidence has been tested in the fire of cross-examination.

There is no mystery about collocations. Features do not become a combination merely because some patent attorney in a moment of irrational exuberance stuffs them together into a claim. You need to look at the claim to see whether it really does relate to a single invention, in the sense that all the features work together to produce a single new function or result, or whether the features fall into distinct groups. If there really is a new combination, you deal with it on its merits, whereas if the claim is simply a set of features or groups of features bundled together, then you deal with them also on their merits, but individually. Nothing mysterious about that - the courts have known that this is what they should do since the mid 19th century. But it is not a high standard, an onerous standard, a standard that can only be appreciated by judges standing on their heads, or any similar nonsense. It is simply a preliminary enquiry that it is prudent to make before you consider the merits of an alleged combination.

If you look at what the Federal Circuit did, they made just such an enquiry, found a credible technical problem, and found that there was a solution with credible advantages flowing from the claimed combination. In other words they applied the existing law just as they should.

Nobody needs to go to the Supreme Court to achieve the the results desired by the 24 professors and those who support them. All that is needed is a healthy scepticism that an alleged combination truly qualifies as such until a new function or result flowing from the combination has been identified which justifies its treatment as such. There is no need in a change in the law: just a little mistrust and determination to check things out. But the collocation test has been misused in the past, and if people want to go back to the days of such egregious miscarriages of justice as the Great A & P case, or Kalo v Funk, include me out, even from the safe distance of England.

Incidentally, if you folk look at the technical problem test that we have developed in Europe you will find that it works much the same as the collocation test, but in a framework that makes it more widely acceptable and less prone to the miscarriages of justice. The reason is that the technical problem approach forces the tribunal (Examining Division, Opposition Division or Appeal Board) first to enquire whether there is any positive merit flowing from the claimed combination, and only then to consider whether the combination is suggested by the prior art references considered collectively. Identifying the technical contribution first, and only after that has been done considering the accessibility of that contribution makes all the difference

Paul Cole

How many Patent Professors does it take to change a ligtbulb?


24 think it's too obvious.

With all due respect to the ivory-towered opinions, only 8 of the 24 professors are registered patent attorneys. Two of whom have freely admitted, almost bragged about, their willfull blindness to patent matters to me at IP conferences. To the best of my knowledge, Arti Rai, probably the best patent mind amongst the group, has not prepared and prosecuted a patent application in at least the last 6 years. The soft IP experts that purport to boil down difficult 103 issues to block diagrams are so far removed from the incremental progress of real science that they cannot have any conception of the ramifications of their ideas. I sincerely hope the Court gives their "expertise" the credit it is worth.

The comments to this entry are closed.

Search & Share

  • The Web Patent Blog

Patently-O Jobs


  • 19,000+ individuals now receive Patently-O via e-mail each morning.



Terms of Use & Disclaimer

  • Terms of Use

  • Patently-O on Facebook
    Connect with Patently-O readers.