Lee Hollaar had these interesting remarks on our brief in KSR v. Teleflex. Basically, there's another side to the issue (not again, why don't I ever get to work on an issue that only has one side? Or defend a popular constituency, or... Oh never mind):

I just finished reading your brief. While I agree that the Federal Circuit has made it difficult to show nonobviousness, resulting in a raft of patents that should never have been issued, I think, as your brief indicates, that they were (over)reacting to examiners doing hindsight reconstructions of a claimed invention.

As a patent agent and patentee and applicant, I've seen first-hand
such hindsight reconstruction. Examiners find patents that describe
one or two elements of the claim, and then make the assertion with
no support that it would be obvious to combine the references.
(It's likely a canned phrase in their word processor.) I've even
had an examiner simply cut-and-paste the first claim, list a number
of patents, and say that they teach the elements of the claim and a
person skilled in the art would combine them.

Unless you have some secondary evidence of nonobviousness, such as
commercial success caused by the invention after a long period of
need, it's hard and expensive to fight such a rejection. You often
have to make some amendment just so the examiner can think that he or
she was right, and then you get less of a patent than you deserve
and a ton of prosecution history estoppel.

The real solution is to give the examiners the time to do the job
right, and have a team of examiners work on each patent. But that
costs money and time, and adding to either of those cause new problems
with the system. That's why the various proposals for patent reform
may just create different problems.

Keep in mind that examiners are graded by how many applications they
dispose of, whether by rejection or allowance.

Any way to get rid of an application helps the examiner meet his or
her quota. We saw, back before the Federal Circuit made it clear
that software-based inventions were patentable, that examiners would
simply say that the application was for non-statutory subject matter
and get their count. (That changed when the PTO issued guidelines
that not only said that such patents were allowable, but also required
the examiner to do a full examination even if rejecting on the
grounds of statutory subject matter, thereby requiring more work
on the part of the examiner and resulting in an end to such
unwarranted rejections.)

I expect that if we back off from the current Federal Circuit tests,
what we'll see is first office actions that are nothing but hindsight
reconstructions and a bald assertion that they can be combined, and
only applicants that want a long battle with the examiner getting
patents.

It's hard to prove that something is not obvious after the examiner
has seen it, especially if the applicant did a really good job of
explaining the invention in the application. The examiner can't
just forget what the applicant has just taught him.

And it's not just examiners. I've seen the same thing in patent
litigation, including simply saying that the problem provides
the motivation or even that the motivation is that the resulting
combination will work. Those will always be true.

I've proposed an alternative to the current patent reform suggestions
of a mini-patent that provides immediate protection against knock-offs
for inventions in commercial use. See:

http://digital-law-online.info/papers/lah/mini-patent.htm

And when there is a more appropriate, low-cost protection available for most inventions, then we can have the examination that the scope of patent protection really deserves.

posted by Solveig Singleton @ 12:06 PM | Academia, Comments from Readers, Patents

Link to this Entry | Printer-Friendly | Email a Comment | Post a Comment(0)