PRESENTATION BY E. ROBERT YOCHES
FINNEGAN, HENDERSON, FARABOW, GARRETT & DUNNER
MR. YOCHES;:  Thank you, Mr. Commissioner.  My name is Bob
Yoches and I am with Finnegan, Henderson, Farabow, Garrett &
Dunner; and I am presenting my own views today, not the views
of the firm, and not the views of the clients of the firm.
What I would like to address specifically are Questions three
through six of topic B.  However, the testimony I give may
apply to other topics as well.
Questions three through six really address the issue of
whether the examination standard for patents on
software-related inventions should differ from patents on
other technologies.  I don't believe it should.  The primary
reason is, I don't think it's possible and I don't think it's
warranted.
I don't think it's possible because as many other witnesses in
San Jose testified, it is difficult, if not impossible, but
certainly impractical to distinguish between software-related
inventions and inventions based on other technologies. 
Certainly the history of software has arisen many times as an
evolution from hardware to firmware and finally to software.
Moreover, we found, and many witnesses have testified, that
software is ubiquitous.  It is in many different technologies
and it is in many different aspects of the life.  It is no
longer a separate and identifiable part of the technology that
can be treated differently.
More to the point though, if some distinction were made, I
fear that what would result is some sort of game playing.  In
the Patent Office we saw this in the 1970s where clever patent
agents and patent attorneys tried to get around the reluctance
of the Office to grant software-related patents by changing
the specification and claims to make it look not like a
computer, even though that's what the invention was.
Another problem that I see arising from having different
standards for examining patent applications for certain
inventions is in the area of litigation.  Because I think if
you have a higher standard for examining applications for
software-related technology that what you'll do is cheapen the
patents on the other technology, because there isn't a patent
lawyer around who when attacking a patent on a
non-software-related technology won't point out to the jury or
judge that this patent didn't receive the special treatment
that the Patent Office gives to computer patents.
Attorneys representing patentees that have a patent based on
the software-related technology will argue just the opposite,
that this patent received that special attention that the
Patent Office has reserved for computer related inventions.
The other practical problem I see in having different
standards for examination is one in the Patent Office, and
that is a training problem.  I don't need to tell you how
difficult it is to train the examiners with regard to issues
of 102 and 103 and obviousness and novelty.
If they have to learn not one, but two different standards,
and if they also have to use their judgment of when to apply
the one standard as opposed to the other standard, I think
that the training costs and the quality of examination will
drop.
There is, however, I think a larger problem in even asking the
question of having different standards and that's a
philosophical problem, because the questions are based on the
underlying assumption that there's something wrong with
software-related patents that issue from the Patent Office
now.  I don't know that that's been shown.
Certainly there's no question but software is a different
technology than other technologies, but you could make the
same argument about any technology.  I don't think that
there's been a demonstration, other than by some anecdotal
indications that the software patents are any better or worse
than patents related to any other technology.
In fact, to the contrary, I have found that especially in
Group 2300 with Director Goldberg, that there's been an
increased effort and an intense effort to improve the
examination process.  In fact, as you may know, the AIPLA and
the Patent Office held a joint program last fall, a program we
hope to continue, where there was an open dialogue between the
Office and between the practitioners to try to improve
communications and improve the examination process.
More to the point, however, changing the standard for
examining patents will not really address the problems which
have been raised.  Those are the lack of prior art and the
inability to retain examiners.
And now to the specific questions.  Question three asks
whether the Patent Office should impose a special duty, a
higher duty, on applicants having a software-related invention
to disclose information.  I'm not quite sure what's being
indicated there because the current duty is quite high.  I
assume that the additional duty would require some sort of a
search.
However, most of the places that patent attorneys search are
the same places that the patent examiners would search.  So I
don't know that you'd get a better examination process.  What
you would get, however, is a lot more charges of fraud on the
Patent Office because information that should have been
discovered wasn't discovered and given to the Patent Office. 
And as the Federal Circuit has already noted, charges of
inequitable conduct and fraud are a plague on the patent
system.
Question four asks whether the standards of novelty and
obviousness accurately reflect the inventive activity in that
area.  I think they have to.  The standards which the Patent
Office is supposed to apply are independent of a particular
technology and they involve what the state of the art is and
they involve what the level of ordinary skill in that art is.
And the way those standards are supposed to reflect the
particular technology, it's just supposed to be applied to
that technology.  So the state of the art and the computer
technology will track whatever those changes are and the level
of skill will also track those changes.
Question five asks whether we should implement, I suspect, a
per se rule, that if the underlying process is known that
merely implementing on a computer is not patentable.  Again,
I don't think so.  I think that the present legal standard
which asks the Patent Office and asks the courts to look at
the claim as a whole is the proper one because I can imagine
situations where either because of difficulties and
practicalities or because of common knowledge in the art, it
was not thought possible or a good idea to implement a certain
process on a computer.
Although I can't give you a specific example from real life,
one that came to mind on my way over here is the idea of a
product I had just seen, which is supposed to improve the
grammar and the style of writing.  Now certainly English
teachers have been doing this for years.  It's a known process
of how to improve grammar and style.
But implementing it on a computer, I suspect, is pretty
difficult.  Although there is a product out there and maybe
it's prior art now, I don't think that we should have a per se
rule saying that type of product does not merit patent
protection.
Finally, Question six addresses the general question of
whether the PTO should change its examination procedures for
novelty and obviousness in this area, and there are three
subparts.  The first asks whether the Patent Office should
require applicants to conduct a search and distinguish their
inventions from the prior art in the search.
Now the Patent Office already has a procedure for doing this
if you want to get expedited examination.  What I think will
happen is two things.  One is, if this rule existed right now
I would pay the extra fee, which is not too much, and get the
expedited examination.
I don't know whether the Patent Office examination though of
software-related patents would improve, because again the
searches that are conducted generally are from the same
database as the examiners use.
Question B asks whether the Office should impose a special
requirement on applicants to show that their inventions are
distinct over the prior art independent of their computer
implementation.  I have addressed that before.  The invention
is the invention as a whole and, indeed, part of the invention
may rely on how it was implemented by a computer.
Question C asks whether the PTO should be allowed to establish
that a software-related invention is not novel or obvious
using a lower standard, in other words, not a prime facie
case.  I'm a little confused here because I don't know what
could be a lower standard.
The prime facie case merely asks the examiner to do two
things.  One is to find art that shows each one of the claim
elements; and, two, show that there's some motivation for
combining those elements.  I suspect the lower standard could
either be removing the criteria for motivation or allowing the
examiner to reject applications based on his or her gut feel.
In my experience, both occur right now.  Whether they should
or they shouldn't is another issue.  But I don't think it's
appropriate that, again, you should be applying different
standards here.  I don't believe that the result will be any
better patents.  It will just be a longer and more drawn out
examination process.
My conclusion is this.  I believe that if you adopt more
stringent or even different examination standards for a
certain class of inventions, whether it be software-related or
others, that you'll be opening up a Pandora's Box that will
create many more problems than it's intended to solve.  Thank
you.
COMMISSIONER LEHMAN;:  Thank you very much, Mr. Yoches.
Next I'd like to ask Stephen Noe, counsel to Caterpillar,
Inc., representing Intellectual Property Owners, Inc. to come
forward and share.

Forward to Jeffrey A. Berkowitz