Subject: Re: on software freedom and patents
From: "Stephen J. Turnbull" <>
Date: Tue, 05 Dec 2006 13:20:19 +0900

Thomas Lord writes:

 > I am justified because the statement, that patents "grant" rather
 > than "create" rights has several sources.

OK, I see what you're saying.  You're missing the point, then.  I
agree, U.S. patent law does not create rights to make, sell, and use
things; those are basic rights of persons in a free economy (which I
*define* for my purpose in this essay below).

What patent law does create is *exclusion*.  So you could say it
*destroys* or *transfers* the rights of all others to make, sell, and
use, in favor of the patentee.  This is clearly what the Constitution
means by "secure".

 >     3. As consitutional and common law gained control over the
 >         patent system in England and later the United States,
 >         some things about the patent system changed, but there was
 >         no shift from "granting" to "creating" rights. 

No, as I point out above, there was a shift from granting a right
vested in the king to destroying rights vested in each individual

 >     4. Certain precedents, such as TRIPS, the right to repair and maintain
 >         a patented invention you have purchased, and the right of
 >         a farmer to avoid infringement by simply destroying any
 >         GMO crops his fields have made without license, prove that
 >         the patent grants of exclusive rights to "make", "sell", and "use"
 >         are not absolute.

I don't know what you're referring to by "TRIPS", the doctrine of
"exhaustion" is AIUI not an exception to the exclusive right but a
definition of the scope of the use license normally granted by sale of
an object embodying a patent claim, and the case of the farmer is not
exceptional---in civil law, it's "no blood, no foul".  OTOH
destruction of intentionally planted unlicensed GMO crops is not a
defense to *criminal* infringement AFAIK.

Note that I don't have an authority for those claims; they are simply
an alternative interpretation that demonstrates that your "proof" is
no such thing.

 > Hopefully you at least now better understand my alternative reading:
 > the rights secured for the inventor are exclusive property rights that
 > would arise naturally by virtue of the act(s) of invention and that would
 > otherwise belong to whoever seized them.

The problem I have with the "arise by virtue of the act of invention"
interpretation is that that would make independent invention a defense
to infringement.  The high rate of independent invention is one fact
that makes software patents so hard to accept.  But quite clearly the
intent of patent law is that it is *not* a defense.

If you insist on a "grant of existing rights" interpretation, I
propose the following.  In the case of a patent granted by an absolute
sovereign, the theory is that all rights belong to the sovereign until
granted to others by him.  In the case of a free economy, which I
*define* by vesting the rights to make, sell, and use objects
(including services) held under valid title in each person, by
definition the theory is that those rights were vested in each person,
and (for the purpose of "encouraging the useful arts") they were
nullified by patent law.  This dovetails with the explicit mention in
the Constitution; without that Constitutional provision, Congress
would have no power to abridge rights of making, selling, and use
within States.

N.B. I don't claim that the U.S. is a free economy.  I claim that the
Founding Fathers, on good days, in odd-numbered months, aspired to
create such in the U.S. ;-) and that that aspiration is in the
background of the institution of intellectual property in the U.S.
Obviously, IP is a *compromise* with that aspiration, rather than an
implementation of it.

 > Your alternative, I think, is roughly that a patent holder's rights
 > to exclude making, selling, and using an invention are roughly
 > absolute and have nothing to do with actual trade per se.

No and no.  The rights to exclude are absolute to the extent that the
law makes them so.  The remedies permitted are correlated closely to
the amount of economic benefit ("actual trade") received by the
infringer, unless the infringement was criminal.

At one time I did naively believe that the "right to exclude" was
absolute, but then you (IIRC) introduced me to the relevant court
decisions.  My understanding now is still probably not 100% accurate,
but it is that there is in principle an "exclusive right to benefit"
from the invention assigned to the patentee, which is implemented as
the explicitly enumerated rights to make, sell, and use.  It is clear
to me that while the patentee may not prohibit a FLOSS project from
"making" and "using" software embodying his patented claims, under the
law he certainly has a right to demand a reasonable royalty, that
"reasonable" is not defined by the ability of the project to pay but
by comparison with the willingness of a commercial enterprise to pay.
As a matter of legal and commercial practice, he may demand it of the
project, making it their problem to obtain revenues.

If the project has no revenues and prefers not to try to obtain
revenues, presumably the maintainers will prefer to shut down rather
than pay from their own resources.  But that is not "absolute
exclusion"; the maintainers had the choice of obtaining a license,
just as every potential shoplifter has the choice of purchase rather
than theft.  They may not have enough money to execute that purchase,
but that doesn't remove the legally available opportunity.

 > Your interpretation has major problems that I think make it
 > ultimately unacceptable: (a) you have squarely assigned the courts
 > a role in social policy making rather than interpreting the law --

Not at all.  Courts are made up of human judges and juries; they will
overstep their bounds.  But when I look at such decisions as those
which concern the doctrine of exhaustion or the more recent decisions
that the right to exclude does not imply a right to extort more than
the value of a license, I don't see courts overstepping their bounds;
I see them interpreting the law by assigning "natural" interpretations
to the implicit license granted by sale of an object embodying a
patent claim and a "natural" method of valuation of the right to
exclude (specifically, one based in the royalty that could be
negotiated during the planning stages of a new product as opposed to
the protection money that can be extorted by threatening to shut down
a running business).

 > (b) all of the concerns of the FLOSS community about the logical
 > absurdity of software patents (giving the highly recombinant and
 > frequently convergent nature of software inventions) come home to
 > roost under your interpretation (but not mine).

That is not a flaw in my interpretation, it's a flaw in yours.  Is it
a flaw in my interpretation of physics to say that FTL communication
is impossible, even though that might be very beneficial to future
computer hardware designers?  Of course not.  The question here is
"what does the law say", and that's why I ask for an authoritative
source for your rather shocking assertions.

 > In short, all of those FLOSS community objections are best
 > understood *not* as an argument against software patents *at all*
 > but, rather, as an argument that the "absolutist exclusion"
 > principle you espouse is simply wrong.

I don't "espouse" it in the sense of advocating it as policy; I
espouse it in the sense that I believe that what I've written (which I
would not characterize as "absolutist exclusion") reflects the reality
of American and Japanese patent law and regulation.

With that caveat, I agree 100%.  However, to implement an exception
for FLOSS would be to implement an exception; it is not a consequence
of "patent" as currently defined.

 > In that light, all of the current controversies, fights, and
 > political actions in the FLOSS community reveal a new meaning.  We
 > have a picked a correct fight: that the application of patent
 > exclusions to the exercise of software freedom leads to absurd
 > results.  Unfortunately, we, the FLOSS community, have offered an
 > over-reaching solution: the elimination of software patents.  With
 > all due respect, RMS and Moglen have gone a few steps too far.

I disagree.  If the issue is implementing software freedom as
promulgated by Richard Stallman, patents which exclude independent
invention must go.

Thus I resolve the conflict in a different way.  I see software
freedom as something to be *approximated* by law and *granted* by
individual rightsholders for their own reasons.

N.B. "The law can not compel an impossibility" is a legal truism.
That's not to say that "the law may not compel an absurdity," though.

 > Nevertheless, at some point we have to stand up and point out that
 > the threats we retreated from were not legitimate and we were not
 > obligated to retreat: a *single* defense is sufficient for *all*
 > copyleft free software projects.

Uh, what do you think Richard Stallman has been doing, with a complete
lack of success, for the last 25 years?  Those threats are
*legitimate*: that is, they are fully in accord with the law, which
was legislated and interpreted by a political system that has a
tolerably high level of support and consent from those governed by it.

The fact that Richard has convinced a large group of intelligent and
highly ethical individuals that the software patent system is
unjustifiable (and for the radicals, even immoral) does not change the
*legitimacy* one bit.  And that's the problem.  It's very difficult to
achieve abolition of a legitimate feature of society, even if it's a
grave mistake from the point of view of many thoughtful people.

 > I think that what I write actually sums up everything you know
 > about patents and draws a simple conclusion from that.   So, if you
 > are still wondering what factual claims I make lack support, point
 > those out and perhaps I can help.

As far as I can tell, the factual claims about the theory and practice
of patent law in the U.S are unrelated to your argument about how it
should be interpreted.  You are saying "here is the way the FLOSS
segment of the software industry should be treated, and (if we
redefine 'patent' appropriately) it *can* work that way without undue
harm to the proprietary, patent-oriented, segment."  Then by an act of
wishful thinking you assert that definition.  As I said before,
"That's not the way the law works."  It would be an inadmissible act
of judicial legislation for the courts to accept your argument.