PATENT LAW

The patent-as-property analogy is limited:

1. A patent is just a right to exclude. No 'positive' rights
2. Patents don't necessarily come with an injunction, unlike traditional property
remedy

A patent is a negotiation between the inventor('s agent) and the PTO.

Bonito
Why didn't Bonito file a patent? Was it that they couldn't have gotten a patent? Maybe.

There's also trade secrecy law, state law which prevents competitors from engaging in
activities that are geared toward figuring out your stuff. Largely a tort doctrine. That's
always the patent alternative. Trade secrecy, unlike patent protection, has no limit.
That's one advantage it has. (Coca-Cola has never patented its formula.) Once the
secret's out, though, your legal protection disappears.

But the Bonito boat hull is pretty easy to copy. It's sort of like the hula-hoop in its
transparency. As soon as you release it, the "secret" is out. When that's true, TS doesn't
provide much help.

So what else. For a patent, you need novelty and non-obviousness. N: If you want to
patent something, it can't be found in the material that existed prior to your inventing it.
IOW: it needs to be new.
Maybe Bonito hull wasn't new. How would you know? Well, first, you could check a
description of the hull against what else has been patented. But what about
other, unpatented stuff? You just submit the patent. You gotta do a reasonable
search, but then that's it. The examiner will do her own search, and the P can
always be challenged later.
Interesting dynamic: For stuff that his little prior art, it's easier to get the patent through,
but it's also more likely to be challenged
N-O: In order to patent a certain invention, the idea you're claiming can't be "obvious"--
obvious to a person who's skilled in the art at the time of the invention.

Also, there's the utility requirement. The invention has to be useful. This requirement is
pretty lax. This only really comes up with inventions that are well "before"
commercialization. (Sort of like originality in copyright?)

And: the best mode requirement. Part of the 'disclosure'/enablement part of patent law.
You have to describe the invention that enables others to use it. You need a 'written
description' of the invention, disclose the "best mode" of making/using it.
Best mode gives your competitors a manual on how to make and use your
 invention. The risk here becomes enforcement. A reason not to patent is you
don't want to have to worry about your protection working. Expensive to enforce.

And best mode allows competitors to invent better products based off your
patent. The "design-around" problem.

So what do we have here. Why didn't Bonito patent the hull? Maybe they were relying
on trade secrecy, possibly combined with the dynamics of the market (by the time the
copy comes out, they themselves have a new product). Or they were unsure about N, or
N-O. Or they didn't want to bear the risk of enforcement, or (especially!) design-around.
*

So we have this Florida statute protecting boat hulls. Why? First, clearly, a big boating
industry. But is there a logic behind the politics. They seem to be trying to patch a hole,
to create something like patents, and patent protection, where there was none. Why?
If R&D costs associated with an invention are high relative to the ease of imitation, that
can turn into a big under-production problem. Why invest when you can wait and copy.

So - why does SC strike the law down? Preemption. But there's all this other stuff states
can do. Trade secrecy, for one. How is this different from trade secrecy, in terms of
meddling? Well, the Constitution says that patent protection must be limited. And the
Constitution has been interpreted to also bar protection from anything that's already in
the PD. Giving unlimited patent-like protection to Bonito, when it's hull has already been
for sale for a while, implicates both these Constitutional issues. And the patent system
has been set up as a quid pro quo, which is not present here.

But this isn't the route the Court takes. They don't say "if the law had x, it'd be ok." Why
is trade secrecy law ok, but not this? It has something to do with property rights. Trade
secrecy is protecting rights in a very common law sense. It's "taking stuff" in a much
more everyday sense. It's about possessory rights. But once you start selling
something, you lose possessory rights over it (obviously!). The Florida law is a lot more
like patent protection.

Does the case make sense? SC says it very well might be a good idea to protect the
sort of thing that boat hulls are, but that it's Congress's call. What do we think?
*
Why do we have a patent system.

1. Incentive to invent
R&D costs + ease of imitation
But who says there's a causal relationship between financial incentives and actual
output of invention? What do we actually know about the creative process? Pushback:
Inventions happen where there's money to be made, and that's not a coincidence.
So patents are incentive to innovate in the areas that patents cover. Patent law pulls
inventors toward protected areas from unprotected areas. From, say, boat hulls to
business methods. More rent there.

And, by the way, there's also a free rider problem. That's the really simple story for why
you want patents.

2. Disclosure
Disclosure leads to spillover benefits. Increase the pace of innovation.
Is there any reason to be skeptical about this? Economists, for a couple hundred years,
have not been particularly enthusiastic about this rationale. For a lot of inventions, once
it's sold, the secret's out. So why do we need patents, if they're just to encourage
disclosure?

In order to make money off *your patent* (as opposed to your product), you need to
enforce it. Litigate, etc. In order to litigate, you have to be able to spot infringers. If your
patent is over just a process, that can be really hard to do.

So there's something odd about the disclosure rationale. On the one hand, there's the
hula-hoop point. No extra need to incentivize disclosure. On the other hand, if trade
secrecy's available, disclosure may just be a really annoying and avoidable thing.

3. 'Property rights'
If you want to do/make x, you gotta go through me. Trade secrecy isn't good for this.
The Disclosure Paradox: If you want to sell an idea, you don't have any rights to it. All
you have is the fact that you have the idea. And in order to sell it, you have to tell it. And
once you tell it, there's nothing to sell. Trade secrecy disappears right when you need it.
Patents allow contracting.

4. Avoids wasteful/duplicative R&D

But this raises a counterpoint, about 'patent races.' When you have a prize (rent) for a
particular achievement, you're going to get a bunch of parties racing toward it. So the
story is, if you offer a patent, a lot of companies are going to engage in the same R&D
to get to the patent first.

One way around this is the fact that patents usually get awarded pretty early on--the
race usually ends pretty early. So the opportunity for duplicative R&D is smaller

5. Incentive to develop & commercialize

There's a lot more that goes into innovation than coming up with new ideas.
*
That's the basic story to why patents. So on the other hand, what are the costs?
1. Price
The flip side of the incentive to produce is higher prices. So how do we know whether
it's good that we grant the patent? For every bit of incentive, there's a bit of price
increasing (not really, but still). A useful over-generalization: Do we need the patent to
get the invention. If you're going to get the invention anyway, then all you're getting is a
higher price. Better not to grant the invention. But if you need the patent to get the
invention, then you're better off with the patent. Better to have higher prices for a limited
time than not having the invention in the first place.

That rule of thumb is tricky for a couple reasons. In a lot of cases, you'll get the
invention without a patent, but it will be delayed.

2. Thickets/ licensing costs

As technologies have become more complex, this has become really important. This
cost is the flip-side to the property story. Huge administrative costs (clearance/search
etc.) may make the innovation not worth it.

3. Antitrust

1/26
• summary: five main patent system advantages
o incentive to invent
• just won't have inventions
o incentive to develop and commercialize
• companies just won't bring things to market
o disclosure function
• by offering patents, people will disclose technical knowledge
• economists don't agree that patent systems actually encourage disclosure of
information
o "property rights" story
• facilitates transactions, ie. technology transfer
• doesn't have to be the same person that invents as markets
• dis-aggregates the R+D process
• part of the disclosure function
o reduces wasteful R+D
• once someone has a patent on something, no one else will duplicate that R+D
• summary: problems/costs with patent system
o core cost of patent system: prices are raised
• patent system rewards innovators by allowing them to charge above marginal
cost
o licensing costs/patent thickets
• when you need to use patented innovations to develop your own product, those
patents function as a tax on your downstream products
• not a big deal if we wouldn't otherwise have the upstream products
• cost is if we didn't really need those upstream patents, ie. university research
• transaction costs associated with licensing
• patent pools can make this more efficient
• patent thickets: so many patents in an area that it becomes not worth entering
• too much licensing, risk of infringement
o wasteful R+D
• deters competitive R+D, the deterred party might have come up with something
better
• competitive balance might spur more innovation
• ie. monopolists are not incentivized to innovate
• patent races
• multiple firms racing to get patent
• solution: grant patents earlier in R+D process
• but this further reduces competition for better R+D
• reading patents:
o hybrid between lawyer and engineering phrasing/terminology
• mixes technical terms with legal formalisms
o inventors have to describe invention in abstract form
o Markman hearing: judges interpret the claims so that analysis can proceed
also known as a "Claim Construction Hearing"
• question of Markman case: is claim construction a jury issue?
• SCOTUS had to go back to 16th century and determine difference between equity
and common law
• Markman holding: a judge issue, not a jury issue
o definition for elements
• distinct structural properties as outlined in the claims
• nicely separated by commas
• not actually a definition
• Larami v Amron, 1993
o words of a patent to be interpreted in their ordinary fashion without regard to
clarifications offered later by inventors
o patent infringement demonstrated in one of two ways:
• literal infringement of every element of a claim
• absence of even one element of a patent's claim from the accused product precludes
literal infringement
• every element of a claim in infringed under the doctrine of equivalents
• every element or limitation of a claim is substantially equivalent in the accused device
o commas nicely separate the three elements
o strategy: break a claim into structural element, ask what we know about each element
o element: housing
• elongation is a limitation
o element: pump
• describes what it does to simplify explanation
• tells us a lot about what it looks like
• functions as a limit
o element: means for controlling ejection
• means-plus-function: claim isn't telling us what this does
• could be useful if you can't actually describe all the ways of doing the function
• stops people from designing around your patent
• these functions are not normally the core of the innovation
o drawing is just one particular embodiment, claims are what matter
o holding: if even one element is different between claims/infringing device, then device
doesn't infringe
o doctrine of equivalents: if infringing device is "close enough"

Paragon v. Timex
Timex (defendant)'s argument. It's a unit, after all, and our product separates the
structures

On the alternative argument, about the 'real-time data.' This is a patent on a device, not
a use or process. Thus, if a device infringe, it infringes never or always--not depending
on the use.

Paragon's big mistake here: Why didn't they just define "real-time" in the specification
section? If you define your terms in the spec, that holds up. And they could have
defined it in a way that made a design-around costly. (Otherwise being specific can be a
roadmap to legal 'infringement')

This gets to what Bessen is talking about. Even if you're used to reading patents, it will
be (purposefully) hard to know exactly what a given patent covers. In this way, the
property-rights story that gets told about patents . . . this is the counter to it.

Atlantic
What's special about these inner soles? A flexible plastic insert in the heel and arch.

This is a different kind of claim--it's a method of manufacturing. A process claim. For a

process, the elements of the claim aren't structures, but steps. Which is to say, verbs
are important.

Then the "improvement" comes along. This is a jepson claim.

Fed Cir had said that if you have a product-by-process claim, your claim covers any way
of making that same thing, even if it's different from the process you've just disclosed.
But here, same Cir a year later, disagrees. Look at the history--why would we allow
someone to describe a product just by describing a process? It was because that was
the only way to describe it. You used to have to show the move, but you don't anymore.
But the compromise, this judge says, is . . .

J says the worry with PbP claims is that it risks issuing patents for products that in fact
aren't new. So the compromise is limits on the claims. If someone comes up with a way
to make the product thru a different process, tough.

So, 10 years later, the split is finally resolved. The rule now is that PbP's process is a
limitation on the scope of the claim. Fine. Why is the dissent so mad? Because
processes are so easy to invent around, the argument goes.

On the other hand, limiting to the process might incentivize innovation of processes.

2/1
Roin: When reading a claim, divide the invention into its separate physical
structures/steps. Frequently in that regard, the grammatical structure helps.
Product by Process - Description of process acts as limitation on claim, most recent
holding says. The rule is a trade-off between broad scope (bigger reward for inventor)
and more room for competitors.

On to stuff for today - patentable subject matter.

The statutory basis is §101: "any new and useful process, machine, manufacture, or
composition of matter"

Diamond v. Chakrabarty
Bacteria being used to break down crude oil. A man-made, genetically engineered
bacteria. Three claims: process, the combo of bacteria with some other stuff, and the
bacteria itself. Only the bacteria itself is at issue. And Court says ok. It's man-made, and
it's not found in nature

So wait: you can patent living things? "Anything under the sun made by man", except
for "laws of nature, physical phenomena, and abstract ideas"? Can you patent a
genetically engineered human?

Suppose you could patent a genetically engineered mouse. You create the embryo etc.
Now: What rights do you have? You have the right to exclude others from making,
using, or selling it. Someone makes, uses, and sells, them. You sue, and get an
injunction. You can stop them from making, using, selling it. But what happens to the
mice?
Trying to get to an important distinction between "owning" what's patented and having
the exclusive rights to make, use, and sell what's patented, which often gets confused in
talk of patenting living things.

Take-home: You can patent living organisms. "anything under the sun."

O'Reilly v. Morse
The telegraph, basically. The eighth of Morse's eight claims is really broad. Any use of
electromagnetism, basically. Court says too broad.

Comparison made to old case, Neilson, where claim was upheld. About blowing hot air
into a furnace. Court needs to frame this case narrowly to distinguish. In that case, there
was a description of the physical machine.

There's a sense here that this patent covers a huge variety of yet-to-be-invented cool
shit. With the furnace, you're not quite as worried. Heating furnaces isn't a whole new
field, we've been heating things for a long time, etc. Not a principled distinction, but a
distinction. So, why are we even asking this question? We have the enablement
requirement, can't we just end it there.

What does the dissent say? Something about "pioneering invention." If you are first in
the field, if you're incredibly novel, then so be it--your invention is broad.

Also talks about principle being applied. Once you put a principle to use, it becomes
patentable subject matter. Frames patentability as utility.

So we have anything under the sun from Chakrbarty. And now we have a slightly
different, but not inconsistent way of looking at it: Decide whether it's an abstract idea by
looking to see the extent to which inventor has enabled. The more the claim goes
beyond the enablement, the more it looks like an abstract idea. The dissent thinks it's
only abstract if there's no use attached. E=mc2 is absract; the idea of using that to
create power is not. A very different rule.

Parke Davis
Two claims at issue. One for a base, and one for a xxx. This is purified adrenaline. Not
naturally produced. Hand says that re base, even if this form *does* occur naturally,
then you can still patent it. Why?

Hand says look, this was very useful, and a big advance over the prior art. But this
doesn't quite get to the problem: these are compounds that exist in everyone's body.
Why is it that this thing, which our bodies produce, is patentable? Because, Hand says,
it's *extracted*, isolated. Even if something is produced in nature, you can still get a
patent on it if you can get it newly outside its original context. So that's the first claim

What about the second claim - the salt forms of the adrenaline. Same problem with
naturally occurring, but that's solved the same way. Here, there's another problem. A
bunch of other patents already knew about this. But the claim in this case is for the salt
form *in a highly concentrated form*. So although the salt isolation itself is not new,
what's different is . . . The very fact that this thing is concentrated *makes it new*. Much
in the way that isolation makes it new. And *why* do we have a rule like that? Because
these features can make something previously not useful useful.

These cases are just rules now. But they're *very important*, especially for the
biotechnology industry.
Dude: Why not just patent the process of concentration, rather than the product itself?
Of course, the patentee wants the broadest patent he can get. Patent on a process is
easier to work around that patent on the product, especially in biotech, where the
specifics of the product are crucial. So that's the upside.

The downside is lower incentive to improve the process. What if someone figures out
how to make it *really* pure. You can get a patent on the new process, but since you
get the same product, you'll have to go through the original patentee to use the process.
Increases costs. It's possible the new inventor will just say forget it.

Diamond v. Diehr
What's patented: a process for curing synthetic rubber. This seems classically
patentable. So, what's the problem? Part of the patent was this formula, and a computer
program

They're using the equation in the way it's designed for. All they're doing is automatizing
it through a computer. The concern, IOW: There's this algorithm, and none of this stuff
is new, it's just the combination that is.

What's the test here? A bit scattered. It's ok if there's a formula/algorithm involved, but it
has to be more than just that. But how do you get to patentability? You certainly have
novelty and utility. And, as Court says, there's "transformation of the article". So part of
the test for a process's patentability is exactly that--the transformation of matter.

How would a sports method--how better to throw a baseball--fare under this test? Prob
not. What about the Pac Man game? Again, no. Anything that has to do with software
ain't gon pass. It always looks a lot like 'just an algorithm'. There was something else to
this case:

"Tied to a particular machine". In Benson, which wasn't assigned because it's

incoherent, software patent is invalid. Here, the process is tied to a particular machine.
That's enough to make it not-abstract.

The concern here: Should mere 'computerfication' be patentable? Amazon one-click

example. It's just a tab, in computer form. Why is that patent-worthy? It's hard to figure
out when you're overlooking important ingenuity. The "hindsight bias," that is, is a
counter-concern. But these lingering doubts have played a pretty big role in the
evolution of the computer-stuff doctrine.

The dissent here was pushing on the question of software patentability full stop.

Alapatt
A squib worth mentioning. What eventually evolved was claiming software patents as
machines. Could get around the 'transforming matter' and 'tied to a particular machine'
tests. Diehr became less important . Then comes

State Street
What's patented: a "hub and spoke" business method for mutual funds. The "hub" is a
partnership, and the "spokes" are mutual funds. Gains and losses from spokes are
pooled together in hub, and some tax benefit is gained, and spokes get some of the
hub's profits. Or...something.

Claim gets rejected on a few grounds. For today (second half tomorrow), the basis is
that it's an algorithm. With an algorithm, the question is always 'can you patent it
anyway'. Here, all the algorithm is doing is punching numbers. And it's not tied to a
particular machine. It has to run on a computer, but that's not thought of as a particular
machine. And on first blush it doesn't seem like it's transforming matter.

So District Court says invalid. But Fed Cir disagrees. "It's useful." But that's not it, is it?
Remember, novelty, non-obviousness, utility are their own requirement, separate from
this subject matter stuff (in theory) (like 102(b) vs. originality, seems like). So where is
the basis for the decision here?

Clearly the court isn't overruling Diehr--they can't. So how do we reconcile these two
cases? Well you can define this as transforming matter. Defining "transforming matter"
as "resulting in something useful" can help you. But is there another way to do this? In a
way, this case is a reformulation of the dissent in Morse: an idea is not abstract if it is
being put to use. This isn't really the rule that SC handed down in Diehr. So, what else.

You could say it's tied to a particular machine, by essentially reading the "particular" out
of it. But what else.

This is sort of a radical decision, seeming to fly in the face of the rule. But it stuck,
because it was clean. After this case, the subject matter issue ceased to be interesting.
If it's useful, you can patent it. It got rid of what we now can see is a sort of really sticky
issue.

But this didn't last. We'll talk about this tomorrow

2/2
Diamond sets out the general rule (sun minus abstract). Then Morse comes in, and
basically describes abstract idea as a matter of breadth. The dissent, on the other hand,
simply thought that "abstract" meant "not useful." This distinction is important. Then, on
the side, is Hand's Parke Davis opinion. A rule about stuff that exists in nature. PD says
you can patent stuff in nature if you've isolated them or purified them. Basically, a way
of getting around novelty requirement.

Those are the three basic background holdings. Then we get into the process
patentability stuff.
From Morse, we have scope.
From dissent in Morse and basically State Street, you have practical application.
From Diehr, you have tied-to-a-machine or transforming-an-article. Sort of a
physicality req.

Before we get into that, a discussion of university inventions (readings from yesterday
you didn't do). Nowadays, unlike before, any research funded as university activity can
be patented by the researchers.

What's the obvious problem here? University researchers are being paid (by the gov't, I
think the idea is) to do it anyway. Think of the analogy to roads. This is like both paying
for the road, as if it's public, and letting the builder charge a toll for it, as if it's private.

What's the first answer? The commercialization/development story. The costs that need
to be recouped aren't just the costs of inventing. There's more. Like, as an analogy,
road maintenance. The government funding may be enough to invent, but not enough to
cover the development and commercialization of what's developed. You need a patent
to get it to the public.

But that's not a great answer. Even if you can't patent a particular compound developed
through public funding, you can still get a patent on a process (say, a therapeutic use)
that uses that compound. But maybe these downstream patents are too weak?

What else, against the d/c story. Well, maybe the government is "really bad" at
negotiating these transactions. They give the rights to sell a drug at higher prices (they
let a corp get certain rights) not for a good price, but for almost nothing.

Another problem: Not all inventions need mo money for d/c, and it may well be the case
that most don't.

Another justification: This is just another model of government funding. A tax on users of
inventions that goes to inventors. And the useful inventors get more of the it.
Also: It's possible that by offering patents, the gov't will shape university research
toward commercializable stuff. But the flipside of that is that it makes people think only
in terms of short term application.

Another worry: Researchers publish their findings anyway. Why do you need a patent?
But publishing is a far cry from full development

[The point of the conversation: Talking about whether something should be patentable]

State Street pt. 2

No business method exception

35 USC 273
Passed after State Street. It's a prior independent invention test. The concern: There's
this change in the rule, and now some company is going to come in a file a patent for
something that the other company has been doing for a while. Why don't we have this in
other areas? Well, business methods can be both old and novel, because they're
usually used behind closed doors.

But it's not just that. Think of manufacturing processes that go on behind closed doors.
Since patent lawsuits are so risky, people are averse to calling bluffs. So like it or not, a
good business model is to go around asking everybody for a little money. This is
especially true in the business method area.

LabCorp (tomorrow) triggered the PTO to start denying patents on a lot of stuff. The big
one was

Bilski
What's patented: A method of hedging risk. Three steps. A really old idea. Not new, very
obvious. But those aren't the issues here. Here is a 101 issue

Court uses the transformation-or-machine test. Machine gets thrown out right away.
Leaves the lingering question of whether it's patentable if it had been specified that it
would be done on a computer. But there's no computer limitation in the claim, so it
doesn't come up.

Court puts further limitations on the TM test. (tied to a particular machine/transform an

article into diff state of thing).
Limiting an abstract idea to a particular field doesn't make it patentable ("only applies
to commodities").
And "insignificant post-solution activity" isn't enough, either.
And mere data gathering added to an algorithm also doesn't help

One way to think about this: Court talks about TM in terms of abstractness. And
abstractness is when a claim preempts the whole field surrounding the claim. So you
can think of that as a separate test in there when the TM stuff gets confusing (?)

Court also talks a lot about what this test isn't. Not a "technological arts" test. Not the
State Street test.
*
So we have tests, and not-the tests. The court now has to look hard at "transformation
of article." The big issue is what you're transforming.

Here, what's being transformed are legal obligations. Are these 'articles'? Transforming
a physical structure is ok, and (see that cardiogram case) transforming data that
"represents" (whatever that means) a physical structure is ok.

(Discussion of option rights vs. possessory rights. Does/should it matter?) Court says
this claim covers any transaction involving commodities. How do you reply as Bilski? If
hedging is an abstract idea, you can't save yourself (see field limitation) by just saying
it's limited to commodities. So you need to argue that hedging isn't an abstract idea,
basically. And the way you do that is to say look, how is a delineated method abstract?
This isn't E=mc2 here. And court says back, hey, this is so broad. Or something.

Roin: But how is this really different from any other patent? Takes an abstract idea and
makes some use out of it. What's wrong with that. Well, it's true that all patents are in
some sense abstract

(Bilski's lawyers: can't patent something that goes on inside your head. If it doesn't take
place in the physical world, it's not patentable. First Amendment issue, sort of, but
definitely an enforcement issue. A broad, clean line. Once x has some physical effect,
it's patentable (as long, and it's a big as long, as it satisfies the other requirements)).

2/8
Hypo: What if the Bilski patent was "a software program contained on a disc that....
[original patent]." Technically, Bilski doesn't apply, because Bilski was about process
claims, not machine claim.

What about an operating system? Now we're under Bilski. We'd look to 'tied to a
particular machine'--the question becomes whether a computer is a particular machine.
(Benson suggests general purpose computers don't count.)
(What about transformation for the operating system? Remember, Bilski says when
you're transforming data it has to be a physical structure or represent an underlying
physical structure, or something.)

Software running a satellite would probably be ok--the more particular the machine, the
better.
What about video game software?

What would the State Street argument be in Bilski? The program transforms
representation of real physical things, the rpt's being financial values.

What about the Amazon one-click thing? If you're going to argue that it stands after
Bilski, you have to say this: Bilski was about options, whereas one-click involves a real
physical thing

Roin thinks this is what Bilski is trying to do (so hard to tell): Fed Cir has intuitive sense
that patents have to have something to do with technology. FC doesn't know exactly
what they mean by technology, or exactly how much the patent has to do with
technology. But they're not comfortable saying something like that. "Has to have to do
with technology" seems like an impossible judicial standard. And then at the same time,
they have the Diehr language. So they're using this test, keeping it incredibly vague so
that they can freely apply their instinct to different situations. This is why the two Bilski
dissents are interesting. One is saying that you just don't know patents on (technology, I
think). The other one says look, we don't know xxx, and we should go with Morse here,
focusing on patent scope.

Lab Corp
Process claim. Measure the level of x, take note of it, and conclude whether it indicates
vitamin deficiency. Claim uses the word "correlate" as the second step to describe the
mental step. All the action takes place at the first step, the test. So we have something
going on in the external world--a science-y thing, at that.

Why doesn't the test alone get them into 101, according to Breyer? Well, the way the
patent is claimed, the first step can be done in any which way, including the 'star trek'
test. It's a lot like Morse. What you have here is an important discovery about a
correlation, and you have a claim that seems to cover the entirety of that principle

Is this right? The basic tension between what you're trying to promote--the discovery of
health-related correlations, say--and what patent law lets you protect (and between
incentivizing innovation and incentivizing subsequent improvements)

(missed a lot) This distinction between pharmaceuticals and diagnostics. Not clear you
need patents to cover R&D for diagnostics.
Prometheus
Immunosuppressant drugs are dangerous. This is a process claim. (1) Give person a
drug. (2) Determine the level of active ingredient in the body. (3) Depending on the
level, a warning goes out to increase or decrease. The "indicating" step.

At least this is how the court construes it. Not just a mental process, though that's
involved. Raises an interesting question about infringement. If the infringer uses a
different exact number, is there infringement? If there's any overlap, at that point there's
infringement.

Fed Cir blows off LabCorp because it was just a dissent. Court says a couple things.
First, administering the drug transforms the body and so it satisfies T (this is consistent
with LC, since LC didn't have that). But then it says even disregarding that, the blood
sample is transformative.

But huh? The step in Prometheus just says "determine"--doesn't talk about blood
sample. But determining requires a blood sample and doing stuff to it.

But really, how is this different from LC. How does this not preempt an entire
fundamental principle. This decisions sort of reads Morse out of xx. Morse is going bye
bye.

Diagnostics is not where you want t experiment with denying patent protection, Fed Cir
is saying. Too important, and genomic stuff is really expensive, or something.
This isn't consistent with LC, and it's only sort of consistent with Bilski, in that it basically
guts the Bilski standards (i.e. how is this not just data gathering).

Nujten
The question here is different. It's about a new kind of technology. Decreases distortion
in sound waves caused by watermarking.

Thing is, he claims the signal itself. Some sort of electrical impulse. PTO rejects. What's
the problem? This is certainly science-y. What's patented is the "signal".

2/9

UTILITY
As a logical matter, any utility rejection under 101 is also going to operate as a rejection
under 112. If you didn't disclose a use, you can't enable it.

Operational utility
Fantastical claims, drafting errors
Beneficial utility
Lowell
Famous Justice Story case. At issue is a pump that's not much better than the pumps
that came before. Story says ok: "All that the law requires is, that the invention should
not be frivolous or injurious to the well-being, good policy, or sound morals of society."

The idea is, let people/the market figure out whether they like the invention or not. The
patent office and the courts don't need to predict. Sort of an institutional competence
point.

Juicy Whip
Regular beverage dispenser made to look like pre-mixed dispenser. So the purpose of
the invention is basically to lie to people. Fed Cir says ok, and in doing so, they basically
get rid of Justice Story's rule. Not entirely, though. If the purpose is *illegal*, you can't
patent it under beneficial utility.
*
The questions presented by this sort of cases keeps coming up. Agency and judicial
competence.
Marcia Angel is worried about 'me-too drugs'--drugs with no use additional to what's
already available--being patented. The market won't necessarily take care of this,
because of the way that pharma marketing works. Wasteful to produce something
identical.

Why would a company spend a lot of money making Lipitor II, when Lipitor I has gone
generic and cheap?I.e. why would a company engage in socially wasteful/duplicative
research. Because whenever you're the first in the market, there's a big market share.
And a second comer, even if they're selling the exact same thing, can capture a good
chunk of that market. This is the common goods problem.
*
The patent race story - Kitch reading (didn't do)
A lot of money to be made if you get the patent, a lot of companies trying to get the
patent. As you get more and more entrants, the expected profit decreases (because the
chances of getting the patent money goes down while R&D stays the same). Which is to
say, the race eats up the social value of patents. Way more being spent on R&D than
necessary.

This is the theoretical foundation of how we talk about the utility requirement. Because
utility changes the timing of when you grant a patent. If you grant a patent early, in
theory you're reducing the waste and duplicative research that goes on. (Btw, the same
is true about the breadth/scope of the patent granted). But, on the other hand, another
word for "racing" is "competition." R&D is unpredictable. Lots of people do it might
create lots of different things, or it might make innovation speed up, or it could create a
competitive end market of multiple similar but meaningfully distinguishable products.

Practical utility
Brenner
At issue, a cheaper process for making a known steroid. A priority fight. The second filer
says he invented first. He can show that he invented it first, but he has to show that at
the time of invention, he had a utility for the invention. Can't get benefit of earlier date
without that.

His first evidence that his process is useful is a paper (written by the other side)
showing that a similar structure is useful in treating tumors. Court rejects this because
(and this is unique to the life sciences), there can be a lot of variation in small
difference, lots of unpredictable stuff.

His second argument is that the process makes the thing that a paper has suggested
could be useful. It's a process for a marketable thing. Court rejects this, too. They talk
about policy .One reason to be liberal about granting patents is increasing disclosure.
But disclosure is so artful these days (to show nothing) that's not a compelling
consideration.

The court also worries that granting a patent on a compound without knowing what it's
good is potentially really broad in scope. (Dissent, probably correctly, doesn't buy this.
You have this problem anyway, because a patent in a compound with only one known
use covers all its future uses).

In his dissent, Harlan points out that so much scientific discovery happens at
'intermediate stages.' You need the intermediate stages, we want to encourage them.
So why draw the line? He doesn't get it.

Brana
At issue a chemical compound. There's this article about very, very similar compounds
and their use. This patent is claiming that their compound is even better at that same
purpose.

This patent ultimately is upheld, even though all the claim says re use is "anti-tumor
properties" and "better than that other compound". This is a specific utility question. You
need to describe exactly what you're using this for. It's not like we don't believe you. You
just have to talk in more detail.

PTO, court says, has to meet a burden to cast doubt on the specific utility. In other
words, there's a presumption of truthfulness toward the utility. It's easiest to do this in
fields where there are few or no past successes.
Once that burden is met, patentee has to overcome it. Animal, in vivo, human--one or
some combination thereof, depending on how well accepted the method is of predicting
future activity. Structurally similar compound use by itself is never enough, but it is
evidence, and it can support other evidence.

Fisher
During the biotech revolution. ESTs. PTO had put forth its 2001 guidelines. You can't
patent a compound (or a process for making a compound) unless you know what that
final thing is good for. What they're getting at: It's ok to patent intermediate stuff, but you
gotta know what the end stage product is good for. An application of Brenner.

Here, at issue is EST for corn. Probably interested for application in seed technology.
You use ESts to find out when a particular part of DNA is being used to produce
proteins.

Part of the reason the court rejects is specific utility--everything the claim listed
generically applied to all ESTs.

(missed)
--Unlike a microscope, which is a generalized platform
--All the uses are hypothetical.

2/15
The question of practical utility breaks into two questions: specific and substantial utility.
The first asks whether a particular use for the invention has been identified. The phrase
we get from Fisher is "not so vague as to be meaningless". From Brana we know that
naming a specific cancer is good enough.

Fisher also told us that a simple laundry list of generic uses is not enough. And the PTO
has a rule about claiming sort of tautological uses. Using the chemical as a paperweight
or whatever.

So that's specific utility. More complicated is substantial utility. The thrust is about
present use in the real world. You have some of this language in Brenner--a patent isn't
a license to hunt. The problem becomes distinguishing intermediary research steps--
which we don't want to protect--from research tools, which we do.

[XX] says you can't patent something like a gene line if you don't know what the end
product does. The court in Fisher is saying, really, is that it's a question of timing. Once
you figure out what the gene does, you're welcome to patent it, along with the research
tools you're using to discover. But until then, you have to wait. In actuality, it's more
complicated that, not least because this analysis interacts with scope stuff and other
stuff. But that's the basic idea. When do you grant the patent.
If you grant the patent early, you stop wasteful/duplicative racing among competing
researchers. You allow the party with the patent to coordinate the rest of the industry's
R&D efforts. The drawback is that we might not trust that coordination process. That we
can't trust the controlling party to license when it's good for society. The earlier you
grant the patent, the more worried you have to be. So. On to

ENABLEMENT
The claims are the metes and bounds. Before that is this long description. That's where
we are now.

Incandescent Lamp
Sawyer and Man use carbonized paper as their light bulb filament. But they make a
broad claim to any fibrous material used for filament.

Thomas Edison figured out that bamboo plants have high resistance, which is good for
light bulbs.
So we have a straightforward literal infringement claim from S&M.

So what's the problem with the case? Well, the vast majority of stuff in the broad class
of filaments that they claimed don't actually work, according to the court. And there's no
info in the application that tells you how to figure out which instances will work and
which won't. This points toward 'undue experimentation', which we'll get to later. But on
its face--broad claim, largely unsubstantiated--this is an enablement problem. The
invention isn't enabled because it doesn't tell someone skilled in the art how to do
whatever the invention does.

Enablement and utility overlap in that they both have to do with scope. But enablement
is a broader issue--even if there's a specific utility, enablement can come into play if the
claim is significantly broader than the particular embodiment. Which it usually is--
otherwise design-around would be trivially easy.

So you need a way for patentees to claim a little/a lot more than their particular
embodiment. The trouble with this doctrine is doing that without going too far.
*
So how do we allow this extra space? First, if you identify what it is about a given
compound, then that's good. You can have variations on that compound.

(missed a bunch of discussion)

Amgen
The claim is over a 'purified and isolated DNA sequence consisting essentially of. . . '
(That 'p&i' part is there for patentability purposes, remember?). 'Consisting essentially
of' is the intermediate construction. Not open, not closed. And they have 'sufficiently
duplicative' + key characteristics, to protect against functional equivalents.

The problem is that there are tons of analogs to this DNA. They've ID's one
embodiment, and they're trying to claim all the other embodiments, but they (1) haven't
shown which will work, or (2) told us how to predict which ones will work. You'd have to
engage in a lot of experimentation to figure those things out.

What if the patent had been claimed structurally, as opposed to functionally? It would
have helped on the enablement test. It wouldn't just be blind experimentation. But--as
soon as you start getting more specific, you can run into utility problems, if you have the
same knowledge gap.

Wands
The patented thing is a diagnostic test using antibodies. Wands wasn't the first to do this
kind of diagnostic. But Wands did discover that the IgM antibodies were a better
predictor than the usual IgG antibodies. So the claim is for a diagnostic that uses IgM
antibodies found through this process with high affinity for a particular kind of antigen.

PTO rejects the claim. In order to enable the claim, people have to go through this crazy
process. Inject an animal with Hep B, extract this, test that, etc. Too much
experimentation. Fed Cir reverses. Uses the "person skilled in the art" concept.
Sometimes PSA is a low threshold; sometimes it's high. Cuts both ways, because
relevant to both enablement and obviousness.

That list of undue experimentation factors on page 4 of the opinion. One thing this case
definitely has is a working example. They even deposited samples in the PTO office.
Why doesn't this end the inquiry? Because the claim goes beyond those particular
embodiments.

A factor that the PTO really doesn't like is "the predictability or unpredictability of the
art." Unpredictability of the result, of the art in general. Suggests you haven't fully
enabled the claim.

Why did Wands want this broad claim. Similar to the problem in Amgen--anybody who
goes through the process will end up with a slightly different cell line. If this patent is
defined this way, it's going to have a really narrow scope. Sort of the opposite of the
product-by-process stuff from the first day (Atlantic). But the pushback is the potential
for a big improvement within this field.

What's different, really, between this and the incandescent lamp case? One thing is that
these guys actually found a category that works pretty well. They couldn't figure out
why, but they did know what worked (unlike the fiber stuff).
The other test--aside from the process being described--is that the starting materials for
the process need to be available.

This case is useful for inventions that you don't know how they work. Wands is seen as
expanding the scope of enablement.

For T - Written description + claim definiteness.

2/16
Incandescent Lamp case shows you have to at least know why it works. And you can
claim a scope of materials, but the fewer of them that work, the more you'll want to show
you know how to pick the ones that will work.

In re Wands comes along and says well, you just need to really show the process.
*
The written description requirement comes from 112, the same place as enablement.
For a long time, WD was mainly used only to make sure that people weren't trying to
stick "new matter" into old, rejected patents.

Gentry
The sofa with two reclining seats next to each other. During the course of the patent
prosecution, the language of the claims shifted. First, the claim said that the controls
must be on the console between the seats. But when they noticed that their competitors
were putting the controls further down, they got rid of that 'need to be on the console'
part of the claim.

The court says no, on-the-console was an essential part of the specification. Yet it's
nowhere in the claim. So that's why, in the court's language, it fails written description.
The mismatch between disclosure and claim.

Courts have backed off from this specific interpretation. Hard to figure out what's an
'essential' element of the specification, such that it *has* to be in the claim. A narrower
reading of this case would have it apply only cases where stuff in the specification says
"we're doing it this way because the other ways are bad," yet the claims cover the bad
ways.

Rochester
COX1 and COX2. We want a drug that only inhibits the latter. A narrowly tailored
compound. Rochester has two patents. The one that's not at issue is the one for the
method to figure out if something is selectively hitting COX2. A patented immunoassay.
The second claim, which is at issue, is the method of selectively hitting the COX2
receptor with a compound, proper. The end aim.
Court says the claim is invalid under written description, because Rochester hasn't
disclosed the actual compound that can be used in that method.

NB: Don't think about this in the same box as Gentry. They end up to be about different
parts of the requirement. We get a standard out of this--showing that you have
possession of the claimed invention.

You need (1) to describe the invention with all its limitations, and (2) to demonstrate that
you have possession of it at the time of filing. Here, part 2 is missing.

What's going on here policy-wise is a concern about scope and timing. University
researchers are good at finding out what's going wrong. Back in the day, the professor
would publish his findings, and the developers would read it and make a drug based on
it. (Missed a minute or two). The patent here would give the researcher a monopoly not
over a particular drug, but an entire class of drugs.

--Maybe this solves the Marcia Angel problem of me-too drugs that are made just
because there's market share to be tapped. A monopoly over the whole class make this
less lucrative
--But there are drawbacks, of course. You avoid the race, but then also the drug might
not even get developed in the first place. The risk of investment is less likely to pay off.
If there's a lot of cost involved with picking the compound to develop and refining it, then
we're granting a patent at a point when there's a ton of risk, and when there's a ton of
ideas that need to be developed that aren't located within the monopolist.

One of the core justifications of the patent system is that it's a big open market. There
are problems; go out and solve one and you get a reward. Not like the NIH, which picks
the problems beforehand. Upstream patents start looking like a privatized NIH sort of
thing, where companies are control of the solving of problems.

But the court doesn't say this. What does it say? What does WD give us in addition to/
separate from what enablement gives us? Say the enablement gives a skilled person
the knowledge of how to make A, B, and C. But the description doesn't cover C. What's
the benefit of rejecting the claim on C?

It's about scope and timing. Written description, unlike the other doctrines that are
manipulated for policy purposes, doesn't have a justifying backstory. For that reason, it's
very flexible. It's easier to get the results you want from WD than from more rigid
doctrines.

There's a big category of cases (not the Gentry style) where claims invalidated on WD
couldn't also be invalidated on enablement. The lawyers on these cases in some ways
really like WD. You described it, or you didn't. You don't have to get into the difficult
discussion of undue experimentation. But it raises a problem: if the policy
considerations at work here are the same as those in the enablement analysis, then we
might be getting the wrong result.

Lizardtech
At issue, a 'wavelet transform'--a way of compressing digital images. Basically, the
specify one way of making a seamless image, but claim 21 covers all ways of making a
seamless image. They also dropped the 'seamless' bit from claim 21--but that gets
resolved. It's obvious they meant seamless.

The infringing stuff gets seamless DWTs but doesn't use the maintaining updated sums
method. Court says WD and E both invalidate this.

The E reasoning is like this: You've described one way of doing this, but figuring out the
other ways would require so much undue experimentation. This is sort of like the
product-by-process claim from the second day of class. Trying to get to the whole class
of end products. This claim is set up as p-by-p, though. Anyway, court says no.

And it's also a WD problem. This seems intuitive. But at the same time, we often allow
people to claim a big area if you're the first person to figure out a way to get there.
Drugs, e.g. You claim X compound, and disclose a way of making it. Someone else who
comes along with a different way of making the drug still needs to go through you. How
is this different?

Well, this is a method claim, not a compound claim. The way of doing it is part of the
limitations.

Ariad
Looks a lot like Rochester. Ariad attempts to distinguish by saying look, our method
makes no mention of compounds. It's just the method. Fed Cir says no, of course you'll
need a compound. And you haven't described them.

But there's another differences. Ariad includes three hypothetical compounds.

Here, court is saying that one example of how to get somewhere doesn't give you a
broad method claim. Now--we don't know exactly what the rule is here. It certainly can't
be that you need to list all of the ways.
**
Aside from that initial Rochester distinction without a difference, there are a couple
important things here.
--Written description is measured from the time of the filing date
--Utility gets dragged back in. For that second compound, you needed to show that it
would work. Narrows the gap between WD and E

2/22
Enablement and written description requirements have a lot of bearing on other parts of
the patent. Today's issues--best mode and claim definiteness--don't.

Claim definiteness

Orthokinetics
Wheelchair that can deposit kid in car. Language about front wheels being "so
dimensioned" as to fit between car door and car seat. District court finds patent invalid
under CD. Subsequent people would have to measure their chairs for every freaking car
if they wanted to know they were infringing.

Fed Cir reverses. The standard is that someone skilled in the art can read and
understand what is being claimed.

Standard Oil
Struck down on CD. The troublesome language is "partially soluble." The patent doesn't
define it, and it has no standard definition.
*
Because we define inventions by the claim, we need to know what the claim says.
Otherwise, the patent is a land mine. You can't tell when it's going to go off. But CD is a
pretty narrow doctrine. Term has to be really ambiguous. "Insolubly ambiguous". When
it does come up, it's often for things like drafting mistakes.

Best mode

Chemcast - leading case

A grommet with varying hardness. Didn't supply the specific material and manufacturer
they use.

Two step test

1) Subjective - Did inventor know of a best mode?
(How'd they show that here? Well, if you're commercializing a particular embodiment,
that's evidence. And anything that makes it look like they decided x for a reason)
2) Objective - Does the disclosure enable someone skilled in the art to make the best
mode?

Before you apply the test, you have to answer the threshold question of what the claim
language claims. The claim language tells you what aspects of the invention are subject
to the best mode requirement. E.g., you don't have to talk about your great employee
incentive program here because it's not in your claim.
So Chemcast first argues that they don't have to disclose the best type of rubber they
used, because they only claimed the hardness. Court rejects this. You can't get around
the fact that you've claimed a material.

So onto the substantive portion of step 2. And court says "one skilled in the art simply
could not divine Rubright's preferred material hardness." Rejecting the argument that
they didn't know anything really about the material, court says you at least give the
hardness and the trade name.
*
Best mode, under step 2, has "an" enablement test embedded in it--enablement with
regard to the particular embodiment, as opposed to the general invention.

If you don't have a best mode, the requirement goes away in step 1.
***

NOVELTY

102 has a lot of stuff in it. First, we'll look at "true novelty," where the critical date is the
date of invention (as opposed to (one year prior to the) filing date). The other interesting
thing about the statutory bar is that the inventor herself can trigger it by disclosing too
early. Besides true novelty and statutory bar, there's derivation, which is simple--you
have to be the one who invented it.

True Novelty
Robertson
Patent covers three things. The fight is about this third fastening means.
PTO says this isn't new. Point of reference is 102(e), the anticipation, "each and every
element" ("all element") test: "each and every element as set forth in the claim is found,
either expressly or inherently described, in a single prior art reference." We'll worry
about inherent anticipation later.

Fed Cir says look, the other patent had no third fastening means. PTO and dissent
interpret the claim language differently, though. The claim language is about "fastening
means"--never says that they physical entities constituting the means have to distinct. If
they can be the same, then the older diaper could function the way described in the
Robertson patent. Now, the old patent doesn't claim that use, though it is discussed in
the specification. But for novelty purposes, we don't care.

But wait. Is a single example of something that fits within the claim language that
existed before the invention enough to kill the patent? Yes!
(Discussion of this proposition as strategy in litigation. In orthokinetics, say - if you can
find any wheelchair that fits between any car, you can invalidate the patent. Some
things don't always count as a 'prior art reference,' but a physical object always does.

The other important thing here is the genus-species distinction. When a patent claims a
whole series of embodiments (a genus), it only takes the prior disclosure of one of those
possible embodiments (a species) to invalidate the patent.
**
So how does In re Schreiber fit in here? Where an oil can top could do the same thing
this popcorn top is doing. The more general, older patent anticipated this more specific
one.
*
Inherent anticipation - prior doesn't explicitly disclose all elements of new thing, but all
elements are inherent in it. We don't want people to be able to claim a patent on what's
already out there simply describing in more detail.

Seaborg
Patent claims the new element. The claim is that this is a new element--it's not found in
nature. Now, technically, this can't be correct (supernovas?). So why don't we say you
can't patent this, but you can patent some sort of purified form or whatever (Clarke-
Davis).

Well, the fact that it existed in supernovas doesn't anticipate it, according to the statute.
Had to have been "known or used by others", under 102(a) (I think). But. A previously
made reactor did produce trace amounts of this element.

Novelty is a very strict, rule-based requirement. We really mean it when we ask if

something is new. But here, the CCPA rejects the PTO's rejection. How'd they get
around this? Well, the reactor people, first of all, didn't even know they were producing
this. And moreover, even if they did, that Am was undetectable. This rationale is related
to the doctrine of "accidental anticipation."

It's important to see that the scope of the underlying rule is broad, so that we need to
carve out this exception. The basic rule is that if something previously existed, even if
not explicitly and publicly recognized, then it's not new. It just has to have been there.

What's the flip-side of this outcome as a solution? Now, the reactor people are
manufacturing what the Am people have patented. Can they get an injunction? Maybe
not. There's supposed to be relationship between anticipation and infringement. What's
one is the other. So maybe what's not one is not the other. But nothing in this case says
that.

2/23

SmithKline
A patent on the anhydrate compound and a later patent on the hemihyrdate (easier to
package the hemi). When anhydrate patent runs out, Apotex starts selling generic
anhydrate, SK sues on hemihydrate. When Apotex makes the anyhdrate, SK says,
trace amount of hemi are produced, so they're infringing. But then Apotex says if that's
true, then your patent is invalid, because the old form had the new form inherent in it.

District court (Posner) rejects Apotex's argument. We don't know for certain, he says,
whether the production of the anhydrate back in the day actually produced the
hemihydrate. What Posner wanted to do instead was to answer the anticipation-
infringement symmetry question implied by Seaborg. If trace production of new in the
older thing doesn't mean anticipation, then can trace production of older thing in new
thing mean infringement? Posner said no in district court, but Fed Cir rejects.

Fed Cir says there was inherent anticipation. You don't need to have actually produced
the new thing--as long the enablement describes the new thing, then that counts.

Why doesn't Seaborg answer this case? Both involve trace amounts and lack of
knowledge. They distinguish it on the basis of probability. In this case, it's 'undisputed
that' the practice of the old patent results in the production of the new one.

There's a stretch going on here. The underlying principle is that you can't get a patent
for explaining the use/benefit of something that's already being used/benefited from.
That's being extended, though--see Atlas Power in page 3 of this op--once it's in the
public domain, even if it's not being used for benefit, it's in there, and you can't take it
out.

Hafner
His German patent, filed in 1961, doesn't disclose a use. His 1960 American patent is
rejected on 112. In 1964, US rejects again. They won't accept the German patent,
because it doesn't state a use, but they also say it is enough to anticipate the 1964

CCPA says yep, it's asymmetrical. Lower standard for anticipation than for enablement.
And court gives a reason. Based on the text of 102 vs 112.

So what does 102 require now? It needs at least to disclose the invention itself
(Missed a few)
*
So two developments so far. The first thrust--can't get a patent just be realizing why
something works--is covered by the basic inherency rule. But since this covers things
that are for all intents and purposes new, because the value wasn't recognized,
exceptions developed. Accidental stuff, Trace and undetectable stuff, etc. So an
enablement requirement grew--whatever the prior art reference disclosed had to enable
the public to make use of it. But then Hafner comes along and says that this enablement
requirement is different from how it is under 112. 112 only requires that you show us the
invention--not that its use has been demonstrated. So this creates the opportunity for
prior art which no one knew to be useful to anticipate a later claim. That's one
development. The other development, from SmithKline, is getting rid of the requirement
that the invention is being practiced. Essentially a repudiation of Seaborg (doctrinally,
not quite, because of a reinterpretation of the facts).

Abbott
At issue, an inhalation anesthetic. But big problem develops where it's toxic. Realize it's
the combination with Lewis acids Realize that you just need a little acid inhibitor; they
use water.

The prior art at issue is a method of purifying the chemical that involves saturating it
with water.
So, basic anticipation, we start with the claims.
--Quantity of sevoflurane? Check
--Presence of Lewis acid inhibitor? Check
--One of a group of LA inhibitors? Check

But the prior art, where water is involved, is just an intermediate step in distillation. The
object of that old patent is to purify the sevo--exactly the opposite of the objective here.

District court doesn't care. The question is, is it new. It doesn't have to be the final
product of a process to have been known and used before.

(Missed a discussion on the argument from the BMS case (page 3) that if something is
directed to a new use, there may not be inherent anticipation, or something).
**
Basically, a distinct trend toward formality, and concomitant strictness

3/1

(Missed first half of class)

Klopfenstein
Same invention, given in a speech. In Klopfenstein, they talk about the MIT case. Found
prior art in a speech, but there were accompanying printed handouts. "Printed
publication" requires something printed. Some sort of printed symbols.

Also, if it's available, but only in a way that you have to ask for it--like a thesis at a
university library--not so much.
What about a website. The Klopfenstein factors make it seem like this would be ok. But
how can a webcast count as "printed."

What about a company intranet? Generally, doesn't count. The company is generally
thought of as a single entity.

The thrust of the inquiry is public accessibility through reasonable efforts. And by
"public" we mean the target audience, the people who could understand it, not some
remote tribe. And that's not a policy question. It's not "why" is it publicly accessible. The
underlying logic is very formalistic

102(e): A patent app is treated as published as of the day of its filing, even though it's
not really published til 18 months later, usually.( apps are published, even if not
granted!) If a patent isn't granted or published, it doesn't count as prior art.
**
Switching to
Priority
If an invention has more than one inventor, who gets the patent?

In almost every other country, the rule is simple: whoever files first. Here, we want to
know who invented first.

102(a) cares about the invention; 102(b) doesn't. If you're in the window of time that's
less than a year prior to filing, and there's a disclosure, you have to inquire about the
invention date.

Mahurkar
Third party publication--a catalog--is said to anticipate the invention. Three months
before the date of filing.

The date of filing is the 'constructive reduction to practice.' But of course, the date of
invention may well be before the filing date. So what Mahurkar is trying to do is to push
this date back.

3/2 - 102(g): Priority

Barbacid
Brown files first, April '90, then Barbacid, May '90. For our purposes they claim the
same thing. Barbacid, claims that they reduced the invention to practice in March '90.
So now Brown needs to show an earlier priority date.

Brown wants to talk about two experiments in September '89. They throw out the first
experiment because there was no assay involved. But Brown says that in the second
experiment, he did it.

(BTW, how did Barbacid get that March reduction-to-practice date? To get constructive
reduction to practice, you file a patent. But you don't get it if that patent isn't enabling
under 112 or if the inventor doesn't know it satisfies the intended use (the inventor
doesn't know whether it works. The Estee Lauder case). For regular reduction to
practice, you need to 'do' the invention with all its limits and elements, and you need to
know it works, and you need corroboration.)

Without evidence otherwise, the date of conception is the date of reduction to practice,
either actual or constructive. As in this case. Brown here is trying to show an earlier
date of conception. Conception is a
"definite and permanent idea of the complete and operative invention, as it is
thereafter to be applied in practice."
Needs to cover all the elements.
The test is similar to enablement, in that it's about whether it would require
unreasonable experimentation to practice the invention from what is known.
You can't "inherently conceive" the thing, and so is more about what's actually going
on in the mind than anticipation. The way you have the invention conceived in your mind
needs to be the way it's actually practiced.

Burden of Proof issues. Here, Barbacid--the junior party, the later to file--carried the
burden to show by preponderance an earlier reduction to practice than their filing date.
Brown has the burden of proof by preponderance of the evidence to show their earlier
date of conception. In most cases, if you want to show an earlier date, you have the
burden, by preponderance. One exception: If junior party filed after senior party was
issued or published the patent, then must b clear and convincing.

Back to the case. What evidence does Brown have? Well there's these lab notebooks.
But the lab notes are uncorroborated. But the lab notes, court says, don't need
corroboration. Physical evidence can speak for itself. But still, you want more. Brown is
trying to take an invention from somebody else here after all.

We also have inventor testimony re the second experiment. But it's 'uncorroborated'
says the court. What about the notebook? You can't corroborate your own story.
Uncorroborated oral testimony doesn't get in the door. But Brown still has the testimony
of his fellow researcher. But there's a further twist, because the colleague didn't know
about the invention until later. But still it was before the other reduction to practice, so
still good from Brown.

Why didn't this case end four pages before it did? Well, just getting the lab notes to be
admitted doesn't mean that the evidence will convince the jurist. The evidence has to be
weighed.
So the court accepts conception. But, since they reduced to practice first, Barbacid can
still say that Brown wasn't diligent from the point of conception. They did say that (note
7, pp. 450-451). You need "reasonably continous activity toward reduction to practice."
Trying to commercialize it doesn't count. And if you change the elements, then there's a
new date of conception. The PTO is surprisingly strict here, micromanaging.

Peeler
A hydraulic fluid used to prevent gas pockets. Miller comes up with this basic idea in
March 1964, and submits a form to Monsanto's patent office in 1966. Company labels it
'ready to file'. But then, due to personnel problems, four years go by with no action.
Patent finally gets filed in 1970. But by then, a competitor, Peeler, had filed the same
patent two years prior.

Peeler appeals under 'abandon, suppress, conceal" clause of 102(g). But first, they try
to argue that Miller and Monsanto didn't actually reduce their invention to practice
before Peeler filed. If they can establish this, then, even granting that Miller conceived
first, they have a slam dunk: no way in hell that four years counts as reasonable
diligence.

Court doesn't buy this first argument. Peeler says that Miller didn't run the experiments
with jets, so no reduction. Court dismisses easily, because Miller had shown it worked in
a hydraulic system in general, and the claim is about that, not jet engines specifically.

The next argument--still under the argument that there's been no reduction practice--is
that Miller had simply abandoned an experiment, thinking it had basically failed.
Abandoned experiment arguments are a way of rebutting a claim of reduction to
practice. But the court rejects again, under a sort of rule of reason, let's figure out the
right answer kind of way. We can see what happened here--nobody thought there was a
failed experiment. The thing worked, the patent was sent off, and then it died. Not an
abandoned experiment within the meaning here.

So the final argument, and the one the case is about, is that Montano suppressed this
patent. But, huh? Clearly there wasn't nefariousness here.

Policy: Don't want people doing just enough to win an interference, then sitting on
patent and not practicing it until forced. But that doesn't seem to apply here. The story is
one of ineptitude. So what's the reasoning? Well, at a certain point the delay is just too
excessive, court says.
Why not shape the application to fit the policy, though? Sam: People will just make their
legal departments look inept. Requiring intent is bad as an evidentiary issue.

Roin: If you can show intentional delay, that is much stronger, but even long delay
without intent is prima facie evidence of suppression.
But if you've delayed a long time, but then pick back up before the conception/reduction
of the competitor, you can still get the patent. It's not like punitive damages or
something.
*
Next week we move to 102(b). The biggest difference from (a) is that is that it provides
a hard and fast date for when the invention needs to novel: one year before the filing
date, if there is any public disclosure. A grace period.

The all-elements rule of 102(a) and (g) get relaxed in 102(b). And there's a difference as
to what constitutes prior art. In (b) we have prior sale.

For M: Statutory bar-public use and SB-on sale material.

3/8

Egbert
Court doesn't buy argument that the use of the corset pin was not public.
No restrictions were imposed on the use of the thing. . .

The story here can't be withdrawing something from the public domain, because only
one person was using it. It has to be something like the encouragement of patenting
early.

Moleculon
Rubik's cube. Grad student Nichols comes up with the idea, shows a construction-paper
model to his friend.
Goes to work at a regular chemical company. His boss Obermeyer sees the cube and
says let's patent. About a year later, after Obermeyer makes some attempts to
commercialize, they file a patent.

Nichols sues CBS after they come out with the Rubik's cube. CBS's defense is 102(b).
They go all the way back to his grad school stuff, but they only need to go back to
basically just before Obermeyer got into the office. So they can't talk about Obermeyer
shopping it. But they can talk about Nichols bringing the thing into his office all the time
and putting it on his desk.

But this isn't a public use, says the court. Well, to begin with, he hadn't commercialized
it. (That'll be important later today). Then the court says that Nichols retained control
over the puzzle's use and the distribution of information concerning it at all times. Sure,
there was no confidentiality agreement, but that's not determinative.

CBS also tries Nichols' assignment of the IP rights in the cube as a "sale" under 102(b).
No: Selling the invention itself can certainly count, and licensing can count, but
assigning rights doesn't count.
*
This is the image of the independent inventor that courts love. But aside from that, is
this a compelling case for granting a patent? It's ironic, because we value this type of
invention story, but the independent, unincentivized nature of it weakens the case that
you need a patent to get the invention.

But sometimes inventions need a champion--someone to believe in it and push it along,

esp. within a company. So maybe awarding the patent would make there be a
champion (or something).

Moreover, the Arrow paradox: Secrecy is not conducive to "vertical transactions" (sell
the idea to a company, who develops it). You want to be able to make sure they're not
just going to take the idea and run with it. How do you sell your idea without explaining
how it works.
--> This is about coordination of the investment effort, (so is the champion thing), rather
than about encouraging the invention itself, which we classically think of as the metric of
patent justification.
*
Look at the facts of this case, though. These justifications don't work out here.
Obermeyer failed at commercializing it. CBS, without a patent (but with a trademark),
develops the idea, presumably from a second inventor. So looking back, was there (in
contingent fact) any reason to grant a patent?

[As an aside, relying on marketing/trademark for return on investment is basically giving

a starting advantage to big companies. Marketing is hard and expensive.]
*
The pushback to a patent in all these args is also a patent race argument. You're
creating a race that results in duplicative R&D.
**
Netscape
System access thing. Successful testing, handing off to patent office in 1990.
Demonstrates to a couple frineds 1991. Patent department at Berkeley files in 1993.

Why is the showing to friends potentially a public use? Well, first, and importantly, there
was no confidentiality agreement. (Learned Hand says that a confidentiality agreement
can be implied, btw). Court rejects implied confidentiality argument

(Missed one point)

He also didn't monitor the testing of it. He just sort of turned it on and let it run.

So Moleculon is like an exception to a confidentiality requirement, because control was

retained.

A broader "totality of the circumstances" test emerges from the case, with three
aspects: the nature of the activity that occurred in public, the public access to and
knowledge of the public use, and whether there was any confidentiality obligation
imposed on persons who observed the use.

They also (rare!) consider policy: "The public use bar serves the policies of the patent
system, for it encourages prompt filing of patent applications after inventions have been
completed and publicly used, and sets an outer limit to the term of exclusivity." We want
you to file early, because if you wait, you're basically extending your patent term. What's
also going on is a general preference for filing patents.

But what's missing here? Why do we have the novelty requirement. So that we're not
pulling stuff out of the public domain. So an important inquiry is whether the public
actually has access to the invention. And that's missing from the policy considerations
here.

Metallizing
What was the patentee doing before they filed the patent? They were selling the metal
on which they had used their eventually-to-be-patented metallizing process. So, the
process is being used commercially but is also secret.

[Remember, they're claiming the process, not the spray itself. It would have been easier
to invalidate under public use if it was the spray itself that was claimed].

Hand rejects, again on a policy against extending patent terms. Combining secrecy and
legal monopoly. And you also have the policy against withdrawing stuff from the public
domain. This second policy is present here, but only indirectly.

Gore
This process for stretching teflon and making some other shit out of it. Gore discovers it
to and makes Goretex. Years prior, in New Zealand, Cropper was doing the same
process, but making a tape out of it, instead of Goretex. The process is protected by
trade secret. (This is important. Under 102(b), public use/on sale only counts in this
country. Publication--which trade secret is definitely NOT--could be anywhere). Made a
machine that completed this process, described it and offered it for sale to an MA
company.

The public use question is whether having the machine was a public use

Two things going on: (1) Public didn't have access to the process of the technology itself
(like in Metallizing), and (2) It wasn't the inventor that was secretly commercializing the
process, it was a third party (distinguishing it from Metallizing).

Why do we make this distinction? We want to penalize people who practice their
inventions in trade secret as opposed to in patent. The justification has got to be that we
don't want people protecting their [things? processes?] via trade secrets.
**
What would stop a person from getting a patent on the Coke formula? How would that
differ from Gore? Because the formula is embodied in the product. On the other hand,
their 100-year-old process for making cans would be patentable, because it's not
embodied.

OR SOMETHING!

3/9
The interesting thing about 102b is that it's come to be more about whether the inventor
could have filed than whether the public truly had access to the invention. We see this
most in the distinction between the third parties and the inventor herself. If a third party
has a process they're selling the product of, that doesn't count as prior art for the
patentee; if the inventor is doing that, that does count. This rule is different for
processes than for products. Even a third party's sale of a product that's being patented
would count as prior art. Then there's the embodiment stuff (see Coke example above)

Today, the on-sale bar and experimental use

Pfaff
Computer chip socket. TI contacts Pfaff with something it wants. Pfaff draws some
sketches, etc., TI likes it and puts in a purchase order. A year and nine days later, he
files his patent application. Was the purchase order, based on the sketches, a sale of
the invention?

Court says that under 102(g), the date of conception is really the date of invention.
Reduction to practice is important, but blah blah. But that's not quite what they say is
needed to pass 102(b) bar. The standard is a description 'sufficiently specific to enable
a person skilled in the art to practice invention' without unreasonable experimentation.

Why not just make reduction to practice the hard line? Because it would give inventors
incentives to just not make a prototype--that way that can commercially exploit for a
long time, then eventually file a patent. Not a terribly strong argument, since a lot of
commercial exploitation depends on a prototype. But the point is correct--rtp would
allow at least some stretching of the patent term.

The two part test that comes from here for on-sale: 1) must be the subject of a
commercial offer for sale, 2) must be "ready for patenting," either by proof of reduction
to practice or by drawings/descriptions that satisfy that standard above.

Note that it's offer. Not necessary to have buyers.

What about a letter/ phone call that says "I have this invention, here's what it is - contact
me for details!" Think back to contracts. There's something missing here--the price.
(And more generally, it's an offer to start negotiating, not an offer for sale).
And the focus in these cases is on the seller, not the buyer. You want to know if the
thing was put up for sale.

The implication of the second part of the test is that you need some evidence for
use/utility, because that's part of enablement (at least for biotech stuff?)

When we use enablement, as we're doing, things get complicated in a way they
wouldn't with conception or reduction to practice. Has elements of both these things.

Abbott v. Geneva
Some other company was selling "Form IV" of the drug at issue.
Abbott had started patenting other salt forms of its original drug, to defend itself against
generics. They started that in October '94. So CD is October '93. The generic
companies had been buying huge batches of the compound, in undifferentiated form,
before that. Some of that was Form IV. Do those prior sales constitute a prior sale?

Abbott argues no. The buyers and sellers of those batches didn't know it was Form IV.
Court rejects this under the Pfaff test. If the standard of Pfaff had been conception, the
prior sale wouldn't have counted. There wasn't "all elements" satisfaction. So how does
the court get around this? That neither the buyer nor the seller knew what it was that
they were buying/selling? Court simply says it doesn't matter. The question isn't whether
you knew it was all there, it's whether it was all there and being sold.

This imports all the inherency stuff from our discussion of 102(a) into 102(b).

(Missed a few)
**
Experimental use exception

City of Elizabeth
Inventor invented a form of pavement to be used for roads. Tests it on this road that he
has an ownership interest in. 75 feet of road near Boston. For seven years prior to his
filing. When he finally gets a patent and sues for infringement, he sues. Defends against
public use, via experimentation. This is a judge-made exception.

The way we doctrine-ize the experimentation exception now is by saying that the
invention is not reduced to practice, it's not ready for patenting, while it's being
experimented on. So more of a negation than an exception.

So how do you avoid people just 'experimenting' for a long time? Well, first, you don't
need to be changing the claimed elements, but you need to be testing them, and testing
them in a serious way. And the testing needs to be bona fide (so intent is important).
And you need to have the invention under your control

It's still the case, since this is public, that somebody could note the pavement, start
using it in their city, then get sued seven years later. A pulling from the public domain.
Why is this ok? Seems like because this is the only way to run this test. But, importantly,
court doesn't say that if it was possible to do this in secret, the experimental exception
wouldn't apply here. And that's the rule--just because secret testing is possible doesn't
mean that public testing doesn't fit under 102b in a given situation.

But the court does mention explicitly some things you can't do. Offering for general sale
bars you from using experimental use exception. And you can't really let the invention
be used by other people.

Lough
Makes a better boat sealant thing, tests it on his own boat and give prototypes to a
handful of friends.

Court focuses on the failure of the dude to maintain control over the inventions and
even to monitor the use in an experiment-y way.

Dissent says it's one thing to fault the experts for not following the rules, and it's one
thing to fault someone who commercializes an invention prior to patenting it (what we're
trying to avoid), but it's another to fault the amateur who doesn't know the rules even
when he's not really even commercializing it. This is really a matter of fact that should
have gone to the jury.

Explicates a 'totality of the circumstances' test that has a list of circumstances (bottom
562-563). Most important is control.

Baxter
Centrifuge for heart stuff. The potential problem is that the same thing had been used in
someone's research lab prior to the critical date. And that dude is really open about it.
Has the centrifuge in his room, anybody can come and see it, no conf agreements, etc.

This case shows how much of an outlier Moleculon is.

(Missed a couple)
3/21

OBVIOUSNESS

First - Why do we even have this requirement? The first instinct is that for inventions
that are obvious, you're going to get them anyway. You don't need to reward a patent to
spur it. (More fleshed out: Take a high risk innovator. 9 times out of 10, his gamble
doesn't work. When it does work, you don't want competitors to be able to hang around
not innovating and then get into a price war with you, making your innovation not worth
it). So you do look at the risk. But interesting, in obviousness, you don't ask about R&D
cost. Given this paragraph's story, you'd think cost would matter.

And when you don't need a patent, there are a bunch of costs that now have no
countering benefit. You're taxing subsequent innovators, for one. And direct deadweight
loss of higher prices. And administrative costs.

What is the counter to this? To this argument that if you don't need a patent to get the
invention, don't patent. The counter is that you still need the patent to get the
development and commercialization of the thing. Also prospect theory, the Ed Kisch
stuff (?). Granting a patent prevents wasteful repetitive R&D, allows one party (the
monopolist) to coordinate R&D efforts.

So the debate really hinges on this the patent being necessary to call forth the
invention.

Graham
At issue, a clamp involved in plowing. It's an improvement (the hinge is improved) from
Graham's old patent. The comparison is between the new patent and the old patent.

S.C. starts off with this very long discussion about Thomas Jefferson. Because the
statute is unclear or something. So they go back to the IP clause of the Constitution
(Art. I cl. 8). What they eventually do is leave aside the 'flash of creative genius'
standard for a lower standard. They're doing this in part because people really hadn't
liked the Kuno standard, this seemingly ratcheted-up standard. (or something. This is
wrong. Read the case.)

So Jefferson is being used as a sort of anti-patent voice by the SC. A great hesitance
toward granting monopolies
*
Congress, SC says, isn't allowed to expand patent protection if they're not good for the
useful arts. Raises interesting question of who gets to decide whether granting patents
in a category of case promote progress. I mean, this is a broad policy judgment, isn't it?
Roin compares to CTEA, Eldred v. Ashcroft. That was the main constitutional challenge
in the IP clause. This stuff hasn't been tested, is probably dicta.

The only rule we get out of this, and again it may be dicta: You can't patent stuff that's
already in the public domain, constitutionally. But as a constitutional rule--something
compelled by the C, as opposed to developed through case law-- this is worrisome.
Development and commercialization thing might make you want to patent something in
the PD. Or a worry about organizing the market (?).
*
As for the doctrine, we get a three part test for obviousness
1 What is the prior art?
2. What is the difference between claimed invention and prior art?
3. Who is the person having ordinary skill in the art? Is the difference obvious to him.
+ "Secondary indications of obviousness": Long-felt need, failure of others, something
about the market I missed.

Really important: When the court says look at the prior art and find the difference, they
really mean it. These steps have a fairly mechanical aspect. You need to really do it
explicitly. On the exam.

Here, they decide the invention is obvious. The court doesn't articulate it, but
*
Benefit is important. If there's no benefit to it, if it's just a cosmetic shift, if it's just purple
Lipitor, then it's obvious. It's hard to draw that out of the opinion, but that's the
underlying logic.

Dembiczak
Assigned because it's important to see what the court in KSR was reacting against. At
issue, garbage bags used for Halloween decorations

Again, the details of the differences matter. Here, the first difference from the prior art,
the arts and crafts thing, is that these are plastic. And second, there's a size difference.

(The majority of non-obviousness opinions go through the Graham test in name, but
then never ID the person of ordinary skill, the third step.)

Court says, trash bags are big and plastic, so that's all there in a prior art. And the
putting the face on thing, that's in the arts and crafts prior art. So you have different prior
arts that together cover all elements. Is combining these elements of prior arts obvious?
The question is whether there's a teaching, suggestion, or motivation in the prior art to
combine them. Court mentions three things you can look at, and focuses on the first one
(go to the case). All the while, you have to make sure not to fall victim to the hindsight
bias--everything looks obvious in retrospect.
MISSED TUESDAY (TYLER'S NOTES)

3/29
With obviousness, first (usually, not a purple pill situation) you get the prior art
references that add up to all the elements. Then you need an obviousness "story"--how
was the idea to combine the refs obvious?
Here we have the TSM test as guidance. First you find the thing that pushes you to
combining them, and then you look at what the PHOSITA would do. Jack-o-Lantern
case

In KSR, SC told Fed Cir not to be so rigid about the TSM test. What does that mean?
Something about the specificity of the prior art refs. Or something. Predictability and
obviousness of result are themes that run through much of the case.

Pfizer (already did half on Tuesday)

Pf has its chemists find the best salt form for this pill. Then they patent the salt form (in
addition to the active ingredient, patented before). Was the salt form obvious given the
core compound?

Court introduces this second step to the standard non-obviousness test, this
"reasonable expectation of success" thing. In biotech, courts do this. Two parts:
compound itself would form, and would work for purpose it was being designed for.

The analysis isn't much different from the basic obviousness inquiry, but it's more
particular. We want to know whether there was a good reason to think that changing the
thing to its salt form would solve this problem they had. You're still talking about the
combination of prior art references. But there's this 'obvious to try' aspect. You have this
problem. Sticky pill. And there's a certain number of salt forms you could try. And you
pretty much know one's going to work. Goes along with the KSR court in this way. Here
they call it 'routine experimenting', but the idea is the same.

When you're using 'reasonable expectation of success,' you're looking at the patented
thing. With 'obvious to try,' the gaze shifts the the prior art refs themselves.
*
Then the court gets to the secondary considerations.
Here, the court considers unexpectedly superior results. Why do we care about that?
Remember, a prima facie case of obviousness has already been established. If it was
obvious to combine the refs, then wouldn't we have gotten the thing anyway?
(Obviousness is a way to not reward things that would have come anyway.) (**Formally,
there are a few ways to answer this, always:**
--Disclosure. Maybe without the patent, there'd be no filing, and other people wouldn't
know about the benefits
--Development and commercialization. Separate, in this story, from simply calling forth
the invention.
--Coordination. Prevents competitors from engaging in duplicative R&D and R&D
around the whatever. Makes one coordinating party, reducing wasteful R&D.

There is always pushback to all these points. Disclosure can be effected often simply
from sale of the invention. And re coordination, maybe giving an exclusive right to one
person, that person would mess shit up. Granting a patent early on that covers a range
of further development may simply deter other good ideas relating to that invention from
being developed.

So, again, you can say this stuff about unexpectedly superior results. But there's
something more here. Think of the Thomas Edison thing. It may be obvious to try all the
different high-resistance fibers. And one might work way better than others. And you
want somebody to figure that out. Increases incentive to go through and find not just a
thing that works, but the thing that works very best. It encourages people to do the
legwork (or, alternatively, to get lucky)

Unexpectedly superior results is a big deal for biotech companies. A big way to try to
fend off obviousness claims.

Ormco
At issue, this package of multiple retainers that progressively move teeth, plus
instructions as to the order in which to wear them. Invisalign.

The patent claims (1) three or more appliances . . . to progressively reposition teeth, (2)
instructions regarding order of use, and (3) a single package for provision of the
appliances to the patient.

There's this prior practice though. In that one, patients were getting the retainers one at
a time. And there isn't good evidence that they had instructions.

But the FDA says you have instructions when you have medical devices. That counts as
a prior ref

Before we do secondary considerations, let's run through the core obviousness inquiry.
We've already done Step 1 - identifying the scope of the claimed invention, plus we've
found the prior art refs--the old invention, the FDA guidelines, and identified the
differences. Then court says it's obvious to combine the two to come up with
instructions, and to put it all in one package. This is edited out of our opinion, but why
does court say that's obvious to do, prima facie? Look back at KSR. That would allow to
say predictable move, predictable results. But that's after this case. Here, you can use
the motivation language.
*
So, on to secondary considerations. Objective indicators of obviousness/non-
obviousness.

-Commercial success might be an objective indicator of non. The idea is if there's a lot
of money to be made, and it was obvious to do, someone else would have done it first.
But this has been criticized a lot. CS might have a lot more to do with marketing than
with non-obviousness of invention. So courts have required a connection between the
claim and the commercial success. There's this nexus requirement for all of the
secondary considerations.

Nexus: Needs to be due to one or more of the claimed elements of the invention, and it
has to be due to the novel component

Here, the commercial success is largely about aesthetics. Align concedes this, but says
that success is also due to the ease with which the patient can use the stuff--no dentist
trip, only one package, etc.

Ecolochem
A way of making deoxygenated water. E was hired by EPRI just make this
deoxygenated water. For the plant's "make-up water". When EPRI finally finishes its
construction or whatever, and they have their own water purification system in place, E
sues EPRI saying that EPRI is infringing on E's deoxygenation system. EPRI says it
was obvious.

With respect to claim 20, Ecolochem conceded that it was prima facie obvious. We pick
up at the secondary consideration analysis. Long-felt but unsolved need, commercial
success, failure of others, (these past three were the Graham factors) simultaneous
invention, teaching away, copying/acclimation

Long-felt: The argument that this doesn't apply is that it was a short-felt need, just
arising as a result of the regulation. Eco says no, the reg actually reflected the already
long-felt need. This doesn't stand up, especially in light of clear error review.

Commercial success: Eco's stuff is admittedly good, and there's a nexus. So Eco wins
this one

Failure of others: (talking to Kristen)

Simultaneous invention: If you can show SI, that's an objective indication of

obviousness. If a bunch of people come up with something at roughly the same time, it's
more likely to be obvious. But SI is only an obj ind of obv if PHOSITA or lower is the one
who simultaneously invents. So an expert inventing it simultaneously won't count for
this.
(Side discussion about setting PHOSITA level. Remember, it's also relevant to
enablement. So the higher the level, the less easy for someone to understand the
enablement. And enablement relates to scope. A high level skill in the art gives you a
wider disclosure)

Teaching away: Not particularly important in this case

Copying/acclimation: Copying is a mixed signal. In one way, it's like commercial

success. But it also could indicate that it's an inevitable thing everyone was moving
toward. But direct evidence of copying can be really helpful for non-obv

Generally, long-felt need and failure of others are useful for non-obv. And commercial
success, when there's a clear nexus, can help a lot. SI isn't used that much, and
teaching away is more common in the prima facie analysis. Copying is pretty rarely
used.
***
PRIOR ART
What counts as prior art for the obviousness analysis? 102(a), clearly, and 102(e) and
(g). These are uncontroversial, but some people still complain. Because patent that are
filed might not be published til 18 months later, someone's invention could be called
obvious even though inventor had no access to obvious-making refs. But if you look at
the purpose of obviousness--to filter out what would have come anyway--it seems
reasonable to do this.

Oddz-on
A weird question in prior art having to do with derivation. Design patent, interestingly. An
invention, not the same, but pretty close, was disclosed to the inventor of the Vortex
beforehand. Q: Does secret prior art, available only to the inventor and not to the public,
make an invention obvious?

An intuitive answer would be "no, it doesn't." But 103 has this extra paragraph that was
added to protect 'team' inventions. And as a matter of statutory construction, are forced
to conclude that secret disclosures do count as prior art.

In re Foster
Does 102(b) stuff count as prior art for 103? This is the statutory bar material. Stuff that
happened more than one year prior to the filing date. So, very often after date of
invention. And 103 defines the inquiry into obviousness with relation to date of
invention.

This court, though, says, that 102(b) contains in it an obviousness inquiry. Part of 102(b)
is strictly about anticipation. But court also reads 103 into 102(b). Why should it matter
whether it's one ref or a combo of refs that anticipate. That's how they say it.

3/30

Clay
So we have the Graham analysis. Two prior art references, and the differences between
them and the claimed invention. Now, is the invention obvious, given the combination of
the references? How do we tell. Would it be obvious, that is, to take the Herrington
invention and change it by using the gel that's used to fill up gaps in an oil well.
--TSM. Where would we look. You could look at PHOSITA or nature of problem to be
solved, but usually you don't. Usually you look at the prior art reference and what it says
about the product's characteristics. What purpose they serve. That would give you
motivation to combine.

But we don't even get here in this case. Prior art doesn't even count for purposes of 103
if it's not analogous. And analogous gets split into two parts:
--Prior art is in same "field of endeavor," OR
--Prior art is "reasonably pertinent" to the particular problem which the inventor is
trying to solve.

Court's trying to get at whether a PHOSITA would look to prior art of this nature in trying
to solve the problem in the second prong. The first prong is based on the assumption
that you're familiar with the work in your field. Partly attenuated, b/c in actuality there's
constructive knowledge, rather than actual knowledge. None of this is particularly
precise
**
So, done with standards of patentability!

INFRINGEMENT

As a procedural matter, claim interpretation and infringement analysis are some of the
first things you do. A famous SC case called Markman said that claim interpretation is a
matter of law to be handled by judges.

The claims delineate the metes and bounds of the invention. To determine infringement,
you look at the allegedly infringing activity and decide whether it fits within the
abstraction of the claim. It's the reverse/equivalent of the 'all elements' rule of novelty.
Infringement activity must meet all the limitations of the patent claim. I.e., if the activity
had happened before the invention, it would have anticipated it.

So Markman sort of dragged claim construction into the light of day. And it turns out it's
a nightmare. But claim construction is very important. It's important for others to know
exactly what they can and can't do in the market. The inquiry turns on how a PHOSITA
would interpret the claim.

Phillips
The main case we have here.
At issue, these "baffles" that attach to walls to protect them against various stuff. It's for
prisons.

There was this circuit split at the time. Sort of intrinsic evidence vs. extrinsic evidence,
the claim and specification itself vs. dictionaries. The district court and the appeal say
no infringement, but then it's reversed. But what's more important than the holding is
how they interpreted the claims.

A dictionary rule is more like a rule. Ostensibly lends itself to an open and more
definitive way of giving scope to the meaning of claims. The problem, though, is context,
of course. Things can get arbitrary quickly with a dictionary--and often overly expansive.
Not always broader, but often. Ends up making patent scope more uncertain. And since
patent litigation is so expensive, leads to patent trolls.
*
The district court--which both the appeals and the en banc overturned--interpreted the
baffle claim as a "means plus function." The classic example is, say, a fastening means:
"a means of fastening a to b." Rather than describing the thing, you just give its function.
The appeals courts say no. M+f doesn't apply when there are structural limitations in the
claim. And here, "internal steel baffles" are mentioned--clearly structural limitations.

So this isn't means+function. Which means you have to do some serious claim
interpretation.
Court: Start with the claim language.
In the background is the question of whether the right-angled structures--the allegedly
infringing thing--fits within the claim language.
(Remember, comprising/consisting/consisting essentially of. "Comprising," used here, is
an open transition--even if there's other stuff there, I'm still claiming that).

Here, the definition of "baffle" is conceded--a thing that impedes flow, or something
Then you have this part of the language that says "extending inwardly from the walls".
What does that tell us? Roin: Just its directionality. So the claim sounds pretty favorable
for the plaintiff. What can the defendant hook onto? Roin's instinct: Because baffle is
defined in terms of impeding flow, stuff at 90 degrees can't count as a baffle.

Before we get to the specification, we also have canons of interpretation. Same terms
are used in same ways. And claim differentiation--make dependent claims add
something. And the validity canon (not used often). Not hard and fast rules. And there
are more, see book. Here, there are dependent claims talking about angles (I think?),
implying that...something.
It's better for defendants to win on infringement, rather than to wait for novelty/non-
obvious, because since after Markman, infringement is considered way before.
*
The specification. It talks about the baffles being good not just for structural support, but
for protection from projectiles. So what are supposed to do with this information. Well,
we're DEFINITELY NOT supposed to add limitations to the claim based on the
embodiments described. That is, if the claim doesn't say "stopping bullets," you don't
read "stopping bullets" into the claim. On the other hand, the specification can help you
interpret the claim. It's a hard distinction.

FOR M: READ CLAIM CONSTRUCTION PROCEDURES, AND CORRECTIONAL RE-

ISSUES. DON'T KILLSELF ON LATTER. HE WILL LECTURE.

4/5
(Fifteen minutes late)
You look at claim language, specification, prosecution history, and extrinsic evidence
when doing claim construction.
(Missed claim language)
Things involved in the inquiry of looking at the specification
-Lexicographer rule: Patentee can define things in the specification
-Disclaimer: patentee is allowed to disclaim some part of the scope (Gentry Gallery):
(dedication to the public: related concept where you disclose something but you don't
claim it)
-Interpret claim in light of patentee's use: lighter version of lexicographer rule
-Look at the purpose of the invention
-Central taboo of specification interpretation: never read limitations into the claim from
the specification. The spec will have particular embodiments. Be careful not to take
those and say you're limited to that. Otherwise, patents will serve as just a minor hurdle
for competitors to get around. One way to do this is to keep in mind that you're trying to
figure out the claim language.

Prosecution history
-Record of interaction between patentee and PTO can be used to interpret claim
language. Not used all that often, usually ends up as subsidiary evidence.

Extrinsic evidence
-Everything else. Whatever else is in the patent, and dictionaries, treatises, expert
testimony, etc. Technically the least important, but often used by judges.

So that's an overview of claim construction, now back to

Phillips (steel baffles)

The specification talks a lot about the main purpose of the baffles as obstructing flow,
and 90-degree baffles wouldn't do it. How does court get around this to support a broad
claim construction? Well, first you'd mention that just because an embodiment has
certain features doesn't mean you read those features into the claim. A major rule,
remember. But beyond this, they cooked something up. Namely, there is another
purpose mentioned in the specification, the compartment-creating function. And 90-
degree baffles can perform this function.

Unique Concepts
At issue, this fabric frame for do it yourselfers.
Claims have preambles at the very beginning, and then a transition, and then the
elements. The preamble to this claim (p. 814) says right away "an assembly of border
pieces." That's what it is. What it does comes after.
Normally, preambles don't serve as claim limitations in claim construction. But they can
help you in interpreting the rest of the claim.
The transition is "comprising," which is open. Additions are still infringing. The thing
comprises "linear border pieces and corner angle border pieces." That's it. Everything
that follows that tells us more about those two things, which are the invention.

(IMPORTANT side note: means plus function elements have to say "means of" and give
no structural limitations)

So how do you get the assembly of linear pieces and mitered pieces as falling under an
assembly of "linear pieces and corner border pieces". The story would be that the
corner border pieces don't have to be pre-formed.

(Missed a few)
There's also this idea that they may have disclosed the mitered pieces but not claimed
it. Mitered pieces are in the specification. But, court says, they were dedicating it to the
public--they didnt claim it. Maybe because they just thought that aspect wasn't good.
Here, the mitered pieces are hard to put together. And the invention is for a do-it-
yourselfer.
There's also the aspect that a claim was dropped during the prosecution history, as an
improperly dependent claim. And we don't know, but this could have been because the
broader claim didn't include the more specific one.
*
What do we think about the outcomes of these two cases? I mean, the competitor in
Unique sort of just invented around by splitting the piece. They didn't really do anything
socially useful. And they can't even really claim to have come up with it. Unique took all
the risks, and blah blah blah. But the counter is hey, if they wanted to cover the mitered
pieces they could have claimed it. And generally, competition is good, and we don't
really like huge patent claims, because they stifle competition. Again, this balance
between encouraging innovation and foreclosing competition. And back to the if you
wanted it, claim it point. Court's saying hey, be clear. Don't make us go searching. So
we balance between broadness and narrowness, and a balance between a demand for
clarity and considerations of "fairness" (i.e. preventing design-around).
**
On Cyber Corp and Phillips dissent:
Markman is where this all started. Claim construction is a matter of law. The theory was
that it would increase predictability. But there's no deference to district courts. And
Cyber Corp says there aren't even subsidiary questions of fact that are deferred to.

4/6
Yesterday we were doing literal infringement. To infringe you need to meet all limitations
of the claim. For that, you need to know what the claim says. Hence, we were doing
claim construction.

Now, quickly, re-examination and reissues. The conditions under which you can correct
a patent. Skim this in the book. 253 is the disclaimer provision. You can say
"nevermind" about a particular claim, and it will disappear. 254-256 are about mistakes.
254 is about PTO error; 256 is about adding an inventor.

255 is a little more. It allows you to correct errors as a patentee that are your fault.
Covers mistakes/typos that are apparent on their face and that anyone would know how
to correct.

251 and 252 cover a broadening reissue. You're basically surrendering your patent and
getting a new one, but the date of the patent is still the date of the original. For
infringement purposes, it can be more complicated. If there's a substantial difference in
the scope of the old and new claims, then "intervening rights" comes up. IR, first,
excludes all infringement prior to the reissuance. (Missed a minute). With a reissue,
you're allowed to alter the claim, but not add new matter to your disclosure. And you
can't claim what you couldn't have claimed originally in your prosecution.

Then there's re-examination. Usually takes place with at the same time as or shortly
before litigation. There's ex-parte and inter parte examination. Re-examination decisions
are retrospective--they apply backwards. But you can only contract, not expand, the
claims. So no intervening rights worry. And PTO will only look at new ar, nothing that
was considered during prosecution or prior sales.
--Ex-parte reexamination was created as a way to address the high costs of litigation,
prevent patent trolls. But without the third party present (usually the defendant), not
particularly effective.
--Two restricts make reexamination unpopular. First, anything you say in there you
can't argue again in court. And potential defendants can't appeal if they lose (thought
they can start litigation) .Patentees, on the other hand, can.
DOCTRINE OF EQUIVALENTS

The tension is between design-around prevention and competitor uncertainty. (We're

not doing this, but there's also the almost-never-used reverse DE, which is when use
does fall into claim language, but it's so different that it's excluded).

Warner-Jenkinson
At issue, a dye purification process that is claimed "around pH 6.0-9.0". Here, the
allegedly infringing activity is at pH 5.0. the 6-9 range was put in the claim because of
prior art above 9.

So we want to know if the plaintiff is doing something equivalent. How do we frame the
inquiry. The Q is--function, way, result. The triple identity test. Does it perform the same
function, in the same way , to get to the same result.
But it's not an "overall" question. You have to line up every element. The infringing
activity has to fit each one. For elements not met exactly, you ask about the scope of
equivalence. And, of course, the perspective is PHOSITA. The weird thing, is that we're
looking at what PHOSITA would find at the date of infringement, not the date of
invention.

But defendants have an affirmative defense here. Prosecution history estoppel. If you
narrow your claim during patent prosecution, you can't turn around and try to re-claim
what you got rid of in narrowing through doctrine of equivalents.
Court says here that the reason a claim was narrowed matters a lot. If it's made to avoid
prior art, PHE applies. If the reason is unclear, patentee has burden to establish that it's
wasn't for a reason to which PHE applies.

But what if the PTO was wrong? You only narrowed it to satisfy their absurd request. In
that case, you're out luck. If you didn't like the PTO's finding, you could have appealed it
in the PTO.

Festo
At issue, a piston. The double-lipped sealing rings and the magnetized surface. The
original patent had to be modified during prosecution. It was a 112 issue, not a prior art
issue. The defendant has made a piston with a different lips and non-magnetized
surface.

The issue here is the amendment, again. The Q is whether amendments made to avoid
prior art are the only ones that qualify in triggering PHE. The answer, says, the Court,
anything that changes the scope of the claim is going to be not-minor, and will trigger
PHE. But what about the arg that 112 scope change--to make claim fit disclosure--is
different from a prior-art avoiding scope change.
The rule: Any amendment made in response to the requirements of the Patent Act is
going to count for PHE. (And that's almost everything.)

Why is this the rule? The problem with DE is notice. If we allow people to expand
beyond what claims encompass, it's hard for competitors to figure out what it covers.
That's the classic story. Kennedy goes a little deeper. He says if the reason we're
worried about design-around is that language can't describe real boundaries of an
invention and we need to police those boundaries, then when we're applying DE,
you . . . (something that's sort of redundant with the design-around worry)
*
The next question is "Does the estoppel bar the inventor from asserting infringement
against any equivalent to the narrowed element or might some elements still infringe?"
They go for a flexible bar. "Still, the Court conceded, however, that there are some
cases where the amendment cannot be viewed as surrendering a particular equivalent.
"The equivalent may have been unforeseeable at the time of the application; the
rationale underlying the amendment may bear no more than a tangential relation to the
equivalent in question; or there may be some other reason suggesting that the patentee
could not reasonably be expected to have described the insubstantial substitute in
question. In those cases the patentee can overcome the presumption that prosecution
history estoppel bars a finding of equivalence."" ----> this is the standard.
"unforeseeable at the time" or "tangential relation" or "any other reason..."

So this puts us in a sort of gray area between two extremes. What do courts do once
they find themselves with an amendment, and thus in the world of PHE. This
"unforeseeable at the time" stuff is important.

So you have amendment, which, if under Patent Act, triggers PHE inquiry. The burden
goes to patentee to show exactly why the amendment was made, and that there was no
way they could have known to come up with literal claim language that could have
covered what came after. This makes the bar pretty high, especially for soon-after
arising technology.
Some people think that the test is now dead. You still have the triple identity test in the
background, but the focus is much more now on literal infringement.

Golf ball case

Literally, no infringement, because there's a tiny amount of dimple-great circle overlap.
The rule is that the scope of equivalency can't include something that would have made
the patent claim invalid (here, because of anticipation).

The test: Imagine that at time of patent application, you had claimed what the
defendants are doing. Would that patent have issued?
IMPORTANT: PHE applies to the doctrine of equivalents. It DOES NOT apply to claim
construction. CC has rules about disclaimer, but you don't apply PHE.

For M: Means + Function and Indirect Infringement

4/12

MEANS + FUNCTION CLAIMS

You're looking for the word "means" (usually), plus some function, without structural
stuff

How do you do literal infringement analysis for means+function claims? You can't look
at the language of the claims for as much.

112(f): Refer to the structures you're given in the specification, and ask it they're
equivalent. And the test for equivalence--and this is different from the doctrine of
equivalents!--you're looking at PHOSITA at the time of the patent claim. And the
overarching term is "insubstantial change"--which is about known interchangeability.
Look at what's in the specification, and see if what's at issue was known to be
interchangeable with the relevant part of the structure at the time.

Even though this looks like non-literal, equivalence analysis, this is the literal
infringement analysis for means+function. The biggest difference between DE and MF-
literal is that DE covers after-arising technology. But after-arising had nothing to do with
this case. For this case's purposes, DE and MF were the same. So court, as a matter of
law, shifts the jury's finding, which had improper instructions, from DE to MF-literal
liability.

So: basically the same as DE, but pay attention to after-arising technology diff, and look
at the structures disclosed in the specification.

INDIRECT INFRINGEMENT
We're not going to spend a lot of time on this, but it's important.

271(a) is what it means to infringe a patent. "Makes, uses, or sells." What if you don't
make, use, or sell it, but you aid and abet that?
Classic third-party liability question: Why not just go after the direct infringers? Because
of enforcement problems.

Aro
The invention is a convertible car top. Aro manufactures replacement fabric covers for in
the interior, for specific models, including the cars we're interested in.
Then there's the CTR company, which has the rights from AB--the convertible
manufacturer--to make convertible tops.
(When you buy a patented product, you are also buying the implicit right to use it. And
you also have the exhaustion/first-sale right to sell it.)

So the first Aro cases were whether the people who bought the convertible tops were
infringing the patent on the convertible top or not. The standard here is repair vs.
reconstruction. Repair is an implied right; reconstruction is not. It's obviously a matter a
of degree. But about the core of the structure, or something.

Since you can't indirectly infringe without underlying direct infringement,

repair/reconstruciton is an important question. The answers with respect to GM and with
respect to Ford are different. Ford never had a license from the patent holder. The Ford
drivers were infringers. So even aiding and abetting any use, not to mention repair, is
indirect infringement.

Aro, then, was selling components of a patented invention to infringers. Again, the
justification has to do with enforcement. How could these guys recover by going after
individual customers.

1. The first question, to repeat, is whether there was underlying infringement. There was
here.

2. Then (not at issue here) you look at whether the infringing thing is a "staple article"--
whether there are SNIU. And no SNIU here--it can only be used on the Ford cars.

3. Then you need the mental element--contributory infinger must know "that the
combination for which his component was especially designed [staple] was both
patented and infringing". So you gotta know how your thing is being used, and you gotta
know about the patent.

Then there's this thing about misuse at the end. Aro says look, CRT, you don't have a
patent on the cover, just on the convertible as a whole. You're using your convertible
patent to prevent us from making something unpatented, in the PD. Misuse doctrine
grew a lot in the 20th century, making it really hard to pursue indirect infringement
claims. In 271(g), Congress basically gets rid of this line of cases. The idea is, patent
right become near-unenforceable this way, especially with all these patents in
combinations. You could still sue the users who are putting all the parts together and
using your patented combination, but are you really going to? We value the knowledge
about how to actually put together and use stuff.

CR Bard
The invention is this method of coronary angioplasty. The invention only has holes in a
particular place (in the artery?) to avoid the prior art. The defendant's invention, which is
allegedly infringing, has holes in a bunch of different places. There's three possibilities
as to how it's used. All holes in the artery, all in the aorta, or some in each. All-in-the-
aorta is practice of part. All-in-the-artery infringes for sure. And some in each, let's just
call it infringement.

So the catheter maker is being sued for indirect infringement.

1-Underlying infringement. The patent is a method. The potential direct infringers are
the doctors/hospitals. So yeah, with some of these uses, we can definitely tell a story of
direct infringment

But case gets stuck at 2. It looks like about 40% of the catheter use would be non-
infringing.
If this didn't work, Ds could claim no mental element. But that's probably a loser here.

NB: You can also get indirect infringement, regardless of 2, if you're "actively inducing"
infringement. Here, instructions are the key thing to look for. 271(b)
Hypo: Drug that is approved for use in depression. But turns out to have good in off-
label use for anxiety. Under 271(c), the generic people are off the hook. But under
271(b), they'd prob be liable, because FDA only has one approved use, so instructions
have to give instructions for depression. But if FDA approves the use for anxiety, then
the generic maker can get off the hook under 271(b), as long as they limit their
directions to anxiety.

Brown
Real old case. Patent on an improvement in boat gaffe construction. French boat
passing through uses it. The claim under 271 would be direct infringement. What the
French boat is doing is concededly covered by the patent. And the patent is valid.

The problem is that patents are territorial.

Court, in part, uses a de minimis argument. A lot of "doing this would be a bad idea"
type stuff. No real textual arguments.

This holding was codified in 272, which applies to ships.

*
Why are patents territorial? It comes from tradition, really (contrast expectations re
chattel rights). TRIPS really changed everything.

Microsoft v. ATT
271(f) says that there's infringement when one "supplies...from the US," for combination
abroad, a patented inventions "components." All that, on a broad level, happened here.
But the "components" sent were material things, but a master CD of software. Does the
immaterial software--the "composition," as opposed to the "sheet music"--constitute the
components that are combined to form the invention? Court says no.

This case was sort of a big deal.

For T, INJUNCTIVE RELIEF

4/12
Is it really true that damages allow competition where injunctions don't? Not
necessarily--depends how high you set the damages.

Calabresi conventional stuff--injunction just forces you to bargain. And, Coase, if there's
no transaction costs, they'll reach a good outcome. How does this lead to overpayment?
Well, say someone invests a huge amount in a new product, not knowing that they're
infringing a patent. If they invest and "miss it," then there's a ton of sunk costs unrelated
to the patented technology. Which allows the patentee to stage a holdout. (There's also
a design-around issue. After investing it may be really hard to do so (or something)).

Existence of injunctions also creates a business model--buy up little-known patents and

start holding up everybody.

But maybe should our answer to the hold-up problem be "tough luck"? You should have
known you were infringing. The modest arguments is hey, there are a ton--a ton--of
patents out there. It takes a lot of money, work, and time to wade through them all. And
that money is better spent on engineers than lawyers. The public notice function, that is,
is really costly. And moreover, what a particular patent covers is up for grabs. Not even
the courts can do claim construction. For these reasons, the holdup problem is more
likely to occur

But there are other arguments. Granting an injunction is just saying damages are
determined between the parties. Holdout can make that number bigger than is socially
valuable. Damages are just court coming in and doing it themselves. Another way to
look at it is as a compulsory license.
A lot of time, damages and compulsory licenses undervalue the patent. Obviously
doesn't have to, though.

If you grant damages and the damages are too high, the infringer will just leave the
business, or negotiate a lower rate. If damages are too low, the infringer will just pay it.
In this way, damages are feared to systematically advantage the infringer. Too high
leads to bargaining, too low leads to not paying enough. (We're talking about future
damages here). Damages seems to systematically lower the expected payment
needed. (This could be a good thing, if holdout is always happening).

Ebay
Before eBay, permanent injunctions are granted as a matter of course. But the statute
only says that injunctions may be issued, in light of the principles of equity

The district court in this case had denied the injunction. Says that the patentee isn't
practicing the invention, is just licensing it, which militates against injunction. Fed Cir
reverses, saying we always give injunctions, basically

SC says no to both. You have to follow the traditional four-factor test. Irreparable injury,
inability to otherwise recover (hard for courts to figure out, monetarily, or no?), balance
of hardships (what are the consequences of the injunction for the infringer), public
interest. Factors 3 and 4 can get blurred when you think about effect on customers.
Factor 4 usually only comes up with public health stuff.

Z4
Patent on anti-piracy log-in software style thing.

(1) II
Regarding irreparable injury, court looks at Z4's business model, and Microsoft's use of
their product. IOW, IMPORTANT: Infringer is NOT a direct competitor. And all that Z4
has lost is licensing fees here--no issue of loss of brand name recognition or market
share. Also, and also IMPORTANT: It's a lot harder for a patentee to get an injunction
when she's not actually practicing the patent--when she's just licensing it.

(2) AR
In the 'adequacy of remedies' issue, court starts talking about holdout problem, in
addition to some overlapping considerations from factor 1. The holdout thing can go
here, or it can go in balance of hardships. But do it somehow (and look here for a
model?). Then go on here to say that they have metrics for calculating a reasonable
royalty. They can count sales here, and they have good estimates as to how much
longer the product will be used for. And, importantly, Microsoft is NOT judgment proof

(3) BH
Here there's the more elaborate discussion about the cost of designing around for
Microsoft (a version of the holdout problem).
(Missed a minute)

(4) PI
Court considers the worldwide impact of an injunction. Again, though, PI not really
moving the outcome unless it's a public health case.
*
Is it really true that an injunction would have given Z4 too much leverage here? Look at
it from Z4's point of view.
Amazon
One-click patent. Able to do it because it remembers all your info. (Does the free rider
story here apply here? That is, is a patent even necessary? Doesn't look like it, but
that's not how our system works).

The issue here is a preliminary injunction.

(1) Likelihood of success on the merits, (2) irreparable harm, (3) balance of hardships,
(4) public interest. Missing the adequacy of damages issue from the normal inj analysis.

From this case--and it might be diff after ebay--after (1) is proven, there's a presumption
for a prelim inj.

Burden framework: For PI, just need to assert infringement/validity defense that
patentee cannot prove lacks "substantial merit"

READ DAMAGES + WILLFUL INF

4/19

Because eBay is so recent, there's not any case law on permanent damages. But a lot
on prior damages.

Lost profits and Reasonable royalties are the two types

Panduit
Lost Profits
The most basic idea is to make the plaintiff whole--make it "as if there were never an
infringement." What does that counterfactual world look like? One possibility is that the
infringing competition would have never been on the market. Another is that the
infringer would have got a license for the use. If we think patentee would have denied
license, we do lost profits. If we think a license would have been reached, we do
reasonably royalties

To obtain lost profits damages, plaintiff must prove: (1) demand for the patented
product, (2) absence of acceptable non-infringing substitutes, (3) his manufacturing and
marketing capability to exploit the demand, and (4) the amount of the profit he would
have made.

Re (2), what's the scope? What counts as an "acceptable" substitute? Well, you first
need a definition of the "market" to find out if there are other players in it. (2) is one of
the main reasons that it's hard to get lost profits, especially in this post-9/11 world.

(1)-(3) are clearly causation questions.

Remember, this is all stuff that plaintiffs must *show*.

For (4), you need to know the number of units the infringer sold (easy), and the profit
margins you would have had on those units (tricky). For the latter part, you look at your
prices before infringement, because infringement's added competition drives the price
down.

The lost profits inquiry tends to force patentees to disclose stuff that they'd rather not
disclose to competitors. Costs, capacity, etc. And if you choose not to share these
elements, you can't get lost profits damages.

Reasonable royalty
The concern is that if we just make them pay whatever they would have in the first
place, there's an incentive to infringe and go to trial. At trial, you'll either win or you'll just
pay what you originally would have.
But there's another side of the story that says routinely challenging patents is actually a
social good. If a patent is declared invalid, the whole public gets the resultant benefit.

Some factors: What plaintiff's property was, to what extent defendant has taken it, its
usefulness and commercial value as shown by its advantages over other things and by
the extent of its use.

You also look at the noninfringing substitutes lying around. Gets to the cost of design-
around, and thus the relative bargaining positions.

(Missed a few)
*
RR is the main type of damages given, because lost profits is both hard to do and
companies don't like to do it.

RR works "well" with products that have huge sales, and the patented thing is a small
component. Because a tiny, reasonable-seeming percentage turns out to have huge
returns. The fact of the matter is that in real negotiations, small components wouldn't
have garnered a percentage, but a fixed fee.

Rite Hite
Scope of damages as a simple causation questions (but for + proximate), which allows
the company to recover for lost sales on things not covered by the patent at issue.

But if you can move the infringing thing to become an acceptable non-infringing
substitute, or if the patent on the second thing is invalid, then these damages are
excessive. But if you can't do either of those things, then the infringer can't continue to
do anything without having to still pay for the second thing.
But none of those questions about the second thing were at issue in this case. So the
court had to decide who gets the benefit of the doubt.

4/20

WILLFUL INFRINGEMENT
Seagate
FC gets rid of affirmative duty of care requirements, and instates a recklessness
standard. No longer burden of defendant to show they weren't willful.

And the standard becomes one like the preliminary injunction: Can the patentee show
that all the defenses raised by the defendant lack substantial merit. But court says it's
not sure it wants to use the PI fuzzy factors--balance of hardship, public int, etc. Looks
like this isn't going to be a part of the test.

Treble damages is a big deal. You already have big damages, then you triple it? Yeesh.

EXPERIMENTAL USE DEFENSE

The policy is that you don't want patents to get in the way of scientific research. You
have a sense of wanting people to be able to play with patented stuff to improve it.

The problem is that a lot of research *tools* are themselves patented inventions. If you
allow people to use those experimentally for free, you basically undercut the whole
market.

Court basically guts it. Doctrine only applies "if the act is solely for amusement, to
satisfy idle curiosity, or for strictly philosophical inquiry." So the doctrine never applies.

LACHES
Another defense.
Symbol v. Lemelson (p. 1130)
This patent from the 50s that kept being continued until Lemelson finally got it, for the
bar code scanner used at grocery stores. Nobody really thinks the patent will hold up in
court. But Lemelson sends letters to all grocery stores asking for a royalty. The mfr of
the bar code scanner files for a declaratory judgment.

Laches isn't about the patent, it's an equitable defense. Laches hadn't been used in a
while, and it's resurrected here. Prosecution laches. Long delays in the PTO that appear
to have been done for reasons unrelated to the task of prosecution, and more related to
strategic delay, can give rise to a laches defense.

Not much guidance as to exactly how this applies.

INEQUITABLE CONDUCT
To see the idea, take a step back and look at the prosecution process. Examiners really
have to depend on the potential patentees to help them find the prior art.

So we try to punish inventors on the enforcement end (since you can't do much after
you allow the patent) to keep inventors candid during prosecution

IC isn't statutory. So the question is what is the remedy for IC

JP Stevens
Invalidity vs. unenforceability. Court goes with unenforceability. Why does this matter?
What's the diff between invalidity and unenforceability.

You don't invalidate a whole patent. You invalidate claims in a patent. (Of course, you
could invalidate all the claims in the patent). But since validity analysis takes place at
claim level, an IC doctrine based on it would have to tie any IC to particular claims and
invalidate--punish--piece by piece. With unenforceability, you don't have to worry about
that.

This remedy covers all patents involved in that prosecution, too. AND it can be the basis
for an antitrust suit!

Critikon
2 prongs of showing inequitable conduct:
Materiality and intent

Materiality--the undisclosed information was material. The standard is whether a

reasonable examiner would think it important for questions of patentability.

There are other possible materiality tests--"objective but for," "subjective but for"

Intent--Knowledge. Have to have knowledge for intent to deceive. Knew about the
reference that they didn't submit, and have to have purpose to deceive the PTO (?)

How do you get in mind of inventors.

Sliding scale: The more material the reference, the less intent you need to show. The
less material, the more intent. But you always gotta have some independent showing of
intent.

You can look at what the attorneys did with the same refs for other patents. We don't
treat the patent attorney as a litigator. We expect a litigator to do all she can in any
given client's interest. Inconsistency isn't a bad (or at least an illegal) thing. Btu patent
proceedings aren't supposed to be as adversarial.

(Missed like 15)

Are we on the same case?

Why did we find IC here. Such a harsh penalty, and there are findings that this
disclosure wouldn't actually have made a difference. Well, if you let people off the IC
hook when it turns out patent would remain valid, then (1) it takes away incentive to be
honest in front of PTO, and (2) you get rid of the general duty to good faith in the patent
prosecution.
(This answer overstates the point a little bit. You can imagine a structure where if the
disclosure would have made invalid even a small part of the patent, the whole patent is
unenforceable)

Kingstown
(Taking a step back) (Ask Tyler?)
Step 1 - Was there omitted or false information?
(here, yes: things that were said to be, but weren't, admitted by previous examiner)
Step 2 - Did attorneys know the disclosure was false?
(here, basically: one of the attorneys was in a position to know that what they gave was
false--but you want to be careful making that inference)
Step 3 - Was the false statement/admitted reference material?
(here, yes. two reasons: (1) any patent examiner would find it relevant what was found
by previous examiner; (2) more subjective--wrt the particular issue here, we know at
least one examiner thought this material)
Step 4 - Was there intent?
(here, eh. this falls really low on the materiality spectrum
Step 5 - Balancing test. On the whole, does there look to be intent here

Here, no finding of inequitable conduct. Easy to make the mistake, they didn't gain from
it, etc. This case is sort of an outlier. IC has started to become a real sword for the
infringement defendant. And that's triggered an info dump on part of inventors, just to be
safe. Fed Cir doesn't care what PTO actually asks for.