Point of Novelty

Mark A. Lemley - Stanford Law School

Posted in ,

We award patents to inventors because we hope to encourage new ideas.  It is curious, then, that patent law itself purports to pay no attention to which aspects of a patentee’s invention are in fact new.  A patented invention is legally defined by its claims—written definitions of the invention.  And those written definitions virtually never call out what it is that is new about the patentee’s invention.  For example, suppose that makers of widgets have long used a three-step process to manufacture the widget.  The inventor of a four-step process that results in higher-quality widgets will not claim to own merely the fourth step she has discovered.  Rather, she is much more likely to define her invention as including all four steps.   Our hypothetical patentee did not invent the first three steps; all three have been known in the art for decades.  But you wouldn’t know that fact from reading the patent claims.

Even if the parties do identify the novel element of an invention, the law purports not to care.  Longstanding patent law doctrine has decried any focus on the “point of novelty” of an invention.  The United States Court of Appeals for the Federal Circuit, which handles patent appeals, evaluates the claim as a whole, not just the piece of the claim that the patentee actually added to the storehouse of knowledge.  As that court has put it, “there is no legally recognizable . . . ‘gist,’ or ‘heart’ of the invention.”1

Refusing to focus on the point of novelty sometimes benefits patent owners, as when a court decides that an invention is novel because no one publication includes all the elements of the patent claim, even though the invention adds nothing new to the world.  But it can also hurt patent owners, as it does when a defendant escapes liability for infringing a patent because, even though he copied the new feature of the patentee’s invention, he did not include an old element required by the patent claim.

It turns out, however, to be hard to sustain a rule that a law concerned with novelty will pay no attention to the point of novelty.  And so point-of-novelty issues crop up in a number of different doctrines in patent law, from figuring out who counts as an inventor to whether the inventor has disclosed the “best mode” of practicing the invention to when the sale of a product exhausts the patentee’s rights in the patent.  Courts are inconsistent in whether and how they consider the point of novelty in these doctrines and others, but when the Federal Circuit is presented with a question in point-of-novelty terms, it most often falls back on the mantra that there is no point of novelty—no “gist” or “heart”—to an invention, even if that means discarding longstanding precedent.

It is time to rethink the point-of-novelty doctrine in patent law.  Ignoring the point of novelty of an invention makes little sense as an across-the-board matter and leads to a variety of harmful consequences.  While refusing to focus on the point of novelty serves some valuable purposes, there are other ways to achieve those objectives.  In the end, a patent regime that pays attention to what the patentee actually invented, not what the patent lawyer wrote, is more likely to achieve the goal of promoting innovation.

I. The Point-of-Novelty Doctrine

To understand how patent law found itself in the odd position of ignoring the very thing that makes the patentee’s invention novel, we must start with the central role of the patent claim in every aspect of the law.  Patent law defines the patentee’s invention not by what she actually built or described, but by the parameters of the patent claims.  Claims are the legal definition of the invention.  Because the words of the claim, not what the patentee actually built, determines both whether the patent is valid and whether a defendant infringes it, claim language has come to have talismanic significance in patent law.

It was not always thus.  In the early decades of the country’s history, patents did not have claims at all.  A patent case was resolved by looking at what the patentee had built, assessing whether it was different from what had been built before, and then comparing it to what the defendant had built.  When patentees began to write claims in the nineteenth century, they served as sign posts calling out the key new feature of the patentee’s invention, not as fence posts identifying the outer boundaries.  It was not until sometime after 1870 that courts converted from a central claiming system, in which the patentee identified what was new about her invention, to a peripheral claiming system in which the patentee sought to fence off whatever territory she could claim.  With the move to peripheral claiming, perhaps it was inevitable that courts interpreting patent claim language would attempt to give meaning to all words, as they do with statutes and contracts, rather than picking and choosing the words that seem important.

In any event, there was an important class of invention for which choosing the point of novelty did not seem to make much sense: the combination patent.  Imagine a patent on a Reese’s Peanut Butter Cup.  Chocolate is known in the prior art.  So, separately, is peanut butter.  Reese, our inventor, combines them.  What is the novel piece of the invention?  Neither chocolate nor peanut butter is new; the novelty and nonobviousness lies in the combination.  The rule against focusing on the point of novelty of an invention developed out of obviousness combination cases.

But the point-of-novelty doctrine is not limited to the law of obviousness.  It extends to anticipation: to invalidate a patent on grounds that it has been invented before, a single prior art reference must include all the elements of the claim in one document; if it doesn’t, the reference can’t anticipate the claim, even if everyone agrees that the omitted element is well-known in the art.  The rule against selecting a point of novelty also applies to the related doctrine of double patenting, which prevents acquiring two patents on the same invention.  It extends to the scope of patentable subject matter: a mathematical algorithm standing alone is likely not patentable, but the same algorithm implemented in a general-purpose computer probably is.  It affects what the patentee must disclose: a patent is invalid if the inventor fails to describe his best mode of practicing the invention, even if the omitted description is of a favored way of practicing an element long known in the art.  It is relevant in claim construction: a patent claim cannot be construed to include elements not expressly present in the claim, even if the text of the patent reveals that those elements are the heart of the invention.  It affects infringement: a defendant will escape liability if she omits even one element of a claim, even if that element has nothing to do with the inventor’s idea.  And while the doctrine of equivalents exists to avoid the unfairness of such a result by finding infringement where new products are only insubstantially different from the patented invention, that doctrine too has been limited so that each element of a patent claim must have an equivalent, rather than focusing on the invention as a whole or on what makes it unique.  Finally, the rule against considering the point of novelty affects damages: a reasonable royalty is calculated based on the royalty base of products actually sold, which is in turn affected by the scale of the product being claimed.  For example, if the inventor of the intermittent windshield wiper claims an improved car containing the wiper, damages may well be calculated based on the price of the whole car, since we don’t distinguish between the novel elements of the claim and the preexisting ones.

II. Cracks in the Façade

Despite the Federal Circuit’s widespread adoption of the rule against focus on a point of novelty, the point of novelty approach keeps reappearing in a variety of contexts.  Perhaps this is inevitable: it often makes little or no sense to ignore what is novel about the invention.

A. Ignoring the Commandment

The Federal Circuit has often ignored the no-point-of-novelty rule when circumstances demanded it.  Here are a few examples.

1. Inventorship.

Joint inventorship is quite common.  And since joint inventors each own an undivided interest in the whole patent, people other than the named inventor often claim that that they should have been included as joint inventors as well.

An individual is a joint inventor of a patent if she contributes to the conception of that invention.  Merely building the device someone else has designed is not invention.  However, if the putative joint inventor contributed to the conception of the invention as claimed, she is entitled to be listed on the patent and to jointly own the resulting invention.

That legal standard sits uneasily, however, with the rule against a point of novelty.  Let’s imagine that the inventor of the intermittent windshield wiper took my advice above and claimed a car with various features, including wheels, a chassis, a steering column, and an intermittent windshield wiper.  Can the inventor’s lab assistant claim to be a joint inventor by suggesting that the car should have wheels?  If we took seriously the idea that all claim elements were co-equal, there would be no more value to contributing the novel feature of the invention than to contributing something from the prior art.  But of course that’s nonsense.  And the law properly recognizes it as such, denying inventorship claims from people who contribute only knowledge found in the prior art.  In doing so, the law focuses on the point of novelty: those who contribute new claim elements are inventors; those who contribute claim elements from the prior art are not.  Any other rule would be unworkable, because it would treat as joint inventors any number of people who did not in fact contribute anything that made the invention patentable.

2.            Inequitable conduct.

Patent law punishes those who lie to the patent office by deeming their patents unenforceable.  But “inequitable conduct” is committed not only by those who affirmatively mislead the PTO, but also by applicants who withhold relevant prior art.  Relevance is, as always with patent law, determined by the patent claims.  Here too a strict rule that all claim elements are equal proves unworkable.  Just as we don’t want the person who suggested adding wheels to a car to get credit for the intermittent windshield wiper, we don’t want windshield-wiper inventors to have their patents rendered unenforceable because they didn’t turn over prior art relating to wheels or car bodies.

Once again, the law accommodates this concern by reintroducing the point of novelty analysis. Relevant prior art is defined as information that a reasonable patent examiner would have considered material to a determination of patentability.  And courts regularly recognize that examiners will only consider prior art to be material if it is at the point of novelty of the invention.  We may act in other contexts as if all claim elements are equal, but in inequitable conduct, as in inventorship, we recognize that they aren’t.

3.            Contributory infringement.

To infringe a patent, a defendant must make, use, or sell a product or process that incorporates each and every element of the patent claim.  If even one element is missing, there is no infringement.  Infringement liability extends not only to those who actually make, use, or sell the infringing technology, but also to those who materially contribute to infringement by another.  A defendant contributes to infringement by selling a product that is a “material part” of a patented invention to another that he knows will use the product as part of an infringing combination.

We worry about holding the sellers of ordinary goods—say, car tires—liable for the uses to which their products are put.  If every seller of tires, or computers, or screws had to worry about patent infringement liability based on what her customers did with the products she sold, modern commerce would be much more difficult.  As a result, patent law creates a safe harbor for the sellers of products.  They aren’t liable for contributory infringement for selling products that are “staple article[s] . . . of commerce suitable for substantial noninfringing use.”2 Because of this safe harbor, whether the seller of a product can be liable for contributory infringement depends not only on whether some buyers use it to infringe, but on whether other buyers have legitimate uses for the product.  If the only substantial use of the product is in the patented combination, the product itself is essentially treated as within the scope of the patent, and the patentee can prevent its sale.

In effect, the doctrine of contributory infringement creates a point of novelty rule.  If a component was already known in the prior art before the patentee’s invention—that is, it has other uses already—the sale of that component will not trigger liability.  But one who sells a new component specially designed to work in the patented invention will be liable.  Liability for contributory infringement accordingly depends on whether the component sold is at the point of novelty of the invention or merely a claim element that recites the prior art.

4.            Exhaustion.

A patent owner exhausts her control over a patented device with the first authorized sale.  That’s why one doesn’t need a patent license to drive a car or to resell it used, despite the fact that the car undoubtedly has patented technology in it.

This exhaustion doctrine is fairly straightforward when the item the patentee sells is the very item she patented.  When the two don’t coincide, however, things grow more complicated.  Selling a car with intermittent windshield wipers on it should exhaust the patentee’s rights over that particular automobile.  So, arguably, should selling intermittent windshield wipers standing alone.  But selling a car without intermittent windshield wipers shouldn’t exhaust the patent rights, even though the car has all the other elements of the patent claim.

In Quanta Computer, Inc. v. LG Electronics, Inc., the Supreme Court resolved this problem by asking whether the product being sold captured the “essential features” of the invention.  The sale there was by a licensee, Intel, of chips that defendant Quanta put into its devices.  LG owned a patent on methods of configuring devices with those chips.  After Intel paid LG for the license to the method, LG sought to get paid again by the downstream device manufacturers of chips, including Quanta, who LG claimed directly infringed its patents by using Intel products incorporating LG’s patents.  The Court concluded that LG’s license to Intel exhausted its patent rights, because Intel made an authorized sale, not of the whole invention, but of the “essential features”—that is, the product that was at the point of novelty.

5.            Occasional Uses: Claim Construction, Obviousness, and Enablement.

Even in claim construction and obviousness cases, the bastions of the “no heart of the invention” doctrine, courts don’t always apply that canon consistently.  That’s because doing so often has silly consequences.  For example, in Slimfold Manufacturing Co. v. Kinkead Industries, Inc., the patent claim was to a device for bi-fold doors with an improved spring that allowed them to be easily installed. Doctrinally, the rule is clear: the invention is a door that has certain parts (e.g. wood, a knob, hinges, etc.).  But the court ignored all that, reasoning that the “invention as a whole” was not the door but a pivot rod assembly for the door, because all the other elements of the door were well known.  In other words, the real invention—the point of novelty—is the pivot rod assembly.

Courts in obviousness cases do something similar when confronted with patent claims in which the only difference between the claimed invention and the prior art is a different numerical range, whether of percentage chemical composition, binding activity, temperature, or any other element.  For instance, in Iron Grip Barbell, the prior art disclosed circular weights with one, two, and four hand-holds, and the patentee claimed a circular weight with three hand-holds.  The worry in these claims is that patentees may be filling in gaps in the prior art, identifying a known technology and claiming around the edges of that technology.  In those cases, courts abandon their focus on the supposedly coequal elements of the invention and, surprisingly for a patent system that otherwise never puts the burden on the patentee to prove entitlement to a patent, presume that the invention is obvious.  To overcome that presumption, the patentee must show that the particular range chosen is “critical” to the success of the invention.  Only if the new range (the novel piece of the invention) is significantly different from the old will the patent be considered nonobvious.

Finally, one recent Federal Circuit decision has applied a point-of-novelty approach to enablement, the requirement that patentees provide sufficient instructions that one of skill in the art can make and use their invention. In Automotive Technologies International, Inc. v. BMW of North America, Inc., the court held that a patent claim that covered both mechanical and electronic impact sensors for airbags was not supported by a patent specification that only taught mechanical sensors. The patentee tried to avoid this result by arguing that people in the field would know how to construct an electronic sensor and that, therefore, the patentee didn’t have to teach it.  The Federal Circuit rejected that argument, concluding that the specification, not the general knowledge of the field, must teach the “novel aspects” of the invention.  The court thus distinguished between the novel elements and other elements; the patent doesn’t have to teach the latter as long as people in the field could figure them out.

Inventorship, inequitable conduct, contributory infringement, and exhaustion are all doctrines that explicitly invoke the point of novelty.  Not so for claim construction, enablement, or obviousness.  But even in those latter doctrines we can find instances of courts paying attention to the heart of the invention where it seems foolish to do anything else.

B.            Blindly Following the Commandment

In many other cases, courts recite the mantra that there is no heart of the invention and use that rule to decide cases.  Quite often, though, applying the no-point-of-novelty doctrine leads them astray.  Here are some examples.

1.            Repair and reconstruction.

We begin with the area of law that gives rise to the no-point-of-novelty doctrine: the law of repair.  Once patent law established that the owner of a particular patented machine was free to use or sell it as he pleased, problems began to arise around owners’ efforts to repair the device rather than purchase a new one from the patentholder.  The exhaustion doctrine was supposed to give the purchaser control over the thing he bought, not exempt him from ever having to buy from the patentholder again.  Courts faced with these cases drew a line between “repair” of a patented invention, which was within the scope of the exhaustion right, and “reconstruction” of the invention, which was tantamount to impermissibly building a new copy.

But how do we know when an owner is repairing and when he is reconstructing?  In Aro Manufacturing. Co. v. Convertible Top Replacement Co., Convertible Top Replacement owned a patent on convertible car top assemblies that functioned with fabric covers.  The defendant, Aro, sold replacement fabric car tops.  The district court and the court of appeals had held that Aro’s sales of the replacement car tops crossed the line from repair into reconstruction because the top was a major feature of the invention, so that no one could call it a “minor repair” to the overall invention.  The Supreme Court reversed.  In language that gave rise to the no-point-of-novelty doctrine, the Court said that it didn’t matter that the fabric tops were the most important part of the invention, because “this Court has made it clear . . . that there is no legally recognizable or protected ‘essential’ element, ‘gist’ or ‘heart’ of the invention in a combination patent.” 3

The Court’s insistence that there was no heart of the invention led it to what may well be the wrong result.  The Court held that car owners could bypass the patent on convertible top assemblies by replacing what is arguably the most important part of the top an unlimited number of times.  The issue is a bit complicated, because the patent was on a combination.  But unless the novel feature of the invention was something other than the whole assembly, the Court’s decision allowed a third party supplier to capture a significant part of the value of the invention.

2.            Best mode

The first paragraph of § 112 also requires that the patentee set forth the “best mode” she knows of practicing her invention.  Unlike most doctrines in patent law, which focus on what the person of skill in the art would perceive, the best mode doctrine is entirely subjective.  Whether the inventor has a preferred mode of implementing the invention and, if so, what it is depend entirely on the mindset of the individual named inventor.

In Ajinomoto Co. v. International Trade Commission, the patentee claimed a method for cultivating a strain of bacteria to produce L-lysine, but did not disclose the particular bacteria the inventor favored.  The patentee argued that it had disclosed the best mode of practicing the “innovative aspects” of the invention, but the court rejected that argument, falling back on the idea that there is no point of novelty in a patent.  The best-mode requirement, it said, “is not limited . . . to vague ‘innovative aspects’ or ‘inventive features’ of the invention, terms that appear nowhere in our best mode case law.” 4

The best-mode doctrine has been criticized as a potential trap for the unwary, because it is not obvious to non-specialists that they must do more than teach people of skill in the art how to make and use their inventions.  Unfortunately, the no-point-of-novelty rule heightens this problem.  An inventor might or might not know that she needs to disclose the technical details of her favored implementation of her new idea.  Unless she reads patent cases, however, she is unlikely to suspect that she must disclose her favored ways of practicing those elements of the invention that have long been known in the art.  Under this doctrine, the court would invalidate a patent on the car with intermittent windshield wipers if the inventor did not disclose her preferred brand of tires, a rather extreme requirement.

3.            Design patents

Design patents quite prominently featured a point-of-novelty rule until recently.  Unlike standard “utility” patents covering inventions, which are defined by the language of written claims, design patents covering new ornamental features of an object are defined in substantial part by a drawing of the patentee’s design.  The basic test for infringement of a design patent is the “ordinary observer” test, which assesses “whether an ordinary observer, familiar with the prior art, would be deceived into thinking that the accused design was the same as the patented design.”5  But that test has long been supplemented by a requirement that the defendant have appropriated the point of novelty of the invention.  If the defendant’s use didn’t include the novel feature(s) in the patentee’s design, it couldn’t infringe.

In 2008, in Egyptian Goddess, Inc. v. Swisa, Inc., the Federal Circuit changed the law of design patents to eliminate the point-of-novelty test for infringement.  The court worried that a focus on the point of novelty would confuse fact finders, particularly in the subset of cases in which the patentee’s invention was a combination of existing features rather than the development of a new one.  The focus, as it is with utility patents, is to be on the overall appearance of the whole design, not on “small” differences at the point of novelty.  Novelty still matters, but now only as a defense that is generally to be assessed separately from infringement.  And subsequent Federal Circuit cases have used Egyptian Goddess as precedent in concluding that point of novelty is no longer the test for the invalidity of a design patent either.

Think about this for a minute.  It is no longer the law that the defendant must incorporate the very thing that makes the patented invention patentable.  As long as an ordinary observer would confuse the two products, the fact that that confusion arises from similarities that already exist in the prior art doesn’t defeat a finding of infringement.  It might or might not create a defense that the patent is invalid for anticipation, though again that seems to depend on what an ordinary observer would think when comparing the patented design and the prior art.  Translated for a moment into terms of utility patents, it is as though we granted a patent on a car having an intermittent windshield wiper as the novel feature and then allowed the patentee to sue a car maker that didn’t include that feature because the cars otherwise had the same elements.  That can’t possibly be the right rule.

A design patent can now be infringed even by a product that lacks the new feature encompassed by the patent, as long as an ordinary observer would think the two were substantially the same.  The legal doctrine that prevented that from happening—the point-of-novelty test—fell victim to the commandment that inventions must not have a gist, even when they do.

4.            Claim construction/infringement

In a peripheral claiming system, the language of the patent claims defines the outer boundaries of the patent.  Application of the principle that all elements of a patent have equal significance creates traps for unwary patent owners.  Many, perhaps all but one, of the elements of a patent claim are efforts to describe existing technology rather than to define what is new about the patentee’s invention.  But a defendant can avoid infringement by eliminating any one of those elements, even if it appropriates the point of novelty of the patentee’s invention in its entirety.

The result is that patentees often lose their infringement cases because the defendant designs around the patent by avoiding a seemingly unimportant but claimed feature of the invention.  In Larami Corp. v. Amron, for example, the patentee invented a piston-pump mechanism for a water gun that allowed the gun to shoot much farther than prior guns. The patent claim covered a water gun with various elements, including a “housing . . . having a chamber therein for a liquid.”6 The patentee didn’t invent holding water in the barrel of the gun; the element was just included to define the prior art gun the patentee had improved.  But because the defendant in that case stored the water outside the barrel of the gun, in a detachable reservoir, the court concluded that it did not infringe Larami’s patent.

In theory, the doctrine of equivalents takes care of problems like this by allowing a patentee to expand the scope of its claims to cover products that are only insubstantially different from the patented invention.  But the doctrine of equivalents falls victim to the same problem: the absence of an equivalent of a single prior-art element will doom a doctrine-of-equivalents claim.  Indeed, the patentee in Larami lost its doctrine-of-equivalents argument for the same reason it lost on literal infringement: it didn’t store the water “in” the chamber or anything like it.

Could the patentee in Larami have solved this problem by drafting the claims differently?  Perhaps.  But the patentee shouldn’t have had to worry about the precise formulation of the words used to describe elements of its product that were already well-known in the art.  By focusing attention on every element, including the old ones, the no-point-of-novelty rule makes it much more likely that patentees will lose the very cases we should want them to win: ones, like Larami, in which the defendant adopts the novel feature of the patent.

6.            Damages

If patent owners can hamstring themselves by including prior art elements in their patent claims, why do they do it?  One answer has to do with the way damages are calculated.  We award as damages either lost profits from sales or, if the patentee can’t show lost profits, a reasonable royalty for the use of the invention.  Those reasonable royalties are calculated as a percentage of the sales of a product; that is, the damage award reflects a percentage royalty rate multiplied by the royalty base (the total sales of the relevant product).

But what is the relevant product?  If it is a small component with limited sales, damages are likely to be small.  But if the royalty base is large, damages are likely to be higher.  And here is where the no-point-of-novelty rule works its mischief.  Rather than focusing on the novel feature of the patentee’s invention, the rule requires us to look at the whole claimed invention, regardless of whether the elements of that invention are novel or not.  As a result, if the inventor of the intermittent windshield wiper claims a windshield wiper, his damages in a lawsuit will be measured by the sale of windshield wipers.  But if he claims the identical invention as a car with an intermittent windshield wiper as an element, the royalty base is the sale of cars, a much larger number.  As a result, patent plaintiffs tend to draft their claims to cover whole products, not the pieces they actually contribute, and they are inclined to sue downstream integrated manufacturers rather than upstream component suppliers.  And the rule that there is no point of novelty is to blame.

C.            The Problem Is Systemic

Treating old elements of an invention with the same force as the novel elements creates a number of problems for the patent system.  Sometimes we avoid those problems by ignoring the commandment and focusing on the point of novelty.  When we don’t, we often go astray.

The problem is not simply with particular applications of the no-point-of-novelty principle.  The problem is with the principle itself.  It makes little sense for a law focused on invention to pay no attention to what is inventive about the patentee’s technology.  A legal system that once focused on what the patentee actually added—demanding a working model of the invention and focusing the inquiry on that actual device—was replaced over time by a system that focuses on the words added by patent lawyers.  In the early days of the patent system, there was no such thing as a patent claim.  The claim developed as an optional device written by applicants to call attention to what was new and different about the invention.  Those claims gradually became expected, then required, then the central focus of the patent inquiry.  Today, we have reified the claims of the patent to such an extent that we are unwilling even to look behind those claims to see what parts of them reflect the patentee’s actual invention.  The result shouldn’t be surprising: both patentees and accused infringers use the ambiguity of those words to achieve their own ends.

IV.            Conclusion

Patent law today goes out of its way to avoid focusing attention on what the patentee actually invented—the point of novelty of the invention.  The results have not been encouraging.  We will better foster invention if we pay more attention to the inventions we are supposed to be promoting.  Courts should focus on the point of novelty in assessing patents.  Doing so would be an important step towards a patent law that focuses on what the patentee actually invented rather than what the patent lawyers wrote.


Mark A. Lemley is the William H. Neukom Professor of Law at Stanford Law School and a Partner at Durie Tangri LLP.

Copyright © 2011 Northwestern University School of Law.

This Legal Workshop Piece is based on the following: Mark A. Lemley, Point of Novelty, 105 NW. U. L. REV. __ (forthcoming 2011).

  1. See, e.g., Loctite Corp. v. Ultraseal Ltd., 781 F.2d 861, 875 (Fed. Cir. 1985).
  2. 35 U.S.C. § 271(c).
  3. Aro Manufacturing. Co. v. Convertible Top Replacement Co., 365 U.S. 336, 344–45.
  4. Ajinomoto Co. v. International Trade Commission, 597 F.3d 1267, 1274 (Fed. Cir. 2010).
  5. Egyptian Goddess, Inc. v. Swisa, Inc., 543 F.3d 665, 672 (Fed. Cir. 2008) (en banc).
  6. Larami Corp. v. Amron, 27 U.S.P.Q.2d 1280, 1281 (E.D. Pa. 1993).

Post a Comment (all fields are required)

You must be logged in to post a comment.