Determining the scope of a patent
A patent gives its owner certain exclusive rights with respect to an invention. In return for these rights, the patent must describe how to construct the invention and how to use it. This way, once the patent rights run out, everyone can use the invention and so more knowledge becomes available to society.
The rights conferred by a patent need to be defined well, so that it is clear to everyone what is and is not covered by the patent. To this end, every patent contains one or more so-called "claims", which provide a definition of what is covered by the patent. Only things that meet the definition of the claims infringe on the patent.
The claims (typically found at the end of a patent document) provide a definition of what the patent protects. Terms used in a claim may be defined by the whole document, but ultimately only what is described in the claims is protected.
To infringe, each and every element of a claim must be present in the infringing product. If even a single element is missing, the product does not infringe. It doesn't matter if the patent document says that that single element is optional: if it's in the claim, it is required.
The claims are used in a very similar manner when judging the validity of a patent (or application). When prior art is found, it must be compared against the claims to determine if the patent is new and nonobvious. Even if the "spirit of the invention" or the "general idea" is the same as in the prior art, if the claims contain one feature that is not mentioned in the prior art, the invention is new. If the one new feature is not obvious, then the claim is patentable.
An example: patenting the button
To illustrate the use of claims, let's assume that at some point in time, only cloaks without fastening mechanisms exist. Or, in patent-speak, the state of the art consists of cloaks whose left and right sides cannot be connected together. Of course this makes using the cloak problematic, since it keeps blowing open in bad weather, and rain can leak in through the opening.
Alice's invention: buttons and holes
Alice recognizes this problem and is the first to invent the button to allow the sides of a cloak to be held together. She applies for a patent and gets one granted, with the following claim:
A cloak with a front opening, with a row of fasteners down one side of the front opening, and a row of holes at corresponding locations down another side of the front opening into which the fasteners can be inserted.
Alice can now sue anyone who sells cloaks with buttons and holes in which the buttons are to be inserted. She can't sue people who sell traditional cloaks, since those don't have the buttons and holes required by her patent claim.
Bob's invention: metallic hooks
Working independently on the same problem, Bob later invents a mechanism with metallic hooks and receptacles into which the hooks are placed. He obtains a patent with the following claim:
A cloak with a front opening, having at least one metallic hook at one side of the front opening, and at least one receptacle for said hook at the other side.
Bob can now sue people who sell cloaks with metallic hooks and receptacles, but not people who use wooden buttons and holes, even though a hole can be seen as a receptacle. A wooden button is not the same as a metallic hook. Bob can also sue people who use a single metallic hook, as opposed to Alice, whose claim requires multiple buttons and holes (since she uses "a row of fasteners", and a single fastener doesn't make a row).
Proving infringement using the claims
To prove infringement of their patents, Alice and Bob must show that the alleged infringer uses all of the elements in their patent claim.
Ian's infringement: cloaks with buttons
Suppose one Ian comes along and manufacturers cloaks with buttons. Ian must now take a license under Alice's patent, or risk getting sued by her for patent infringement. After all, Ian uses buttons, and so his cloaks meet the definition of Alice's claim because he uses a row of fasteners down one side and a row of holes down the other side. But since Ian's buttons are not metallic hooks, he does not have to take a license under Bob's patent.
Jack's infringement: cloaks with hooks and holes
Jack produces cloaks with metallic hooks on one side of the front opening, which are plugged into holes at the other side. Clearly Jack must now take a license from Bob, as his cloaks meet the definition of Bob's claim. However, Jack must also pay Alice royalties, since Alice's claim is not restricted to traditional round buttons, but in fact covers fasteners of all shapes, including metallic fasteners that are hooked into holes.
Keanu's infringement: the anorak
Keanu produces cloaks with fasteners, but instead of holes in the other side of the cloak, he provides loops made of string through which rectangular wooden buttons are to be put (in other words: Keanu produces anoraks). Since he uses wooden buttons, he falls outside the scope of Bob's patent claim. However, since Keanu's anoraks do not have a row of holes down one side of the garment, but instead use loops made of string that are put on top of the garment, he also falls outside the scope of Alice's claim.
Leo's workaround: buttons on shirts
Now shirt manufacturer Leo comes along and notices that Alice's buttons would also be very useful for his shirts. He uses the very same buttons that Alice uses, but puts them on shirts and not on cloaks. Since Alice's claim is restricted to "cloaks", Leo doesn't infringe on her patent.
Although the literal wording of the claims is the most important, there are still ways to extend the scope of a patent beyond this literal wording. One way of doing this is to redefine certain terms in the patent to give them a broader meaning. That makes the claim broader than at first glance. Further, if the patent holder can show that the infringer is doing something equivalent to the patented feature, he can still prove infringement.
Infringement through equivalence
Alice obviously doesn't like Keanu and Leo selling their variations of her inventions, and so starts lawsuits to get them off the market. However, since in both cases the products do not literally match Alice's claim, there is no direct infringement. Alice has another option though: the doctrine of equivalence.
Using this doctrine, someone infringes on a patent even if he doesn't use all the measures of a claim, as long as the measures he substituted are regarded as "equivalent" of the measures in the claim. When and how something is an equivalent of a term in a patent claim is a matter of national patent law. The scope of equivalence varies from country to country, and even within one country it can vary with time.
A popular "test" to see if something is equivalent is the so-called function-way-result test: does the allegedly equivalent measure perform the same function, and achieve the same result in the same way as the measure in the claim? But also important is how relevant the substitution (or omission) is to the invention.
In Keanu's case, Alice would argue that Keanu literally infringes on all elements, except on the "row of holes down the other side of the cloak". Keanu has instead provided holes in the form of loops on that other side. These loops have the same function, namely to act as receptacles for the fasteners. Using these loops, you can close the cloak (the same result) in the same way - by putting the fasteners in the corresponding receptacles. So, using the function-way-result test Keanu infringes on Alice's patent.
In Leo's case, Alice can argue that the use of a shirt instead of a cloak is an equivalent, since the only reason the cloak is mentioned is because the fasteners are used to close its front opening, so the fact that it's a cloak is irrelevant. Leo could argue that if that fact were irrelevant, then why did Alice specify that it had to be a cloak?
Leo could also argue that Alice's patented invention provides a solution for the problem that cloaks blow open in strong winds, which does not occur with shirts, as those would not normally be exposed to such winds. That makes the cloak an essential part of the invention, since the invention wouldn't perform its intended purpose (preventing the garment from being blown open) on garments like shirts that don't get blown open.
The easier it is to prove that something is equivalent to a measure in a claim, the broader the scope of the claim gets. In a country with no provisions for equivalence at all, Alice would be out of luck, since cloaks are not literally the same as shirts. A country that uses claims only as a rough guideline, and is willing to regard almost anything as equivalence, would find that the shirts did infringe. This makes the doctrine of equivalence a potentially powerful weapon to control the scope and usefulness of patents.
Limits on equivalence in the USA: Festo
Recently the US Court of Appeals for the Federal Circuit has issued the so-called "Festo" decision. According to this decision, an element of a patent claim cannot be infringed by an equivalent measure if that element was "amended for reasons of patentability" during the prosecution of the patent application.
For instance, suppose that Bob's patent was issued before Alice applied for her patent. Alice's claim now is too broad, as her "fasteners" are a generalization of Bob's "metallic hooks" and so would claim part of the state of the art, which is not permitted. Alice thus gets a rejection notice from the Patent Office. To overcome this rejection, Alice rewrites "fasteners" to "round wooden discs" and successfully argues that this is patentably different from Bob's metallic hooks. Alice now gets her patent issued as above, with "round wooden discs" instead of "fasteners."
Now Alice has a problem going after Keanu, who manufactured cloaks with rectangular wooden buttons, and not round ones. Although it would be easy to argue that Keanu's rectangular buttons are an equivalent to Alice's round ones, Alice is prevented from doing so due to the Festo decision. After all, she amended this particular element (the fasteners) for reasons of patentability (to overcome a rejection for lack of novelty).
Although before Festo there already were other restrictions on the doctrine of equivalence, those were more limited. In particular, so-called prosection history estoppel or file wrapper estoppel could prevent Alice from going after Keanu if she had argued that the fact that the buttons were round was essential to her invention. But if her argument was based on the wooden aspect of her buttons, she could still claim that rectangular buttons rather than round ones were an equivalent.
In May 2002, the US Supreme Court decided that infringement by equivalence with an amended claim element is only possible if "at the time of the amendment one skilled in the art could not reasonably be expected to have drafted a claim that would have literally encompassed the alleged equivalent." In other words, only unforeseen equivalents can still be covered by an amended claim.
Redefining terms to increase the scope
Alice could have avoided all her problems with Leo if she hadn't used "cloak" but rather "garment" in her claim, or if she had stated something like "Whenever the word 'cloak' is used in this document, it should be read to also mean other types of garments, such as shirts, jackets, pants and so on." In the latter case, Alice can argue that Leo's shirts meet the definition of the claim, since she has defined "cloak" to mean all types of garments, including shirts.
Such arbitrary defining of terms in patents is usually permissible, although if stretched too far it probably won't help much. For instance, it would be a lot harder to get away with using "round" in a claim, and then stating "The term 'Round' as used here means any geometrical shape, including rectangular or square" in the description.
However, the use of broad terms also has its limitations. If all that Alice shows in her description and figures are round wooden buttons, and used just "fastener" in her claim, a court could rule that the scope of her claim is too broad. This is why many patents contain lists of possible alternative ways to do the same thing. Alice would therefore state something like "The buttons could also be square, rectangular or of any other form, and could comprise hooks for easier fastening into the receptacles."
Another issue is what is known in the state of the art. If many types of buttons are already known, Alice could not get a patent on "a cloak with fasteners" as such, but might be able to get a patent on a specific type of button. If she claimed "A cloak with two complementary fasteners on both sides of the front of the cloak", and the zipper already existed, she would have to narrow her definition.
The doctrine of equivalence is also influenced by the state of the art: something that is already known cannot infringe on a patent through the doctrine of equivalence. So Alice couldn't claim that zippers are an equivalent of her row with buttons and row with holes, if zippers already existed.
Dependent and independent claims
Typically there will be more than one claim in a patent. Some of these claims refer back to other claims. They are called "dependent" claims, since part of the definition they provide depends on those other claims. For instance, in the above example, Alice could add further claims like this:
A cloak as claimed in claim 1, in which the fasteners are round wooden objects.
A cloak as claimed in claim 1 or 2, in which the holes are formed as rectangular slits.
Dependent claims often cover advantageous ways to realize the invention. They can be used as fallback positions in case the independent claim is rejected or annulled by a court. For example, if Bob were first with his metallic hooks, Alice's claim as given above could be rejected since it covers part of the state of the art. Alice could then argue that her use of round wooden objects is still inventive compared to Bob's metallic hooks, because they are cheaper and easier to replace, and they do not set off metal detectors. This way, Alice's patent wouldn't be annulled as a whole, but only claim 1 would be annulled, and Alice can still go after people who infringe on claim 2.
While there are many ways to draft claims, there is one style worthy of particular mention: the two-part claim. In such a claim, the measures are grouped into two parts, separated by a phrase like "characterized in that", "with an improvement comprising" and so on.
In the USA, such claims are sometimes also called Jepson claims, after the first patent attorney to use them. They are rarely used, however, since they introduce a potential risk. Everything before the separating phrase (typically "with an improvement of") is assumed to be known, whether that is in fact true or not. This assumption can then be used to argue that the improvement is obvious in the light of the "admitted prior art", even though in reality the improvement wasn't obvious at all. This makes US patent practictioners wary of using this construct.
European patents will almost always have two-part claims, since this is required by Rule 29 of the European Patent Convention, unless there are very good reasons not to use this form (for example, when the improvement is a step of a method to be performed in between two known steps).