“Anything You Say …” — Terms with Inconsistent Definitions Are Indefinite
Last week, the U.S. Court of Appeals for the Federal Circuit (CAFC) held the term “computer” to be indefinite. That such a familiar term could be indefinite should warn patent practitioners to take care when arguing the scope of a term — even in otherwise unrelated proceedings, possibly separated by years.
Infinity Computer Products, Inc., owns a U.S. patent on using a fax machine as a printer/scanner for a personal computer. As detailed below, one element of an independent claim was “a passive link between the facsimile machine and the computer.” The prosecution history had inconsistent definitions of “passive link,” creating indefiniteness about where the “passive link” begins and the “computer” ends. The CAFC ruled that both terms were therefore indefinite, which invalidated the claim.
Indefinite claims are invalid
Patents — being a kind of sanctioned monopoly — must give the public notice about their precise scope. Conversely, the public needs clear notice of what is still open to them. To this end, Section 112 of the Patent Act requires claims “particularly pointing out and distinctly claiming the subject matter” of the patent. If a patent’s claims fail to do that, we say it is invalid for indefiniteness.
The Supreme Court set the gold standard for indefiniteness in Nautilus v. BioSig: “[A] patent is invalid for indefiniteness if its claims, read in light of … the prosecution history, fail to inform, with reasonable certainty, those skilled in the art about the scope of the invention.” One way to fail is for the specification or the prosecution history to use inconsistent definitions of a claim term. For example, in Teva Pharmaceuticals v. Sandoz, the claim at issue had used the term “molecular weight.” But the specification did not say which of three competing definitions of “molecular weight” was intended. In fact, the prosecution history reflected use of two different and inconsistent definitions. So Teva Pharmaceuticals’ patent was invalid for indefiniteness.
Infinity’s claims are indefinite due to conflicting definitions
But how could a term so well understood as is “computer” be indefinite, even to one skilled in the art of interfacing a fax machine with a PC? The answer lies in its interaction with the other indefinite term: “passive link.” Infinity had conceded that, for the claims to be definite, one skilled in the art must be “reasonably certain where the passive link ends and the computer begins.” Indefiniteness in the term “passive link” infected the term “computer” with the same degree of indefiniteness.
The term “passive link” is not in the specification, nor was it in the originally submitted claims. Instead, “passive link” had been added during prosecution to overcome a prior-art reference, “Perkins.” Between the fax machine and the PC, Perkins includes an intervening component, e.g., a fax modem. The intervening component could even be a card located “in the computer.” So, Infinity had limited its claims to a passive link between the fax and PC, distinguishing itself from Perkins as not requiring “any intervening apparatus.” More specifically, Infinity had argued that Perkins’ intervening component “should be regarded as a peripheral device” because it “processes data before it is transmitted to the I/O bus of the computer.” Infinity had argued that its own invention had no such intervening device, because the passive link connects directly to the I/O bus of the computer. Crucially, to make this distinction, Infinity relied on drawings that had been added to the specification in a continuation-in-part. The USPTO issued a patent to Infinity on the basis of this distinction (the “’811 Patent”).
Later the ’811 Patent was subjected to reexamination. This time, the prior art at issue (“Kenmochi”) preceded Infinity’s continuation-in-part. So, Infinity needed to argue the claims were supported by drawings in the original parent application. But each of those earlier figures depicted, internal to the computer, circuitry similar to the internal card of Perkins — i.e., an “intervening component.” To survive the reexamination, then, Infinity fatally argued that “the use of a modulation procedure within the PC and facsimile machine … does not insert an intervening apparatus.”
Infinity had therefore assigned “passive link” to two different, inconsistent scopes. During initial prosecution, to overcome Perkins, “passive link” excluded devices with an internal-card fax modem between the fax machine and the PC. But during reexamination, to pre-date Kenmochi, the scope “passive link” must include such devices. In the first instance, the passive link ends at the port of the computer. In the second instance, the passive link extends all the way through to the I/O bus.
Litigation with Oki Data
The conflicting definitions proved to be a time bomb for Infinity. In litigating the ’811 Patent against Oki Data Americas in the District of Delaware, Oki Data exploited the latent inconsistency. The District Court held the ’811 Patent invalid for indefiniteness. The CAFC agreed.
In its opinion, the CAFC analogized Infinity’s position to that of Teva Pharmaceuticals. Without a consistent definition of “passive link,” one skilled in the art could not know what scope remained open to the public. There could be no reasonable certainty “where the passive link ends and where the computer begins.” So, the ’811 Patent failed to meet the standard for claims under Section 112.
Infinity as the Miranda of patent law
The lesson here is hard, but clear. Say little, and be consistent. Infinity was in a tough spot: Threading the needle between Perkins and Kenmochi would be difficult for anyone (especially if Infinity was unaware of Kenmochi until the reexamination). However, had the “passive link” amendment been left to speak for itself — rather than arguing in detail about the “intervening component” theory — Infinity would have had more room to construe the term carefully. Importantly, it may have been able to delay giving any definition until the reexamination. In that case, Oki Data could not have complained about inconsistent definitions, because there would only have been one. Patent practitioners are friendly, and we often want to help the USPTO side with our clients. But remember: Anything you say can and will be used against you.