When does "a" mean "the"?

Claims use "a" and "the" in ways that are at odds with plain English usage, which can lead to some strange results.

Let's say I'm describing a text processing algorithm which counts how many times a pattern is found. In plain English, it's perfectly natural to express this as "THE number of occurrences". However, claim drafting reserves "the" for previously-introduced elements. So a claim would probably use "a" instead, to read "A number of occurrences".

Note that in plain English this construction has a completely different meaning: "a number of" often means "some" or "several" (e.g.,"a number of widgets were defective"). Even so, usually no one makes an issue of using "A" to mean "THE" even though it's counter to the plain English meaning. Our brains have been trained to apply a sort of patent filter when reading claims, and we intuitively understand that the drafter really meant "THE number of occurrences even though the claim reads "A number of occurrences."

But what if you don't assume the drafter was using patentese rather than plain English? What if you instead take the position that "a number of" has two different meanings. Then isn't the claim indefinite?

In the recent decision Ex parte Whitney (PTAB 2013), the Board said Yes, it is indefinite. The claim language at issue was: "comparing a number of occurrences of a text pattern in the collection of text at one moment in time with numbers of occurrences of the text pattern in the collection of text at other moments in time.” The Board explained as follows:

     Applicant's arguments ... focus on the failure of the prior art to disclose or suggest comparing an actual numerical count of occurrences, i.e., “a number of occurrences,” of a text pattern in a collection of text at one moment in time with the actual numerical count of occurrences of the text pattern in the collection of text at other moments in time. However, it is equally proper to interpret the phrase, “a number of occurrences,” as the Examiner does, to refer to an indefinite group of occurrences of a text pattern, e.g., some occurrences or several occurrences of the text pattern that are compared at different points in time based on some unspecified criteria.
     Because the claim language is susceptible to two equally reasonable interpretations, the claim language is ambiguous, and, thus, indefinite. See Ex parte Miyazaki, 89 USPQ2d 1207, 1215 (BPAI 2008) (precedential). Therefore, we will enter a new ground of rejection of claims 1-50, 53, and 54 as indefinite under 35 U.S.C. § 112, second paragraph.

The Examiner did not raise this issue. And because the claim was indefinite, the Board refused to decide on the prior art rejection. (In re Steele, 305 F.2d 859, 862 (CCPA 1962)(A prior art rejection cannot be sustained if the hypothetical person of ordinary skill in the art would have to make speculative assumptions concerning the meaning of claim language.)

A few months earlier, the Board reached the opposite result in Ex parte Qu (PTAB 2013). The limitation at issue read "wherein an arithmetic sum or difference of twice a number of stator teeth and a number of the stator poles equals a number of rotor poles." The Examiner asserted that it was unclear whether "a number of" ... "intended to refer to the total number of teeth or poles or some subset thereof.” The Examiner then interpreted the phrase to mean "subset of teeth" in the prior art rejection. The Examiner explained as follows: 

[T]he devices disclosed in Qu satisfy equation (1) because any number of rotor poles, stator poles, and stator teeth can meet the claimed limitation since ‘a number’ can be any integer, up to the maximum number of poles/teeth disclosed by Qu.

The Board found the claim was not indefinite in view of the specification:

In determining whether a particular claim is definite, the claim’s language must be analyzed in light of the content of the particular application, the prior art’s teaching, and the interpretation that would be given to the claim’s language by a person of ordinary skill in the art at the time the invention was made. In re Skvorecz, 580 F.3d 1262, 1268 (Fed. Cir. 2009).

The Board found that an equation in the specification made clear that "A number of" meant "THE number of":

The number of the rotor poles, the number stator poles and the stator teeth are configured to satisfy: 2 * S_teeth ± S_pole = R_pole, wherein STEETH refers to the number of stator teeth 30, SPOLE refers to the number of stator poles and RPOLE refers to the number of rotor poles.

Using this construction, the Board then reversed the prior art rejection.

I think these two decisions are outliers. In the first place, I think most Examiners use the same "patentese" rules of antecedent basis that claim drafters do. Most don't reject as indefinite, and most don't take the position that "A number of" means "SOME of". So the Board probably doesn't see this issue raised by the parties very often. I also looked for other decisions where the Board raised the issue, and didn't see any.

But these decisions do highlight how very different claim language is from plain English.