The Likely Indefiniteness of Coined Terms

by Dennis Crouch

AGIS v. Life360 (Fed. Cir. 2016)

In some ways the case here can be thought of as placing a higher definiteness burden on patentees when relying upon non-industry-standard language such as coined-terms in the claims. This result makes sense to me because coined-terms are most likely to be found at points of novelty within the claim — the points where precision in description is most important. 

The AGIS claims all require a “symbol generator” to track mobile phone user location.   See U.S. Patent Nos. 7,031,728 (claims 3 and 10) and 7,672,681 (claims 5 and 9).  During claim construction, the district court found the term lacked definiteness under 35 U.S.C. 112 ¶ 2 (now 112(b)) and, although it would seemingly be a foregone conclusion, the parties stipulated that the claims were therefore invalid.

On appeal, the Federal Circuit affirmed the indefiniteness finding under its strict means-plus-function approach. The appellate panel first held that the “symbol generator” element should properly be interpreted under 35 U.S.C. 112 ¶ 6 as claiming a means for performing a specified function without reciting (in the claims) the supporting structure.  Under 112 ¶ 6, means-plus-function claim elements are However, the statute requires that MPF claim elements be tightly construed to cover only “the corresponding structure . . . described in the specification and equivalents thereof.”  Further, the Federal Circuit has repeatedly held that MPF claim elements that are not supported by corresponding structure within the specification are indefinite and thus invalid.

Step 1: Traditionally, claim elements intended to be interpreted as means-plus-function elements include the word “means.”  Here, the word ‘means’ was not used – and that leads to the a rebuttable presumption 112 ¶ 6 does not apply.  Prior to 2015, this presumption was seen as a “strong” presumption.  However, in Williamson (2015), the en banc Federal Circuit eliminated the “strong” portion of the presumption and in favor of one that appears easily rebuttable.  Under Williamson, 112 ¶ 6 will apply when the proper construction of the words of the claim fail to provide sufficiently definite structure.  The standard is “whether the words of the claim are understood by person of ordinary skill in the art to have a sufficiently definite meaning as the name for structure.” If not, then 112 ¶ 6 applies.

Here, the court noted that the term “symbol generator” was a term coined for the purposes of the patent and thus, cannot be said to be already known to one of skill in the art. As such, the court fell-back on its textual analysis – finding that “the combination of the terms [symbol and generator] as used in the context of the relevant claim language suggests that it is simply an abstraction that describes the function being performed (i.e., the generation of symbols) [and]  by itself, does not identify a structure by its function.”  Of interest, at this stage, the court did not delve into the question of whether the specification had properly defined the term.  I believe that omission was a result of the fact that the specification did not so define the term (as discussed below).

Step 2: Once a term is defined as Means-Plus-Function, the court must then look to the specification to determine whether corresponding structure is available to define the term. Here, because the symbol generator is a computer implemented function, the court requires disclosure of an algorithm for performing the function. Here, that algorithm was not provided. Quoting Aristocrat Tech, the court wrote: “A patentee cannot claim a means for performing a specific function and subsequently disclose a ‘general purpose computer as the structure designed to perform that function” because this “amounts to pure functional claiming.'”

Coined Terms and Circular Reasoning: Looking at the specifications, the only mention of the term “symbol generator” was found in one of the two specifications and that specification stated only that “The CPU also includes a symbol generator for creating touch screen display symbols discussed herein.”

Because MPF analysis involves circular reasoning, it is difficult to know what the result would have been if the specification had sufficiently and particularly defined the symbol generator as an algorithmic module.  That structural definition certainly would have been enough to satisfy structure requirement of 112 ¶ 6.  However, if it was sufficient to satisfy 112 ¶ 6, then it likely would have been sufficient to ensure that the proper construction of the term was non-MPF.  This leads to the conclusion that, at least for coined-terms, the whole game is won or lost at step 1 from above.