There are few things more relatable than the frustration of a hard taco shell shattering unpredictably with the first bite. General Mills attempted to solve that age-old problem with a deceptively simple structural modification: arcuate lines of depressions formed in the shell so that, when bitten from the end, the shell would break in a controlled, predetermined way rather than collapsing chaotically. What followed at the PTAB is a textbook illustration of how functional language can carry real patentable weight when properly supported — and how examiners sometimes misapply “capable of” reasoning in both anticipation and obviousness analyses.

The independent claims recited a taco shell with arcuate lines of depressions extending up the sidewalls, where the shell was configured to break along a predetermined break line defined by the first arcuate line of depressions when the first end of the shell is bitten.  Claim 1 on appeal is reproduced below (Appeal 2025-001922, Application 17/565,622):

1.  A hard taco shell comprising: a first sidewall; a second sidewall; a base interconnecting the first and second sidewalls; a first end; a second end; and a first arcuate line of depressions extending up both the first and second sidewalls at a predetermined location spaced further from the second end than the first end wherein, the hard taco shell is configured to break along a predetermined break line defined by the first arcuate line of depressions when the first end of the hard taco shell is bitten.

The Examiner initially treated the functional language as little more than a statement of intended use, asserting that prior art shells with random or grid-like depressions were inherently capable of breaking along curved paths depending on where a consumer happened to bite.

The Board rejected that approach. Relying on Federal Circuit precedent, the Board emphasized that “configured to” is not satisfied by mere capability when the specification describes a purposeful structural design. The decision explained that the claim language “requires the hard taco shell to be made or designed to break in this manner when bitten as recited,” citing In re Giannelli for the proposition that “‘configured to’ when recited in a claim requires purposeful structural design, beyond mere capability in the prior art.”

That distinction proved fatal to the anticipation rejection based on the cite art’s controlled-break taco shell. The cited art disclosed evenly spaced break points across the shell intended to cause the shell to fracture close to wherever a bite occurred, producing irregular break patterns rather than a predetermined arcuate fracture line. The Board observed that the Examiner “does not identify any disclosure in [the cited art] indicating [its] taco shell is designed or constructed so that it will break as recited in claims 1 and 7 when bitten,” and noted that cited art’s own figure showed a jagged, irregular break rather than a designed break line.  Mere theoretical capability was not enough.

This taco case is a reminder of how functional language can operate when properly anchored to structure. Under long-standing USPTO practice (e.g., MPEP § 2111 and § 2114), functional limitations can carry patentable weight, especially when they describe how the claimed structure is arranged or operates with structural support in the specification, as compared to aspirational results divorced from disclosed designs. The Board treated the arcuate depressions not as decorative features that relied on hope, but as structural elements deliberately configured to produce a specific mechanical outcome.

Where the applicant here ultimately lost was not on functional claiming itself, but on obviousness. The Examiner also relied on obviousness, combining an article showing where consumers tend to bite tacos, a 1940s confection patent teaching arcuate lines of weakness to control breakage at bite locations, and the 102 reference’s general concept of weakening taco shells with depressions. The Board agreed that, taken together, those references would have suggested forming arcuate depressions along bite paths to produce controlled fracture behavior.

For patent prosecution, the lessons cut in two directions. When responding to office actions where functional language might be at issue, it can be helpful to remind Examiners of the correct legal footing. If the specification describes purposeful structure — geometry, placement, material properties, or mechanical interaction — that achieves the recited function, then “capable of” reasoning may be legally insufficient. At the same time, functional language does not insulate claims from obviousness when the art provides building blocks that collectively teach how to achieve the function. When multiple references show where forces are applied, how fracture lines can be engineered, and that food shells can be weakened in controlled ways, the structural pathway to the claimed result may be deemed predictable. In those situations, conclusory arguments that each reference “does not show depressions” or “is in a different field” rarely gain traction. The combination analysis must be confronted head-on.

For drafting applications, this example reinforces the value of explicitly tying functional results to specific physical features. To maintain breadth, a wide variety of example structures can be disclosed. However, such broad disclosure can be a double-edged sword, as the examiner can use such breadth to bring in a wide variety of prior art structures as capable of achieve the desired function, even if no such function is expressly disclosed.  Nevertheless, the stronger and more mechanical the linkage is in the written description, the harder it becomes for an examiner to reduce functional language to mere possibility.

Functional language, when used carefully, remains a useful arrow in the patent professional’s quiver for the right situation.  It can capture real engineering insight without locking claims into unnecessary geometric minutiae. But as Taco Tuesday now reminds us, it only works when the function is inseparably anchored to disclosed structure — and when prosecutors are prepared to enforce that distinction in prosecution.