The Estoppel Argument Hiding in the File Wrapper: A Worked Example

By Contist AI

A claim chart arrives. Every limitation has a column of screenshots and spec cites asserting your client's product practices it. You have to answer all of it, and the budget says a week.

Most of that answer is technical, and should be. But a second argument is sitting somewhere nobody on the other side will look for you: the patent's own prosecution history. If a limitation was added during prosecution to get the claim allowed, the territory between the original claim and the amended one may be surrendered — and the patentee cannot take it back through the doctrine of equivalents.^[1]1Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722, 740–41 (2002).

This post works one real patent end to end: U.S. Patent No. 8,046,721, Apple's "slide to unlock" patent.^[5]5U.S. Patent No. 8,046,721, "Unlocking a device by performing gestures on an unlock image," assigned to Apple Inc., from U.S. Application No. 12/477,075, filed June 2, 2009 and granted October 25, 2011. All documents cited here are public records retrieved from the USPTO Open Data Portal. It is used here because it is public, heavily analyzed already, and its prosecution is completely ordinary — three rejections, three amendments, allowance.

The finding is one paragraph. Most of this post is what had to be eliminated before that paragraph could be trusted, including a defect we found while preparing this post.

The record

Pulled from the USPTO Open Data Portal, the prosecution resolves into twelve steps between June 2009 and October 2011: three rejections, three applicant responses, an allowance, and the grant. Twelve claims stood through all three rejections; three more were added at the final amendment.

Claim 1 is the independent claim. Its length at each step is a crude proxy, but it shows where the movement is:

One step carries real movement. The first amendment adds ten characters; the third adds three. The amendment answering the final rejection adds 156.

The finding

Claims 1, 7, 11 and 12 — amended 2010-11-24, answering the final rejection of 2010-09-17. Direction: narrowed. Reason: related to patentability. Estoppel: presumed.

The amendment added, in the claim's own words, "continuously moving the unlock image on the touch-sensitive display in accordance with movement of the contact while continuous contact with the touch screen is maintained" and "wherein the unlock image is a graphical, interactive user-interface object with which a user interacts in order to unlock the device."

The reason is not inferred. It is quoted from the remarks filed with that amendment:

"The Examiners agreed that amending the claims to recite, 'continuously moving the unlock image on the touch-sensitive display in accordance with movement of the contact while continuous contact with the touch screen is maintained, wherein the unlock image is a graphical, interactive user-interface object with which a user interacts in order to unlock the device' would overcome the current rejection."

That last clause is the whole ballgame. The applicant states, on the record, that the limitation was added to overcome a rejection. A narrowing amendment made for a reason related to patentability raises the Festo presumption that the surrendered territory is unavailable under the doctrine of equivalents.^[1]1Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722, 740–41 (2002).

And note what happens when a reason cannot be established: it does not become neutral. Under Warner-Jenkinson, where no explanation is established the court presumes a substantial reason related to patentability.^[2]2Warner-Jenkinson Co. v. Hilton Davis Chem. Co., 520 U.S. 17, 33 (1997) ("Where no explanation is established, however, the court should presume that the patent application had a substantial reason related to patentability for including the limiting element added by amendment."). Silence favors estoppel, not the patentee — so an analysis reporting "we could not determine why" is reporting something closer to bad news for the asserting side. It just has to be able to tell you which of the two it found.

What had to be eliminated

1. The reason lives in a different paper than the date

This is the one we found while preparing this post, and it is the most instructive, because the wrong answer looked completely normal.

An amendment step routinely produces two documents on the same day, and the USPTO codes them differently. The A.NE paper is the amendment — two pages here, whose only statement about the response is that it is "in response to the Office Action dated September 17, 2010." The REM paper is the remarks — four pages, carrying the argument quoted above.

Our analysis was reading whichever paper came back first, which was the amendment. The result: one claim reported as "narrowed for a reason related to patentability," grounded in a sentence that contains no reason at all — it establishes when the response was filed — and two more claims reported with reason unknown, while the actual explanation sat in a document filed the same day.

The Festo question is entirely why. Reading the papers remarks-first turned three shaky findings into four solid ones with a stated reason. The ordering keys off the code the USPTO put on the paper, which survives ingestion in the filename — the kind of document is never inferred from its contents.

A finding sourced from the wrong paper is worse than no finding, because it is specific, legally shaped, and wrong.

2. Twelve claims were reprinted. Four were amended.

37 C.F.R. § 1.121(c) requires every amendment to present a complete listing of all claims.^[3]337 C.F.R. § 1.121(c) (requiring a complete listing of all claims in every amendment, each carrying a status identifier such as "(Original)", "(Currently Amended)" or "(Canceled)"). An untouched claim is reprinted in full, verbatim, beside the one that changed — so to a text diff, every reprinted claim looks exactly like an amended one.

Each amendment in this file reprints all twelve claims. A naive comparison would report twelve narrowed claims at every step: thirty-six estoppel findings in a prosecution that has four. The status identifier the paper actually prints is what distinguishes them, and it has to be read rather than re-derived from OCR'd text. What the paper says it did beats what a diff infers.

3. The steps that cannot amend anything

Two rules are applied structurally rather than discovered per patent, because getting them wrong produces confident findings in exactly the places attorneys would not think to check:

• A grant cannot amend. It publishes the allowed claims and renumbers them. Comparing by claim number across that boundary compares different claims.
• An allowance cannot amend either — unless the Notice of Allowability declares an examiner's amendment, which is a checkbox on form PTOL-37.^[4]4MPEP § 1302.04 (examiner's amendments, entered with the applicant's authorization and recorded on the Notice of Allowability, form PTOL-37). Two papers are mailed at allowance and both look like "an allowance": that notice, and the Notice of Allowance and Fee(s) Due, which is a fee form with no claim listing and no checklist. A form that cannot carry the answer is not the same as an answer we failed to read.

The rule underneath all of this

"We looked and found nothing" and "we could not look" must never render as the same sentence.

On this angle they are opposites. "No narrowing amendment" reads to an attorney as "no estoppel available here" — an adverse conclusion. A tool that produces that sentence from a crashed comparison, a missing document, or an unread form is not merely unhelpful; it is arguing the other side's case.

So every run states which angles ran and which did not, and that statement is printed into the exported chart — because naming only the gaps lets a reader assume the rest was a complete analysis.

What it is worth in the matter

Four claims, one amendment, a presumption of surrender on a limitation the applicant said in writing was added to overcome a rejection. Not case-dispositive by itself. What it is:

• A constraint on claim construction. The applicant's characterization is intrinsic evidence and binds under Phillips.
• A limit on the doctrine of equivalents. If the contention needs equivalents to reach "continuous contact" or the "interactive user-interface object," the presumption runs against the patentee.
• A question they have to answer. The three Festo rebuttal grounds — unforeseeability, tangential relation, "some other reason" — are theirs to establish, not yours to disprove.

It took about as long to produce as reading the first office action.

Caveats, stated plainly

This is a methodology example on a public record. It is not legal advice, not an opinion on this patent's validity or enforceability, and not a comment on any litigation involving it. Prosecution history is intrinsic evidence, but claim construction rulings, expert discovery and art nobody has found yet still decide cases. Every finding above is a proposal for an attorney to accept, reject or sharpen — in the product that produced it, nothing is argued until counsel ratifies the underlying triage.

Contist builds prosecution-aware claim charts for both sides of a patent dispute. If you are answering a contention, send us the patent and the chart — we will come back with the prosecution history and the angles it opens.