Sunday, November 1, 2009

Claiming "only" as modifying an action (Ex parte Webb)

Ex parte Webb
Decided March 21, 2009
(Appeal 2008-3006; Appl. No. 09/905,298; Tech. Center 2100)

The only real issue in Ex parte Webb was claim construction. The word at issue was "only".

I got several things out of this case. One, though patent drafters and prosecutors think a lot about narrow constructions that could lead to avoiding infringement, maybe we don't think enough about broad constructions that the Examiner and/or Board can use to say that claims read on prior art. Two, dictionary definitions often hurt as as much as they help. Three, word position really matters in a claim.

The claims at issue related to a GUI feature: a dialog window that is displayed in one of two states, collapsed and complete (uncollapsed), and the cursor movements that cause the dialog to switch between the two states. Here's what the dialog looks like in the complete (uncollapsed) state:

And here's what the dialog looks like in the collapsed state:
Here is the relevant portion of the claim on appeal — edited and formatted for clarity.
displaying the complete dialog window in response to the cursor moving only from outside of the collapsed version to within the title bar of the collapsed dialog window;
At issue: exactly what cursor movements cause the complete dialog window to be displayed?

According to the Applicant, the cursor must move into the title bar to result in display of the complete dialog window. Thus, "only" operates to exclude: no other cursor movement causes this result.

Under this interpretation, the reference used to reject the claim was distinguishable. The dialog in the reference has an invisible state, where only part of the dialog is visible yet the dialog reacts to cursor movements within the entire extent. Looks something like this, with the dashed lines representing the not-shown-yet-reacts-to-cursor portion:

The reference teaches that moving the cursor anywhere within the existing-but-not-shown portion of the dialog results in the complete dialog being visible.  Since the Applicant's claim construction requires the cursor to move into the title bar, the reference is distinguishable under this construction.

The Board summarily agreed with the Applicant's position, basically repeating the Applicant's argument and reading of the reference. In fact, this isn't an interesting case at all, if you read only the opinion, because the Board didn't explain the Examiner's position.

The Examiner agreed with the Applicant on the teachings of the prior art, but has a different interpretation of the claim limitation discussed above, and specifically, about "only". According to the Examiner, other cursor movements can also result in display of the complete dialog window.

Huh? Doesn't the claim say "the cursor moving only...into the title bar" ??

The Examiner did not interpret "only" as a word of exclusion. As the Examiner read it: the claimed cursor movement results in display of the complete dialog window "without further cursor movement or positioning necessary." The way I see it, the Examiner reads "only" as "merely": the cursor moving "merely" into the title bar causes display of the complete dialog.

Under the Examiner's reading, the reference teaches the limitation at issue. The complete dialog is displayed when the cursor moves to anywhere within the invisible dialog window (including a title bar) and no further cursor movement is necessary to produce the complete dialog.

Once I read the Examiner's argument, I understood his thinking. Though no dictionary definitions were trotted out during the arguments, or by the Board, Merriam-Webster seems to support both positions:

1 a : as a single fact or instance and nothing more or different : merely 
1 b : solely, exclusively

Here's where I come down on this: both interpretations were reasonable in the abstract, but "exclusively" is the better choice. Here's the single paragraph in the Applicant's specification that actually describes the behavior:
FIG. 4 illustrates a collapsed version of dialog window 300. As illustrated in FIG. 4, cursor 302 was moved outside of complete dialog window 300 thereby resulting in dialog window 300 collapsing into a smaller area. When cursor 302 is moved back into collapsed dialog window 300, the dialog window 300 will expand once again into the complete version as illustrated in FIG. 3.
No way can you get the Examiner's interpretation of "cursor needs to merely move into the title bar, no further" from that paragraph. The Examiner's broad construction is simply unreasonable in view of the spec. So I think the Board got this one right.

Note that the spec does not say that this behavior is the *only* one that results in the collapsed dialog going back to full size. Yet that's how any ordinary reader would interpret it — after all, that's the only behavior that's described as having that effect. And certainly a patentee that tried to enforce this claim on a dialog that went from collapsed to full size upon some *other* cursor action could expect an invalid-under-written-description argument from the accused. So I think the winning claim construction is supported by the spec, even though the word "only" isn't used.

I'll wrap this up by making my final point, that word position matters.

In the final Office Action before Appeal, the Examiner explained that "placement of 'only' within the phrase is significant to the scope and meaning of the claims". Examiner further states that moving the adjective — so that the claim reads "displaying the complete dialog window only in response to the cursor moving from outside the collapsed window into the title bar" — does require the cursor to move into the title bar in order to collapse the window. [I do think the claim is more clear when phrased this way.]

The Applicant seized on the Examiner's statement, and filed an After-Final amendment to move the word as suggested. But the Examiner refused to enter the amendment after final, since "as indicated in the last Office Action, such an amendment significantly changes the scope of the claim, and therefore a further search is required". Apparently the Applicant was unwilling to file a second RCE to get this amendment in, and chose to appeal instead.

Like I said, I think the Applicant had the better argument. And this time the Board agreed. But I think most practitioners would have filed that RCE instead. The Examiner had all but promised that an RCE with amendment will get over the current art, and I see no reason to worry that a "new search" would have turned up a better reference. After all, it's not like the Examiner was actually limiting the search to references in which a cursor "merely" moved into the title bar, right? On the other hand, perhaps the Applicant was concerned about spending two or three more response cycles on another devious claim interpretation.

No comments: