Inventive Step Rejections in Japan (Article 29(2)): How the JPO Examiner Builds the Argument — and How to Take It Apart

The most common ground in a Japanese Office Action on a national phase case is Article 29(2) — lack of inventive step. The good news for foreign counsel is that JPO examiners are required to build the rejection in a prescribed way, step by step, under the Examination Guidelines. Each step is a place where the reasoning can fail. This article walks through the examiner’s method as the Guidelines describe it, then shows where a response should push.

1. The Statute and the “Person Skilled in the Art”

Art. 29(2) denies a patent where a person with ordinary skill in the art could have easily made the invention on the basis of the prior art listed in Art. 29(1) before the filing date. The Guidelines describe the skilled person as someone who can use ordinary technical means for research and development, exercise ordinary creativity such as material selection and design changes, and treat the entire state of the art in the relevant field as their own knowledge — a notional person, not an individual.

2. The Examiner’s Method: One Main Cited Invention, the Differences, the Logic

Under the Guidelines the examiner (i) selects from the prior art the single cited invention best suited for the logic — the main cited invention, normally one in the same or a close technical field or with the same or a close problem; (ii) identifies the differences between the claimed invention and the main cited invention; and (iii) tries to construct a logic (論理付け) by which the skilled person would easily arrive at the claimed invention from the main cited invention, using a secondary cited invention or common general knowledge. Two independent cited inventions may not be combined into one main cited invention, and inventive step is assessed claim by claim.

3. Factors That Build the Logic — and Factors That Break It

The Guidelines add that if there is no secondary invention corresponding to the difference and the difference is not a mere design variation, the logic cannot be constructed — and that examiners must guard against hindsight.

4. Where to Attack: a Response Strategy in Four Moves

5. Amend or Argue? Combining the Two

Because inventive step is assessed per claim, an effective first response usually does both: argues the independent claim on the flaws found above, and amends — within the original disclosure (Art. 17-2(3)) — to add a dependent feature that the cited combination does not reach. If the Notice is a final one, amendments are confined to the categories of Art. 17-2(5); see our guide to new matter and final Office Action amendments. Deadlines and extensions for overseas applicants are in the Office Action hub.

6. Practical Notes for Translated (PCT) Cases

Two issues recur in national phase cases. First, the claim wording examined is the Japanese translation: a term translated more broadly than intended can pull in prior art that the original claim would have avoided — check the translation of the distinguishing feature before arguing. Second, effects relied on must appear in the description as filed; a foreign-drafted specification that describes results only qualitatively can still support an argument, but data submitted later can only confirm, not create, a disclosed effect.

Frequently Asked Questions