Patent Examiners often use the theory of "design choice" in formulating rejections. The typical situation is where one or more primary references show all of the elements of a claim except for what the Examiner feels is a minor modification, for example a certain structural feature such as a taper. These rejections can be particularly difficult to overcome when the specific element alleged to design choice is not disclosed in the specification as having any unique advantages.
Of course, one reason that advantages of a particular element are absent from the specification is that the drafted does not have any idea about the prior art that will be cited by an Examiner many years in the future. Often the drafted does include such description for features that differentiate the art available at the time of filing. As Examiners often do not use art supplied by the Applicant, the situation arises where a feature not previously thought to be relevant to differentiate the state of the art becomes highly relevant. Consider a case where the Examiner cites prior art far afield and thus elements that simply put the invention in context become critical elements to differentiate the prior art. These elements of course enable key advantages relative to art from far afield, but what patent drafts person would have though to add such explanation to the application? Likely the inventor reviewing the draft would have thought the drafter was simple-minded and failed to fully appreciate the invention.
And so the situation arises often that patent applicants are left staring at an obvious design choice rejection with little in the specification. Fortunately, the Federal Circuit has spoken on this very issue - In re Chu, 66 F.3d 292 (Fed. Cir. 1995).
In the Chu case, the applicant was arguing about the non-obviousness of placing a particular component (SCR catalyst) in a specific location (bag retainer), asserting that the Board had not properly established the theory of design choice, and providing various pieces of evidence (e.g., articles, etc.) outside the specification in support. The Board relied on the theory that Chu's evidence and arguments were unpersuasive because Chu's “specification is virtually silent on the matter of any purported advantage to locating the catalyst within the bag retainer” and “does not state that the claimed location of the catalyst ‘inside the bag retainer’ solves any particular problem or produces any unexpected result.” Sounds familiar, right?
However, the Federal Circuit reversed because the Board is required to consider the totality of the record and erred by requiring Chu's evidence and arguments responsive to the obviousness rejection to be within his specification in order to be considered. This makes perfect sense for the very reasons noted above - doing otherwise "would be to require patent applicants to divine the rejections the PTO will proffer when patent applications are filed." Any experienced patent professional will tell you that no one can predict the rejections proffered by the PTO.
However, this does not mean the path forward is easy. One must still have evidence, and attorney argument is generally insufficient. However, the evidence required does not necesarily rise to the lever required to rebut a prima facia case of obviousness. Rather, one must simply provide some evidence that the difference alleged to be design choice does provide some advantage or provide a solution to a problem. See In re Gal, 980 F.2d 717 (Fed.Cir.1992) (finding of “obvious design choice” precluded where the claimed structure and the function it performs are different from the prior art).
So, if your specification is unhelpful, consider using evidence countering a finding of design choice, such as an inventor declaration, journal articles, etc.