Guest Post by Prof. Dmitry Karshtedt: Nonobviousness and Time

Dmitry Karshtedt is an Associate Professor of Law at GW Law whose work I’ve followed for years.  Below he introduces the core idea underlying his new article on nonobviousness forthcoming in the Iowa Law Review.  -Jason

Nonobviousness and Time
Dmitry Karshtedt

Over the years, courts and commentators have said many thoughtful things about secondary considerations evidence and its role in the law of § 103, and reasonable minds have expressed significant disagreement about the value of this evidence. Consider, for example, the Federal Circuit’s fractured en banc Apple v. Samsung decision in 2016 and the never-ending debate over whether secondary considerations are best treated as rebuttal evidence to a prima facie case of obviousness based on the prior art, or whether all obviousness evidence should be analyzed holistically and at once.

My take is different: the whole primary/secondary categorization is unhelpful, especially in litigation and during post-issuance PTAB review. Let’s take one illustration. Commercial success and the skepticism of experts are both classified as “secondary considerations,” while teaching away is usually considered to be a part of the primary inquiry. However, skepticism has much more in common with teaching away, which directly tells us that the claimed invention would have been challenging to come up with, than with commercial success, which represents a market response to the invention that may indicate nonobviousness only indirectly. So what is the point of these evidentiary silos?

My answer is that there is no point. In Nonobviousness: Before and After, forthcoming in Iowa Law Review, I examine the different types of evidence used in the obviousness analysis and conclude that the primary/secondary framework does not make sense or have a strong precedential basis. The framework comes from throwaway language in Graham v. John Deere, in which the Supreme Court was grasping for some kind of a label for non-prior art evidence in § 103 cases. It has no grounding in established pre-Graham obviousness (and “invention”) precedent—which, for example, repeatedly supported the proposition that evidence like failure of others can be highly probative of validity. And it is inconsistent with KSR v. Teleflex, which sang a veritable paean to non-prior art evidence (e.g., market pressure and design need) in rejecting the Federal Circuit’s “blinkered focus on individual documents”—as the court acknowledged in one post-KSR opinion. Yet despite its lack of a coherent normative or precedential basis, classification as primary or secondary is well-entrenched and matters a lot in practice because the label can lead to arbitrary bolstering or discounting of certain evidence. On the one hand, the “secondary” scarlet letter can lead courts to ignore good evidence like failure of others, harming the patentee’s case. On the other hand, a boost from the “primary” label can sometimes cause courts to overvalue evidence like unexpected results in favor of the patentee.

I propose a better approach to structuring § 103 analysis based on the timing of the evidence. Under my approach, the filing date provides a useful default dividing line between different kinds of obviousness evidence: rather than categorizing obviousness evidence as primary or secondary, we should instead view the evidence as ex ante or ex post. With this framing, we can see precisely why skepticism has more in common with teaching away than with commercial success: skepticism and teaching away reflect the pre-filing state of the art and thus speak directly to a PHOSITA’s lack of motivation or reasonable expectation of success to pursue the invention at issue, while commercial success is (usually) a post-filing response to the invention and thus constitutes only indirect evidence of nonobviousness. The readers of this blog are very familiar with the concept of nexus—in the sense that a proponent of objective indicia evidence must establish a link between that evidence and the technical merit of the invention as claimed.  My article’s proposal extends this concept to timing in that the party relying on ex post evidence must demonstrate a connection between that evidence and the elements of motivation and reasonable expectation of success at the time of filing. With ex ante evidence, however, temporal nexus is not an issue.

In my scheme, commercial success has more in common with evidence like copying or licensing than with the skepticism of experts and failure of others, which is classic ex ante evidence. And unexpected results, which the Federal Circuit has had great trouble fitting into Graham’s silos (some cases characterize that evidence as primary while others, as secondary), are instead divided into ex ante and ex post unexpected results.  As I show in the article, there is (pre-Graham) Supreme Court precedent supporting this general scheme and even some modern cases sometimes (almost) get it right. However, there is too much confusion in the case law to sustain a consistent approach, suggesting the need for an explicit adoption of the proposed ex ante/ex post framework. The tables below illustrate the key moves (indicated in italics), and the article explains why the proposed scheme will help decision-makers determine the relevance and weight of various nonobviousness evidence with greater accuracy.

Table 1. Graham’s Primary-Secondary Framework

Table 2. Proposed Time-Based Scheme

        [1].     Question marks next to “unexpected results” reflect their inconsistent classification

Read the draft article here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3820851