Defining patent scope by the novelty of the idea.

• Author: Chiang, Tun-Jen
• Position: III. Upstream Capture: A Different View of the Problem through Conclusion, with footnotes, p. 1239-1268

3. UPSTREAM CAPTURE: A DIFFERENT VIEW OF THE PROBLEM

   In this Part, I want to offer a somewhat different perspective on the patent scope problem than the approaches of the prior literature. The approach can still be considered doctrinal rather than economic, but it differs from the existing doctrinal literature that generally focuses on disclosure, in the sense of the number and types of embodiments taught by the specification--an approach that leads right back to the levels of abstraction quagmire of existing case law. Once I lay out the approach, I will also explain why my proposed solution produces good economic outcomes, though my starting point is not an economic analysis.

   1. Upstream Capture Versus Downstream Capture

      My approach in this Article differs from the existing literature in that my focus is not on the specification disclosure. Of course, the patentee must still file a specification, and the specification must still disclose at least one working embodiment to receive a patent; but the amount of disclosure (in the sense of the range and number of embodiments, in contradistinction to the novelty of the insight that lies underneath) would have very little to do with the scope of the patent under my approach. What I argue is that patent scope should focus on the novelty of the patentee's intellectual contribution. A more detailed explanation follows.

      A starting point is to note, as in Part I, that patent law uses the concept of "invention"--which is synonymous with the patentee's contribution--in two contexts, which I shall call the "upstream" context and the "downstream" context. In the upstream context, the law compares the patentee's contribution against prior inventors, or in other words assesses novelty. In the downstream context, the law compares the patentee's contribution against later improvers, which is the patent scope problem and which, in current law, is framed as an issue of disclosure under [section] 112. And as has already been noted, patent law's definition of the invention differs across these two contexts. In the upstream context, patent law defines the invention as an embodiment or set of embodiments. (168) In the downstream context, patent law defines the invention as an idea. (169)

      This inconsistency can be explored more deeply. The intuitive--not strictly economic--justification for allowing patentees to cover later improvements is that we assume that a later improver copies ideas from the patentee, (170) and that these copied ideas have value. (171) For example, when the Wright brothers create the first airplane and a later improver creates a slightly improved airplane, we assume that the later improver nonetheless copied some principles of aerodynamics that the Wrights first discovered. In the downstream context, patent law recognizes that an earlier inventor contributing ideas to a later improver is a contribution of value, and thus requires the later improver to pay for this value. (172)

      Not so in the upstream context. As already described, in the upstream context patent law defines the invention as an embodiment, not an idea. (173) But the inconsistency goes deeper. Patent law defines the invention as an embodiment because, in the upstream context, it deems ideas to have no value. This sentiment is deeply entrenched in patent law, and was most clearly expressed by Learned Hand as the reason for the enablement requirement for prior art:

      No doctrine of the patent law is better established than that.... [i]f the earlier disclosure ... does not inform the art without more how to practice the new invention, it has not correspondingly enriched the store of common knowledge, and it is not an anticipation. (174) Thus, when the Wright brothers assert their patent against later airplanes, patent law assumes the later improver used ideas from the Wright brothers, and that using ideas amounts to a misappropriation of value. But it never asks whether the Wright brothers themselves used preexisting ideas from even earlier inventors. It never asks the question because upstream ideas are deemed to have no value, absent embodiments.

      The gap that emerges is that the idea that the patentee asserts against later improvers--the value the later improver is alleged to have misappropriated--may not in fact originate with the patentee, but rather with an even earlier inventor. Giving the patentee too much credit upstream then leads to giving the patentee too much credit downstream.

      What the enablement requirement for prior art fundamentally misses is thus the reality that prior ideas do have value to the patentee (just as they do to an even later infringer) because invention is a continuous process of ever-more-specific ideas. The patentee's working embodiment is the culmination of this chain of ideas. To take the example of the cure for AIDS again, in order to create a working cure for AIDS using X radiation, the following ideas must be developed:

      (1) Someone must diagnose AIDS as a distinct disease that requires a separate cure. (175)

      (2) Once AIDS is identified as a distinct disease, someone must suggest radiation therapy (as opposed to pills or other methods) as an avenue to pursue. (176)

      (3) Once radiation therapy has been identified as a possible avenue, someone must suggest X radiation in particular as a possible cure.

      (4) Once X radiation has been identified as a potential cure, someone must find the right dosage and calibrate a radiation machine to make X radiation work.

      Once we understand that invention is a cumulative process, what emerges is that patent law places far too much emphasis on, and gives far too much credit to, the eventual creator of a working embodiment. Each of these ideas is necessary to the eventual creation of the working embodiment, and are thus contributions of value. But the early inventors will receive no credit from the patent system because they fail to create embodiments: the first person to think of curing AIDS is unlikely to find a cure, and the first person to think of using radiation is unlikely to find the right type of radiation. The patentee who eventually creates the working embodiment should not be able to claim sole credit for all of the prior ideas. (177) In other words, in order to be fairly credited as the inventor of the cure for AIDS, a person should be the first to have the idea as well as to create the working embodiment. To do otherwise would be stealing credit from upstream inventors.

      This upstream-focused view is different from the standard disclosure paradigm that dominates today's patent law, where the patent scope problem is viewed through the lens of whether the patentee has taught every cure for AIDS, or every radiation cure for AIDS, or every dosage and machine variant of radiation X for curing AIDS. The standard disclosure-focused and embodiment-centric view asks an impossible and thus useless question. My standard, by contrast, asks a question that matches much of the moral intuition of patent law: was the inventor the first to contribute the idea to the public?

      From this new perspective, current patent law essentially creates the levels of abstraction problem for itself. What the enablement requirement for prior art does is artificially deem every patentee (who by definition creates the first working embodiment) to be a sui generis pioneer. Because all prior ideas by other inventors are excluded from evidence, the first person to create one cure for AIDS is artificially made to seem as if he invented all the ideas within that embodiment, from curing AIDS in the abstract down to discovering the specific dosage of X radiation. The levels of abstraction problem thus arises: the law must choose one idea among all of these options. Abolishing the enablement requirement for prior art then solves the levels of abstraction problem. The next section describes this in more detail.

   2. A Proposed Solution: The Patentee's Insight

      Once we adopt the perspective that patent law should be logically consistent across the upstream and downstream contexts, a solution to the patent scope problem emerges. What I propose is that a patentee should receive the level of abstraction where he is both the first to create a working embodiment and also the first to articulate the idea.

      It is important to understand that "articulation" is not just a slightly-weakened version of enablement or a disclosure test by another name. Conventional thinking equates disclosure with the teaching of how to make an embodiment: to enable the cure for AIDS means teaching how to make that cure. My articulation standard literally requires no more than a sentence that suggests the possibility, with no details on how to achieve it, such as: "It would be a good idea to have a cure for AIDS." (178) The difference between articulation and the conventional understanding of disclosure are thus sufficiently distinct to rise to the level of a difference in kind, not merely of degree.

      How my proposal operates can be made clearer by reference to the AIDS example. In order to obtain any patent at all, the patentee must first create a working embodiment of a cure for AIDS. Even then, the patentee would not be allowed to claim the highest level of abstraction of all cures for AIDS, because the idea of curing AIDS has already been articulated. But if he was the first person to think of using radiation to cure AIDS (where everyone else, for example, had only thought of using pills), then the proper level of abstraction is the cure of AIDS using radiation, which would encompass every type of radiation. Conversely, if the idea of using radiation had already been articulated by others, but no particular type of radiation had been identified, and the patentee is the first to suggest X radiation, then X radiation would be the proper level of abstraction (and would cover every dosage and every machine using X radiation). In short, the patentee who creates the first working embodiment would...