An invention is not a thing. An invention not a “cotton gin” or a “light bulb,” even though a cotton gin and a light bulb were once inventive. It doesn’t help to use things as proxies for inventions.
An invention is a set, a category of possible things (things = objects or techniques). Often lots of “things.” A patent teaches the category and claims as much of the category as the inventor chooses to claim. The category consists of variations and functional equivalents–all the ways of practicing the invention. Each “way” may be a thing or a subset of things. The “invention” is all the ways of practicing the invention. A sum, a collection, a set.
The invention set itself is not strictly defined. Federal patent law does not define invention. Well, except for the Bayh-Dole part of federal patent law does define invention–but that definition isn’t particularly helpful. Here it is (35 USC 201(d)):
The term “invention” means any invention or discovery which is or may be patentable or otherwise protectable under this title or any novel variety of plant which is or may be protectable under the Plant Variety Protection Act (7 U.S.C. 2321 et seq.).
An invention then, for patent law, is what patent law pertains to. (Plus, through a silly drafting cleverness, also non-patentable plant varieties.) Even here, the definition is expansive–“is” patentable restricts an invention set to the scope of possible claims for the set, even if an inventor does not make any or all of those claims, but “may be” is strangely broad. The scope of “may be” includes all those ways of practicing the invention that are not patentable but this condition is not obvious at the start–through lack of awareness of the prior art, through not having the test data to demonstrate that those ways perform as conceived, through lack of actual “utility,” and the like.
The point here is that a definition of invention can bend the scope of the invention. For Bayh-Dole, the definition is (i) stuff that’s patentable; (ii) stuff that may be patentable in someone’s opinion (whose?) but turns out not to be patentable; and (iii) stuff that up front is not patentable but is an invention anyway because plant variety certificates.
Linger on that “may be patentable.” “Is patentable” gets a confirmation when the PTO allows claims. Then we have a pretty good idea that the invention or discovery is patentable certifie. But “may be patentable” is different. In whose mind does the possibility of patentability live? The “inventor’s” mind? Here’s the PTO guidance on the matter, from the MPEP (2138.04):
There must be a contemporaneous recognition and appreciation of the invention for there to be conception.
This mattered more when the US operated on the first to invent model. Now, it’s muchly a matter of whether inventors are willing to declare themselves inventors when they file a patent application.
However, an inventor does not need to know that the invention will work for there to be complete conception.
That is, the invention does not have to be tested for there to be an invention–a new set of ways of practicing the invention. An inventor might hope the invention works–that at least some of the possible ways of practicing the invention the invention will work, even if the first way tried doesn’t work. Just because one way in the invention set doesn’t work doesn’t mean all the other ways also won’t work. One can file a patent application based on a way that ultimately is demonstrated not to work and still get claims issued that do cover ways of practicing the invention that do work. I could cite an instance from university practice.
Furthermore, the inventor does not need to appreciate the patentability of the invention.
The inventor must recognize the inventiveness of the invention but not the patentability. That is, the inventor does not need to know–or even consider–that the idea/discovery/work either is patentable or may be patentable. The invention is “made” when the inventor recognizes it as an invention, without knowing if it will work and without any regard for patentability.
Thus, with regard to Bayh-Dole, we see a weirdness–what does “may be patentable” reference? For there to be an invention, the inventor does not have to have any “appreciation” of patentability. Can someone else look at the inventor’s work and form the impression that that work “may be patentable”? Just anyone? A university technology transfer manager with three months of experience and programmed to receive? A patent attorney with a decade of experience? A federal grants officer who reads a report of research work and decides that some aspect of that work is patentable? I could have a full time job just receiving reports of research activity and declaring each “may be patentable.” Would that be good enough? It is very strange. At least Bayh-Dole is somewhat consistent in producing strange clauses.
The problem for Bayh-Dole is that it wants to reach behind the moment of patent application to apply its management requirements. Once there is a patent application filed with the PTO, at least we can expect that there are inventors and they have appreciated that their work is inventive, even if they don’t know if it will work, and that they think some aspect of that work at least “may be patentable.” But prior to the point of filing a patent application, it is not at all clear what may be patentable or who gets to decide. It’s one of those hand-waving parts of Bayh-Dole.
One could put things differently, and clear them up. Let’s say that Bayh-Dole made the first filing of a patent application on an invention owned by a contractor the threshold for the operation of Bayh-Dole. Then we would have reporting on the filing taking place afterwards–the federal government would get a copy of the application–which the federal government already has, of course, at the PTO. Why not just give federal agencies notice of the filing? Or why not just have the PTO forward applications received that cite federal funding–agency and contract? Simplify.
Now the issue would be whether the contractor could retain “title” to the invention, having filed a patent application. Well, in the current apparatus of Bayh-Dole, failing to file timely a patent application (and continue the prosecution of the application, and maintain and defend any issued patent) is one condition under which a federal agency can request title. Thus, if a contractor files a patent application, then contractor either expects to keep title or is, out of the goodness of a patriotic heart, filing the application on behalf of the federal government, which however hopeful one might want to be, is not at all likely. Thus, Bayh-Dole could be simplified wonderfully on this point–if a contractor files a patent application with the proper federal funding statement, then there’s absolutely no need for the contractor to disclose the invention directly to the federal agency or notify the federal agency that the contractor elects to retain title to the invention it has gone out and acquired title to. All that drops away. Nice. Unlike present Bayh-Dole, which is decidedly unnice.
Now consider other inventions–those that the contractor acquires but does not file a patent application on. This is where Bayh-Dole feels the need to reach into the period before the contractor files a patent application. Why? The apparent reason is so that the federal government can request ownership of the invention. Well, how strange is that? In pre-Bayh-Dole practice, if the federal government obtained ownership of an invention, it would release that invention to the public either by (i) dedication to the public domain–not filing a patent application; or (ii) if the federal government chose to use the patent system to disseminate knowledge regarding the invention, royalty-free non-exclusive nondiscriminatory licensing, often without the bother of paper formalities. Just go do it. Pre-Bayh-Dole, the attorney general’s position had been that it took an act of Congress for the federal government to grant an exclusive license to an invention. Bayh-Dole undid all this. Thus, in effect, the reason that Bayh-Dole has to reach behind any patent application is to get rights for the federal government to either (i) use the patent system to disseminate knowledge regarding the invention while licensing the invention royalty free, etc.; or (ii) to withhold the invention from public use while seeking to license the invention exclusively, even to the point of assigning the invention to a non-federal entity.
In all of this, as a matter of policy, option (i)–federal patenting to release to all–is pretty lame relative to the bother of all the paperwork of disclosure and assignment and the like. It was exactly this option (i) that the advocates of Bayh-Dole railed about as ineffective. What would be available to all would be used by none, and hoppyflap like that. Thus, (i) could not be the intent behind this bit of Bayh-Dole. Instead, we look to option (ii)–federal exclusive licensing to the point of assignment. Bayh-Dole makes a huge deal of federal exclusive licensing–both in 35 USC 207 and especially 35 USC 209. It’s clear that all the pre-patent application activity by a contractor is directed at serving federal exclusive licensing. It’s not “if you don’t file a patent application, then the federal government decides how to give the invention away–either publishing the invention by filing a patent application or publishing the invention without a patent application” but rather “if you don’t file a patent application, then the federal government gets to file a patent application to attempt to license the invention exclusively.”
That’s the Bayh-Dole deal set up by 35 USC 202(a) plus all the 35 USC 207, 209 stuff: either you contractor file a patent application or the federal government gets to seek an exclusive license, owing you and your inventors nothing. All in all, it’s a pretty crappy deal all around–for the public, for contractors, and for inventors. But you won’t read about this in the university tech transfer literature.