We have looked at projects. Small projects can be pieces of bigger projects. It’s the big project that controls. Now let’s look at inventions, small and greater. We will see roughly the same thing: the idea of “invention” can be narrow (and made narrower by a patent strategy) or broad (as, formally and expressly defined by many university IP policies).
Invention does not have a formal definition in patent law. Even Bayh-Dole bites its tale with regard to the definition of invention. Here 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. 2321et seq.).
That is, invention isn’t defined at all! It’s just restricted to what’s “protectable” by patent law. The function of the definition is to eliminate the need to repeat “discovery” along with “invention” and to include plant varieties under Bayh-Dole’s contracting requirements, as if they were patentable. The definition doesn’t do anything at all for an understanding of invention–other than to create cognitive dissonance because in Bayh-Dole, a part of federal patent law, we now have to remember that invention doesn’t mean even an invention “protectable” under patent law–patent law has been twisted to include non-patent law–the Plant Variety Protection Act, which is distinct from plant patents. Why not also mask works? Or rights in technical data? Sigh. This definition is contract language elevated without much thought to be made patent law.
The key point in all of this is that when “invention” gets used, it does not follow that the entirety of an invention must be “patented” by any single patent. The definition of invention is not grainy in this way–it’s not “an invention is that subject matter claimed in a single patent application which subsequently issues as a patent.” What gets claimed in any given patent is first a choice of the applicant (claims can be rejected by the USPTO, but claims don’t get added by the USPTO). An applicant can draft the claims however the applicant wishes, to be narrow or broad, to address specific products or not, to cut out the contributions of particular inventors, even.
The invention, however, that is defined by a patent–the claimed invention–is not necessarily the full invention that has been conceived, not necessarily the only patentable elements of the invention. The invention that has been conceived might end up resulting in multiple claimed inventions, might end up with portions in the public domain because applicants have chosen not to claim those portions (or didn’t realize they could), might end up with claimed inventions owned by different owners (because an invention owner may assign the invention, or because the original owner was not first to file a patent application on some aspect of the invention). A claimed invention, like a small project, may be part of a greater invention, one that has been conceived but not yet reduced to practice in all of its parts. Claiming a part of such a greater invention does not exhaust the invention–the remainder of the invention may still be patentable. And in any case, it may still be regarded as an invention, even if the remaining elements are not patentable.
A license to a greater invention, an invention regarded broadly, involves necessarily a license to every aspect of the invention under the control of the owner of the invention when the license is granted, claimed or held as a trade secret or otherwise “protected,” and including any assignee of that owner.
Let that sink in. If I invent a 3d cookie, claim the method immediately but have to work on getting the dough conditioners right, any license I grant in the 3d cookie invention includes not just a license under my method patent but also a license in my dough conditioner work–all part of the greater invention. It does not matter that I might cleverly assign my invention to a company, a “commercialization” partner, to work out the dough conditioners and extend the methods; or obtain “research” funding from a “commercialization partner” in exchange for agreeing to assign any inventions I make to that partner–the license I grant runs with the assignment. Even if when I formalize my license of the method and sneakily restrict that license to just the claimed invention of the method and expressly state that nothing more is licensed, I still have not exhausted my obligation to license the entire invention established by the sponsored research agreement, established by my policies and practices with regard to inventions, established by the equity of the circumstances.
Only if a sponsor agrees that in a given license to a claimed invention that this license fully meets my obligation to grant licenses do we reach a point of common understanding regarding the promised licensing (and only a fool of a sponsor would accept such a provision).
Just because a university claims a part of an invention in a patent application does not mean that the invention automagically reduces to just what the university claims in that patent application. There may well be a greater invention, one that is not bounded by patent claims or even by patent law. The fact that patent law does not define invention, then, is an important thing. If a university defines invention broadly and promises that a sponsor has a right to any inventions made in supported research, then that promise runs with a broad idea regarding invention–at least as broad as the university’s own definition of invention (if a university administration is fool enough to attempt a definition–and many are). A broad idea of invention does not require either conception or reduction to practice as these terms are used in patent matters. A broad idea of invention does not have to involve absolute novelty or patentable subject matter. The promise of a license cuts across patent and copyright, across trade secret and access to data and tangible property such as prototypes.
The folks writing university patent policies go haywire trying to define invention–they end up making a random list of stuff, often including “invention” in the definition of “invention”–proving that they don’t have the slightest clue what they are talking about. Their lists, too, are random and despite reciting many buzz words–ambiguous, strange, silly. Thus, someone might invent a new therapeutic drug (a “big” invention) but actually only discover that a particular compound creates a biological reaction that can be exploited for therapy (a “small” invention). A patent attorney, working with the small invention, seeks to establish claims that are as broad as the biggest invention possible–any similar compound, the use of any of the class of compounds as a therapeutic, methods of preparing the compound, and the like. It may be that multiple additional inventive steps must be undertaken to produce a commercial drug, but the “small invention” (a compound with claims) implies the “big” invention (the drug in final form, with all the necessary inventions attached).
Here is a university patent policy definition of invention (University of Mississippi Medical Center):
for the purposes of this policy, refers to any item that reasonably appears to qualify for protection under the United States patent law or other protective statutes, whether or not actually patentable, or which appears to be commercially licensable.
The first part is workable–any “item” that is or may be patentable. The second part is totally screwball–“or which appears to be commercially licensable.” What is the possible standard by which something can be determined to “appear” to be something? Who makes the determination? Who announces the apparition of commercial licensing? I see commercial licensing appear in that cake doughnut! That cake doughnut is therefore an invention! I so own that cake doughnut! Jiminy.
Another policy, from the clueless University of Utah:
“Inventions” shall mean and include: inventions, technological advances, scientific discoveries and improvements (whether or not patentable); all trade secrets and know-how; trademarks and service marks; and the tangible and intangible results of research (including, for example, data, lab notebooks, charts, biological materials, cell lines and samples). “Inventions” shall also mean copyrightable works that include subject matter that is also patentable, as determined by the TCO Director, and software, notwithstanding the extent to which software may be governed by University Policy 7-003 (Ownership of Copyrightable Works and Related Works).
Here, utter nonsense. “Inventions” shall mean and include “inventions.” Holy crap, Batman! Let’s have our definitions “include” the term being defined, so that the term now must mean more than it otherwise would mean. Next, a random walk:
results of research
These, with the additional point of “whether or not patentable.” Thus, the definition does not concern patentable inventions, but rather defines a new, apparently massively encompassing bureaucratic creature, the “invention,” that takes in any “result of research” or “trade secret.” A trade secret need not be inventive, need not be a result of research–it can be most anything, so long as it is not generally known, is subject to efforts to keep it from becoming generally known, and has economic value. (That the policy includes both results of research and trade secrets indicates a usage in which some trade secrets must be something other than the results of research–or there would be no reason to include both. Sigh.) What is an invention in such a policy definition but a matter of bureaucratic convenience in asserting control over whatever? Wait, there’s more.
An invention in the Utah sense also includes “copyrightable works” (sheer nonsense–there are no copyrightable works–copyright happens when a work meets the statutory definition, but the clueless folks at Utah don’t know this and don’t care, obviously, that they don’t know this). Yes, it is only those works that include “subject matter that is also patentable.” Wrap your mind around that. Software is *not* an instance of this, since software is treated separately. So what possibly could be an original work of authorship and “include subject matter” that is “also patentable.” Must be some sort of unicorn or flying pig–a creature made out of words, but with utility in the construction of stories for sleepy administrators. Note, too, that the standard for whether the “subject matter” is patentable is merely the whim of the director of the technology transfer operation. And software is included in the definition of Inventions, regardless of whether it has anything patentable about it.
What sort of vision of invention comes from such a mess? A fog of darkness through which walk university bureaucrats announcing that they have control over most anything they choose to control. Here a cell line, there a cake doughnut. The definition amounts to the claim “university administrators will control whatever they choose to control.” But “whatever” provides no clarity with regard to invention, even if it makes things perfectly clear with regard to the intended scope of bureaucratic convenience.
Here’s the problem. The idea of invention matters for contractual deliverables, for compliance with the standard patent rights clause under Bayh-Dole, and for the scope of a sponsor’s license in inventions. The bureaucratic idea of “invention” matters for research contracting and licensing of inventions. But because university administrators are clueless with regard to the relationship between policy definitions and contracting practices, they create an even more awful situation for their institutions. If one has a “whatever” class definition of invention and then contracts to provide research services, then the invention deliverables may reasonably be expanded to include the university’s defined whatever. Such an expansion makes the university’s compliance obligation all the greater. The university’s patent policy, by carrying a hopelessly broad definition of invention, makes the university liable for the reporting, owning, and licensing of all “items” and “subject matter” that fall within that definition. Awfulness. A simple, clear approach would be to report what may be patentable, and be responsible only with that which comes to be patented. But that would be, perhaps, too difficult for the bozonet to grasp.
Administrators are so intent on getting control of whatever that they fail to see the obligations and liability and obstruction and suppression that they create in doing so.
Projects can be much bigger than the bit of work done with a given budget or even under a particular sponsorship agreement. A bit of work can support other work. A university that insists that all inventions are to be owned and commercialized whenever possible introduces a default bigger project to any bit of work–whatever other bigger projects might also exist based on the statements of sponsors, statements in requests for support, and any other documents or communications.
Similarly, expanding the definition of invention necessarily then also expands the default commercialization project and the scope of any license promised or equitably expected by any sponsor in any bit of work on that project. Any such license is broader than any particular element of a greater invention that might be claimed in any particular patent.
This result is confirmed in terms such as “in part” and “conceived” and “whether or not patentable.” If a university claims ownership of a thing (even if this is a coarse administrative fiction), then it also claims the right to license that thing (regardless of the theory of ownership, including none other than an assertion of ownership) and any promise to license necessarily is as broad as the university’s claim to ownership–and is not restricted to any particular patent issued to the university.