The BioBricks Foundation supports a repository of “parts” for use in synthetic biology. It has implemented an interesting two-part arrangement called the BioBrick Public Agreement or BPA. This agreement aims to manage the interests of contributors and users of the repository and is worth reviewing. The BPA establishes what BBF calls “open wetware”–making synthetic biology parts available for use while minimizing the threat of IP claims and the overhead of bilateral licensing.
Bilateral negotiations are, in a deep cultural sense, a good thing. When two people meet, they trade, and to trade, they negotiate their interests. Trades this way may not be “fair” in the big picture of relative gains, but even so, such trades may be appropriate, beneficial, and satisfying to those involved. However, in the world of trading conventions, especially in the context of intellectual property, one encounters complexities that make it next to impossible for individuals to conduct their own negotiations. These complexities come about in two categories. First, there are the details of IP–it’s easy to get these wrong, forget important stuff, and generally create arrangements that won’t hold up later, if there is a need to rely on the written agreement for an answer. Second, there is the effect of many deals that become related. If one is doing only a single, independent deal, then have at it. But if one needs to conduct ten or fifty such deals before having access to sufficient assets to do anything useful, then having each deal be custom means quickly that deals will be incompatible with other deals–each one with its own requirements. Deals easily then may lack “interoperability” and doing one deal may foreclose doing the next deal.
Thus, to operate at scale, to build platforms and libraries of technology resources, public licenses provide a way to normalize a set of transactions so that they are interoperable. Doing so allows any participant to build at scale across any of the deals that are possible within the platform. This sort of thing, called a “reef” by Steven Johnson, provides for great opportunities for mixing and matching, recombining and applying resources that are made available in the deals commons. Anywhere that a single organization is not capable or willing to undertake the development of an entire system, then one has to look at how a group will do the work. Once one is looking at a group, one has to switch from proprietary dynamics to group dynamics. That’s what open source has done, for instance. That in part is what Henry Chesbrough has sought to underscore in his development of “open innovation.”
Group dynamics is what one is immersed in when one decides to work with academic software. The conventional development of invention-patent dynamics is anchored by the assumption of a single inventor, single invention, single patent, single commercialization partner, single license, single product, and piles of cash. This conventional development then can be extended for a couple of more inventors, multiple inventions and patents within a patent family, field of use licenses to split of products, and the like–all of which are the usual complications that technology managers work with every day and take some pride in understanding.
The problem is, this line of reasoning is not extensible to group dynamics. It fails. It is a source of recurrent tension in technology licensing offices that find themselves dealing with software that should go on an open model but the office is rigged for proprietary licensing. In working for years with academic software as a starting point for technology management, I learned how to think in native group dynamics, and learned as well how to work this dynamic with copyrights, but also with research information, with inventions, and with “non-IP intangible assets”–NIPIA–stuff that’s not owned, generally, but may be possessed and may be exchanged for value.
As an aside, one can see how offering a “software licensing” course to university technology managers can become how to cram software into a proprietary patent licensing model or how to recast one’s entire approach from a proprietary default built around hoping to find a big-hit patent to a group dynamic default in which one aims to build a platform or library or reef from which things may happen at scale, indeterminately. Entrenched, self-confident, orderly administrators, having sold leadership and faculty on a proprietary model, are uncomfortable when presented with a more general model that asks them to learn new things, revise their claims and practices, and restrict the proprietary approach to a selective, special case that happens perhaps five times a year out of 100 to 300 or more reports of invention.
In group dynamics, the plural is the default. There is nothing complicated in this, at the start. Everything is multiple, everything is at scale, everything is indeterminate beyond the core rules at the interface by which transactions take place. Expanding this interface is part of the activity that creates NIPIA. I find that a “commons” approach helps to frame the matter. A commons is a social space with an interior, a border, and an exterior. A commons may be a repository, as in the case of the BioBricks Foundation, or a group of people, in the case of a club or professional society. A university laboratory is a commons, as is in a broader way, the university itself. Within the commons, people share and exchange under the rules of the commons. With those on the outside, the commons manages its borders: people or assets may enter the commons, or may depart the commons, and the commons may trade for value with folks outside the commons. In a university, research data may move from one lab to another without a transfer agreement. Inventions developed in one lab may be used in another without a license agreement. It is a commons: there are shared understandings, and a common legal framework. Try to transfer data or inventions to a lab in another organization, especially if that organization has a for-profit tax standing, and university officials will argue that this transfer crosses the border of the institutional commons and requires formalities.
There is nothing, however, preventing the formation of commons that cross institutional boundaries. We (yours, Charles Williams, Gabrielle Campbell) did this, for instance, with the Rosetta Commons, which sets up an arrangement by which individuals and their institutions can establish a common set of rules by which software and data may be exchanged within the commons without triggering institutional demands for license formalities.
A commons may grow–adding assets and members–as well as trade in what it chooses to release beyond the commons. The commons is, for many initiatives, the intangible asset. In a proprietary approach, the intangible asset is often a patent. That’s what the activity of a tech transfer office creates, and then, in time, perhaps, a license as well, as another intangible asset. Combine those two assets and one has a relationship involving patent rights. In a group dynamic, the commons is that primary target. It does not depend on any particular patent right–it depends on a critical mass that makes it more valuable to participate than to remain isolated. One’s contributions may be few, but one gains access to the many more contributions provided by others also participating. Thus, in the metrics for group dynamics, one does not count patents and licenses, but rather public commons that reach critical mass and become platforms or reefs for follow-on activity, much of which by design will be indeterminate–owing little or nothing to any particular rights holder. In the case of the Rosetta Commons, the foldit protein folding “game” came about, in part, because the underlying modeling engine was available within the commons and not tied up, say, in an exclusive commercial deal. Foldit was not anticipated by Rosetta Commons–and that’s the entire point–that by keeping open things open, the commons is able to support startup companies, existing companies, and games that advance science.
Open source licenses tend to favor a commons with an open boundary. By this I mean that the act of obtaining an asset of the commons is the same as joining the commons. Trade with the commons means becoming part of it. A given piece of software, released under any of the major open source public licenses–GPL, BSD, Apache, Mozilla–provides the basis for an open, indeterminate commons. Anyone obtaining the software agrees to the terms of the public license, and may use, copy, and modify the software, and redistribute original or modified form, following the conventions of the public license. What modifications will there be? No one really knows, and that’s the whole point about this sort of innovation.
I once created a game called “Tradition.” It had only one rule: you could make a rule or make a move. At first, the only move possible is to make another rule. But depending on what rules you make, you can also then make moves, especially if you create a rule defining what a “move” is. The game is indeterminate. It may play differently each time it is started. Where does it end? That, too, is something that is framed later in the game, if ever.
It’s odd, but true: software is more helpful than invention in coming to understand group dynamics, and group dynamics is a more general approach to research assets than is a proprietary approach. Group dynamics can accommodate proprietary interests–these are special cases where the anticipated pluralities don’t show up. But models built on proprietary interests generally break when pushed toward a commons. Why patent at all just to let everyone practice? Why go to all the work to establish a monopoly position and then not make any money from it? How can anything new be created if monopoly investors do not have the benefit of a patent to recover their investments? Yup. There’s no real answer within a proprietary model to these questions, but that does not mean there are no real answers–there are. But those answers lie outside the proprietary model–in group dynamics.
In group dynamics, a patent right may manage early associations, may provide for cross-licensing access to other technology rights (what a commons does!), may manage quality assurance, may induce others to contribute rights to a standard (another form of commons), and may document and make visible contributions for their priority, provenance, and scope. These are all valuable functions for a patent in group dynamics. It perhaps does not need to be said that in group dynamics, there are a great many opportunities to become wealthy, if that is one’s goal–it’s just that these opportunities do note arise because one has a monopoly position–they arise on the contrary because one has many associates and is recognized for one’s contributions. In such a setting, one may also have intellectual property positions, but they are established, and respected, by others. It runs against everything a proprietary approach is made to believe about the world, as if everything is run by blind self-interest checked only by the rule of law as enforced by the power of central governments with courts and standing armies. I don’t think this is quite so. There is plenty of room for self-organization, for convention and tradition that are not established by statute and are not enforced by a government power. That, in fact, is what commons are, in the general case–a kind of social discipline that establishes identity and distributes control over a set of subject matter, to put it in terms of Identity and Control.
The BioBricks Foundation IPA consists of two connected transactions. The first is a contributor’s agreement that places an asset in the BBF repository. The second is a user agreement that comes into play when someone desires to use an asset from the repository. Both of these agreements use a semi-structured approach. Part of the agreement is fixed, and part requires the participant to supply information and choose options. The agreements are linked–each references the other. The contributor agreement anticipates the user agreement as a kind of grant of rights, and the user, in accepting the terms of the user agreement, also accepts that grant of rights, forming a contract on the matter.
The Contributor Agreement identifies the Contributor and the material submitted to the repository, and allows the Contributor to indicate whether it should have attribution in certain subsequent uses by a User. If so, then the Contributor can specify the form of that attribution legend.
The fixed part of the Contributor Agreement consists of allowing modification of the material to include a BioBrick identification tag, a general non-assert of proprietary rights in the material, a statement of any known proprietary rights that may apply to the material, a discussion of protocols for use and fees that may be charged, and sections disclaiming warranties and limiting liability.
It is worth noting the choice of non-assert rather than license. The non-assert extends not only to infringement actions but also to threats to assert rights, bringing complaints before the International Trade Commission, and enforcement of rights. In effect, a license is an agreement not to assert rights, but in practice a non-assert does not distribute rights, while a license may convey rights that may then be sublicensed to others. The challenge then is what happens when a User wishes to involve others in that use. Do they all have to sign up as Users?
The User Agreement is in many ways a mirror image of the Contributor Agreement. It has its own semi-structured format, with the User providing contact information, along the lines of a registration procedure. The User Agreement is set up as “click through”–no signature required for execution. Such an agreements depend on there being some form of right that is conditional on the click through acceptance. Otherwise, where is the offer and what is the consideration that makes the contract a contract?
The User Agreement provides that scope of the non-assert in the Contributor’s Agreement extends to anyone operating under the User’s “authority or control”–thus anyone in a relationship with the User. If one wants protection from an assertion of rights, therefore, one either signs up as a User–pretty straightforward–or establishes a relationship with an existing User that controls the use of the material.
The User Agreement has one additional requirement, identified as “No Harmful Uses.” The agreement stipulates that the User will “refrain from using the Materials in connection with any intentionally harmful, negligent, or unsafe uses.” One may wonder, what happens if a User engages in such activity? Does the non-assert covenant end? Or is there a cause of action for something other than assertion of proprietary rights–such as breach of the agreement?
The overall effect of the BPA is to place materials into circulation in a commons, suspend IP claims on materials within the commons, and provide for users to gain access to those materials by agreeing to conditions of use–particularly attribution (to BBF and to Contributors if so requested) and risk.
The BPA has its limitations, of course. It does not discuss what happens if the User breaches the User Agreement, does not flow down background IP that the Contributor might have access to (“belonging to Contributor” is narrower than “under Contributor’s control”), does not provide for ending access if the Contributor is forced to suspend use or supply of the material based on claims by third parties or because of identified hazards or other problems with the material. One might want a Contributor to distinguish rights held by other Contributors, which shouldn’t be a problem, from rights held by others who are not participating in the BPA. For these others, it is important to know what conditions are placed on the material being submitted–and how it is that the Contributor can legally submit the material if there are conditions on the material.
One situation that may arise here involves method patents, since the BPA’s interest extends to “uses” of materials. Thus, one might imagine that a Contributor comes up with a new material, but also knows that this new material is within the scope of a method patent held by another that controls certain uses. The material may be submitted, but the uses are then limited by the external party’s claims. An obvious exploit, then, to submit a material with limited use rights is to move the proprietary use claims to an agent (and thus outside “belonging to” the Contributor). This is equivalent to having a semi-structured arrangement where the Contributor may directly limit a particular use based on its own proprietary position. A response might be to reject submitted materials that carry external claims of any kind, but that also might reduce the number of materials made available in the repository.
Another matter has to do with who has authority to submit a material, and who has authority to agree to the use terms. These are also issues for open source licensing. Is it the individual? Or does the individual represent a group, such as a research lab? If so, is a contributing individual making arrangements for others in the lab who may hold patent rights, say, personally? That might make it important to have everyone in the lab listed as the Contributor, or in the alternative make sure that the Contributor is indeed one of those that does have the full set of proprietary rights in the material (since the rights are those that “belong” to the Contributor). On the User side, this is all taken care of if the User is one that has control over the others in the lab. Otherwise, again, everyone in the lab needs to sign up, or the User has to be defined as the lab generally, or the institution that hosts the lab.
This raises the open source question, is the contributing individual making arrangements on behalf of his or her employer? Universities don’t like to delegate contracting authority in this way, and bark and whine about it in policy statements. Courts on the other hand may look to principles of agency and ask whether others should have reason to know that an employee is not acting properly within delegated authority. Sometimes that is hard to ascertain by outsiders. The delegation of authority can be rather more an internal matter for discipline than it is a way of preventing employees from making arrangements that the institution does not want to honor.
Obviously, if a university has a compulsory ownership policy that operates without review, then the Contributor may well have to be the university and not individuals who have invented the new material or a new use. Individuals might submit the material, but their non-assert won’t go very far, as it is the non-assert of the university that is required to make the deal work. One solution here is to not think of the material as an invention and submit it anyway. In the first to file world, that will pretty much blow patent rights, especially past the grace period, putting things much more clearly in the hands of the individuals. University administrators may not like such behavior, but it’s not a patentable invention unless an inventor recognizes it as such. So just don’t.
All this would be much easier to navigate if universities built their primary research asset management policy on a group dynamic rather than an individual proprietary default. In a group dynamic, the expected deployment would be non-exclusive and if there were available opportunities for doing so, those groups doing the work would be authorized to deploy. Within a group, decisions would be made based on group governance–the role of a principle investigator, a lab director, or team leader would come into play to adjudicate any competing interests. In our development of group dynamics, we formed “projects” that could help groups establish internal governance, or recognize (for the benefit of others) a governance that they were already working with. In such settings the university plays the role of mediator if anything unresolvable required additional mediation. Clearly, if the university asserts ownership of everything with the primary intent to profit from controlling IP rights, then this mediator role doesn’t work–it’s not needed, to start, and even if it were, it couldn’t happen because of the overwhelming organizational conflict of interest between allowing materials to be submitted for general access vs. the prospect of making profits from proprietary positions. (This is actually a false dichotomy–one can make profits from open positions, and one may provide general access going through an exclusive agent–but it’s the kind of dichotomy that tends to present, so while false, it is also something of a social fact).
In this, one can see an asymmetry. In group dynamics, the aim is to extend the commons–this can be done in a variety of ways that speak to the interests of those involved. In group dynamics, non-exclusive is the default for access within the commons, and for exchanges with those outside the commons. One can make any number of non-exclusive arrangements on this model and still have more arrangements to make. However, in the proprietary model, any non-exclusive arrangement is ruled out until it is clear that there is no exclusive position to be had–a delay that may mean that the technology is obsolete before it is finally made available. Thus, defaulting to a proprietary scheme disenfranchises many open, collaborative arrangements such as the BioBricks BPA, turning them into a “special case” to consider after other opportunities to exploit an IP position have been explored.
This asymmetry means that a proprietary default turns an institution from a primary contributor to commons, platforms, libraries, and reefs to one that aims to exploit such things for an institutional gain at the expense of the commons. Multiply this default across all major research universities, and it is clear that as a matter of institutional policy, administrators prefer a model in which they defect from a research or use commons in an attempt to secure a prior position in the sale of products. The effects of such a position are multiple. It delays access and use while a product is being made. It adds costs because product development expenses must be borne by a monopoly partner. It makes investment decisions much more difficult because of the costs and the threats that others may design around, and the lack of any visible use of the new technology (there’s the construction of an artificial “funding gap”). It limits what can be used to specific product implementations, not generally all the variations possible when the invention was made (as happened with the exclusively licensed disease assay patents). It limits development of standards by fragmenting rights into multiple efforts to establish monopoly positions, and where such positions are not possible because of the costs or the scale or complexity of the product or the unknowns that have yet to be characterized, it prefers to hold things rather than release them.
There are cases in which a proprietary position is indicated, and folks should know how to recognize these cases and proceed accordingly. But in university research, these situations are rare–on the order of one or two a year, maybe even one or two a decade. That would be a roaring success for proprietary efforts. The roaring success for group dynamics, however, is the establishment of various sorts of projects, commons, and platforms by which research assets may be made available, evaluated, tested, modified, varied, characterized, reported, combined, developed, and used. All this stuff cannot happen if the authority for doing so is removed and groups cannot make their own choices and each situation is handled as a undesirable special case or exception to a proprietary “commercialization” first approach.
A big sign of dysfunction is when those authorized to make decisions for an institution know less about a situation than those doing the work. In a university research setting, most of the practice smarts are at the periphery, in the labs, and not with “managers” or “licensing officers” no matter their industry or legal credentials. One has to get out to that periphery, and be on it and stay on it, rather than aiming to shoot prospective inventions and bring them back to a central administrative lair to be counted and databased to see if they might be good eating. Open strategies such as the BPA provide an alternative that positions group decisions in practical ways, with uniform terms, that allow indeterminate new work to be built on the contributions made by many. That’s the native approach of university research. There’s no need to “change academic culture” away from such an approach. Indeed, the native approach, the group dynamic approach, is also the entrepreneurial approach. In this, the core value of the entrepreneur is knowing how value is created and captured for a new enterprise. That kind of decision does not happen if every possible asset is claimed and reserved for proprietary licensing.