Vannevar Bush and the Unexpected Model of Innovation

In Science and Technology Policy in the United States: Open Systems in Action, Sylvia Kraemer spends a section of a chapter discussing Vannevar Bush and Science the Endless Frontier. Kraemer agrees that Science the Endless Frontier is an important document in the development of federal research policy but points out that Bush was frustrated in almost all of his goals in proposing a national research foundation to support basic scientific research. All he got was “the dedication of a single federal agency to the support of ‘basic’ research” (46). And even that wasn’t what Bush wanted–he wanted the organization to stand outside the federal government, operating more like the Smithsonian Institution, but supported by federal funds. Kraemer points out that Truman did not share Roosevelt’s (and Bush’s) views on the role of the government in supporting research and insisted that any new research organization be run by a presidential appointee, and be responsive to the will of government. Thus, the National Science Foundation rather than a national research foundation. The “Foundation” part got retained, as a gesture to the original idea, now much lost.

Kraemer takes up Bush’s treatment of “basic” research and attributes his interest in “curiosity for the exploration of the unknown” to New England Transcendentalism–Emerson and Thoreau meet research policy. Kraemer can then connect Bush’s ideas with a version of Platonic philosophical realism, in which science is the search for truth for truth’s sake, and contrast such an endeavor with a vision of productive pragmatism catalyzed by central government policies. Perhaps. But it appears that Kraemer chooses the pattern of analysis to fit the frame of her overall narrative, rather than to get at what it is that Bush recognized and sought to implement as governmental practice. In the clash of titans that are “open” and “closed” systems of research, it is easy to miss what Bush was proposing–and it wasn’t to use Walden as a guide to research policy.

I can’t say Kraemer gets the overall argument wrong. She might well be right. But let’s look at Bush’s work in more detail, and consider a different line of development. Kraemer notes that Science the Endless Frontier–the “Bush report”

has been memorialized in countless science policy discussions for its linear argument for the public support of scientific research. This is the view that science, uniquely and necessarily, generates the knowledge which engineers, in turn, draw upon to invent new technological devices and processes. Upon these new devices and processes depend economic growth and human betterment. This is the “Bush paradigm.” (34)

Apart from asserting a Kuhnian “paradigm” status to an argument for public funding of basic research, this passage from Kraemer lays out a version of the “linear model” and no doubt she is right in her description of how Bush’s work was treated by academics and wonks in science policy discussions. But the “linear argument” is not in Science the Endless Frontier. Even if a linear argument was attributed to Bush by the folks who have made their careers out of discussing science policy (only one level removed from the philosophers of science that Feynman had little use for), Bush’s argument was not that argument at all. Bush did not argue a linear model. He argued something else that got conflated with a linear model and attributed to him.

Benoît Godin has written a series of clear, helpful accounts of the development of the concept of innovation. In “The Linear Model of Innovation: The Historical Construction of an Analytical Framework,” Godin argues that Science the Endless Frontier does not set out a linear model of innovation:

One would be hard pressed, however, to find anything but a rudiment of this model in Bush’s manifesto. Bush talked about causal links between science (namely basic research) and socio-economic progress, but nowhere did he develop a full-length argument based on a sequential process broken down into its elements, or that suggests a mechanism whereby science translates into socioeconomic benefits. (5)

Godin shows that the arguments for the linear model come from scientists and then economists and business schools, and the linear model gets embedded in policy mainly because it comes to have a robust set of statistics built on its premises, and other approaches lacked such statistics. The statistic as a form of rhetoric. In a similar way, AUTM uses its metrics to entrench the idea that licensing patents owned by universities is the pivotal step in the linear model. By not gathering information that would argue against such a role (such as, say, the number of unlicensed, patented inventions, the number of failed startups, or the number of exclusively licensed patents for which no commercial product has been produced), AUTM creates an environment in which someone would have to first build (and finance) an organization to compete with AUTM in the collection of data–in order to have its own statistical epaulets with with to engage a public debate that would get the attention of “policy-makers.”

Vannevar Bush wasn’t proposing a linear progression from basic research to applied research to commercial development to public benefit. He was proposing something rather different, though related. Sylvia Kraemer wants Bush to be arguing for the funding of transcendental science in the mode of Emerson, but I suggest that Bush had something rather different in mind–he wanted more funding for folks like John von Neumann. Further, Bush’s interest was in how discoveries–of whatever age–become available to an established enterprise that has no knowledge of them. We might call Bush’s idea one of “unexpected shift” rather than “linear development.” What Bush wanted was an engine that created the potential for such shifts. He didn’t get what he wanted, and the subsequent institutionalization of university research, culminating with the Bayh-Dole Act, has ensured that Bush’s idea stays buried. We should change that. Let me explain.

Bush led the nation’s efforts to harness scientific ideas for the creation of technology to support the military in the second world war. As director of the Office of Scientific Research and Development, he saw first-hand how research could be conducted to produce results altogether unexpected by the military establishment and using resources and making applications altogether outside what the military set for itself as its own goals for research and development. Where the military looked for better optics for spotting scopes, Bush’s people worked on radar. Where the military looked for better anti-aircraft guns, Bush’s people developed monte carlo modeling for aiming. Where the military wanted more powerful conventional bombs, Bush’s people developed an atomic one. What Bush’s research teams did was obsolesce whole areas of military expertise, problem-posing, and research commissioning. Bush’s idea of innovation–developed in practice–is one of obsolescence of status within established social orders. His idea is not how to give that order what it wants, but faster, further, and better.

In Modern Arms and Free Men, published in 1949–after Science the Endless Frontier but before the formation of the National Science Foundation, Bush describes how the teams worked that built the new technologies used in World War 2. His effort is not to describe science in abstract terms, as if the “advancement in science” takes place independently of individuals. Instead, he gives as his purpose “an attempt to explore its [what science did in the war] meaning in the relations between man and man, as individuals and in the organizations they create” (7). Bush compares two “controlling motivations” in organizations:

One is fear, utilized in the elaboration of systems of discipline and taboos. The other is the confidence of one man in another, confidence in his integrity, confidence that he is governed by a moral code transcending expediency.

Bush bases his development on this contrast between a controlling “system of discipline” and “confidence” in a “moral code transcending expediency.” It is an important distinction, one that is entirely ignored in the development of university patent policies, post-Bayh-Dole. “The philosophy that men live by determines the form in which their governments will be molded,” argues Bush. “Upon the form of their government depends their progress in utilizing the applications of science to raise their standards of living…” (9). “Government” here can be read broadly–the federal government, to be sure, but also state governments, and the government a university administration imposes on its community in the form of policy. What one believes produces a form of government, and that form of government–an established order–has a great deal to do with how readily and with what judgment it adopts new things arising from new knowledge. Bush seeks to follow the “thread of the impact of science in altering the world and the relations of men therein.” How does one accomplish such a thing? That’s what Bush is about. Hardly the stuff for most university administrators to care about–dismissable as “idealism” (in that administrative sense of “useless to me in my political battles for status”).

Here’s what Bush argues with regard to the “real reason” that his research teams were productive:

The real reason we made such great progress was not bright inventors or clever gadgets. It was the fact that we had thousands of men who understood the underlying science in the field, who skillfully practiced the necessary techniques, who were good gadgeteers. They were in our universities, throughout industry, and in all sorts of queer places in the general population. Enough of them were gathered together and saved from senseless expenditure in tasks far removed from their skills to do the job, all the way from the research laboratory through pilot manufacture and engineering design to mass manufacture and skillful use in the field. We made great progress because we had the background for it. (244)

Bush goes on: “A regimented or governmentally operated system of universities and industry would not in a thousand years have produced the background for such an abrupt advance.” And after discussing advances in various disciplines, Bush recommends public support of university research, but with cautions:

This is a hazardous thing that we do, the giving away of public monies, and we need to surround it with safeguards for the many years ahead. With the Federal government plunging into the support of research on an enormous scale there is danger of the encouragement of mediocrity and grandiose projects, discouragement of individual genius, and hardening of administrative consciences in the universities.

Bush argues that the “ultimate control of policy” should be “in the hands of a representative body of citizens, selected and confirmed with care, bound to justify their program annually to the Congress in order to secure funds…” (247). This is the idea, argues Bush, behind the bill to create the National Science Foundation. Unfortunately, what actually came of it was not a citizen-directed institution reporting to Congress but rather a federal agency answerable to the President. Worse, perhaps, is the “hardening of administrative consciences in the universities,” something that is reflected in the changes in university patent policies from ones advocating judgment to ones demanding “a system of discipline and taboos”–the approach based on fear that Bush argued against.

The idea that Bush lays out in Science the Endless Frontier is not one merely of support for university research in the form of government grants, but rather is one of a way to deal with innovation in the context of prevailing established orders. His experience was with making new technology for the military–not the new technology that the military wanted, but what he and others saw was possible. That is, the change agent came from outside the military, did what they did not expect, what they could not have imagined. In Science the Endless Frontier, Bush turns from the military to his next target–established medicine:

Discoveries pertinent to medical progress have often come from remote and unexpected sources, and it is certain that this will be true in the future. It is wholly probable that progress in the treatment of cardiovascular disease, renal disease, cancer, and similar refractory diseases will be made as the result of fundamental discoveries in subjects unrelated to those diseases, and perhaps entirely unexpected by the investigator. Further progress requires that the entire front of medicine and the underlying sciences of chemistry, physics, anatomy, biochemistry, physiology, pharmacology, bacteriology, pathology, parasitology, etc., be broadly developed.

The idea is not that the medical establishment should fund more of what it already has–certainly there would be plenty of need there–but rather to broaden the scope of study in a host of areas, without defining a study problem based on what established medicine wants. This is perhaps counter-intuitive. Certainly an established order–a government, an industry, a company, or a profession–can set for itself goals of development and make new things. This is a form of innovation, of course. But it is not what Bush is after, it is not “science the endless frontier.” Folks read Bush’s title and insert a colon or dash to create “science: the endless frontier” as if Bush is talking about science rather than science as a frontier, as a movable line between the known and the unknown. “Study the unexpected, the rare, the different”:

Progress in the war against disease results from discoveries in remote and unexpected fields of medicine and the underlying sciences.

That is, Bush repeatedly identifies the importance of unexpected results, of studies and knowledge remote from any particular immediate problem. This I call the “unexpected shift” mindset. If one wanted a mental geography, it is of the explorer bringing stuff back, not of a linear impulse to go on from discovery to commercial product. Stuff gets brought back, not because there is a ready market for stuff, but because people document what they observe–and the critical point, as Richard Feynman argues–is knowing what to observe. Feynman tells the story of watching a student tossing a plate in the cafeteria, frisbee-style, and working out the equations of motion for it, and then realizing that this was the same pattern of analysis he needed for quantum electrodynamics–an unexpected shift, studying something that wasn’t in the problem definition to change the problem definition. This, then, is the big challenge in creating obsolescing innovation, and especially innovation that comes from outside an established order–how to know what to observe?

Bush argues then to observe “broadly” and “remotely” and “rarely.” Don’t get fixated on the problems as defined by an established order, the problems that give one status or even a career. The problems in an established order persist, often, in a chronic state of unsolvedness because careers, markets, social orders depend on the existence of the problem. It’s not that anyone will admit it openly, but it’s rather like the naval officers in Patrick O’Brian’s series–war is the only way to advance in rank, so much as officers might like peace, they like war even more. As Godin makes clear, Bush’s idea of “basic” research is nothing of the later formalizations but rather simply a distinction with “applied” work–the difference being less a matter of a sequence from one to the other than that some work is done in the context of creating a tool or product–“applied”–and some work is done without that context and so has a different set of bounding conditions–“basic” or “not applied”:

Apart from teaching, however, the primary obligation of the medical schools and universities is to continue the traditional function of such institutions, namely, to provide the individual worker with an opportunity for free, untrammeled study of nature, in the directions and by the methods suggested by his interests, curiosity, and imagination. The history of medical science teaches clearly the supreme importance of affording the prepared mind complete freedom for the exercise of initiative.

I’ve boldfaced a two phrases, as these address the problem of what to observe. Bush argues we need two things: people with “prepared minds” and “complete freedom for the exercise of initiative,” or put another way, people with training who then are given the opportunity to observe what they choose to observe. They are, after all, on the frontier, for which there is no possible guidance from the established order. On the frontier, there is no reason (and thus Paul Feyerabend bids “farewell” to such reason). On the frontier, one has to be guided by something other than the opinions of the established order, or its systems of disciplines and taboos–one may as well be guided by interests, curiosity, imagination. The noted scientist Carl Sagan, in Contact, makes a similar case, pitting the careerism and safety of the established order, represented in a twist that reflects history by the National Science Foundation, against the curiosity of Ellie Arroway, a scientist willing to risk career to search for extraterrestrial messages from space.

Bush’s insight, if I can summarize it, is not that basic research leads to applied research leads to commercialization. That can happen. It may even be useful to describe sequences of events with these categories. But Bush is not arguing for such a thing, and certainly not arguing that if the government creates categories of funding for research and then mandates people to fill those roles, that innovation will happen. Nuts to that. A “fun party” may be described as consisting of “planning,” “invitations,” “greetings,” “main events,” and “wind down,” but requiring all parties to demonstrate that they have followed this pattern is a recipe for changing a party–a fun time shared by a number of people–into a party–a matter of discipline and taboos. Bush argues instead for the availability of new knowledge created at the frontiers of science, brought back to be drawn upon by people who can bridge both science and human needs.

It is in this way that science–at the frontiers along with what is generally known–may change the relationships among people between people and established orders. This is not the stuff for just anyone to do. Bush may have had in mind people such as John von Neumann, a mathematician who led the development of the digital computer. See George Dyson’s Turing’s Cathedral for an account. John von Neumann combined training, curiosity, and a willingness to work outside the established order to bring something–digital computing–into being that would prove valuable even to the established order. My argument is that Vannevar Bush wanted to create an environment that would allow others with capabilities similar to von Neumann’s to have opportunities to draw on any relevant area of science for the resources to do what the prepared mind suggested. One was not expected to “apply” the finding, but rather to “use” the finding to guide one’s design sense.

This is rather like comparing the output of a Sirius Cybernetics Corporation Nutrimatic drink dispenser with actual tea. In this case, the linear model and the governmental administration that has arisen around it (and the Bayh-Dole law that relies on that administration to be there) are the Nutrimatic, capable of preparing a drink that’s “almost, but not quite entirely unlike tea.” The actual tea–Bush’s idea–is that with a broad frontier of new knowledge, combined with training for many people in science and making things (the “gadgeteers”), then “prepared minds” can lead the “coordinated attack” on a problem from outside the conventions of the established order–necessarily from the outside, because based on information from the frontiers as well as on conventional information. For medicine, that frontier information may come from physics (MRI, say) or from chemistry (DNA) rather than from a frontier of medical research proper. It is not that someone discovers a thing in physics (spin resonance) and then money is thrown at trying to develop a commercial product in some linear sequence. Rather, that the thing in physics is studied as something to observe, a thing found by imagination rather than for market or expediency or desire to make money because it could only be found by imagination (or accident or curiosity) and not for any good, official reason.

In Bush’s experience, innovation happened when one brought together folks on the frontiers, folks who know the foundations of science, and folks who can make things–academics, industrial scientists and engineers, and fabricators–all of them gadgeteers, able to prototype what they can imagine. In this, the key role is played by those on the frontiers, who put together an imaginative vision of what is possible. It is not merely that the frontier knowledge must be applied, but that it points out a way to go that could not be imagined by those lacking in that knowledge. Certainly it is not that the only way to apply frontier knowledge is to patent it and license its use exclusively to some one company on the condition that the company make a commercial product of it. In this context, such a patent-license process is much more likely to impeded and delay than to encourage and speed up use of the frontier knowledge. Such as how things are now, Nutrimatic-style, with administrative hearts–consciences–hardened by policy.

For Bush, the idea of government-funded basic research is to create the conditions for the assembling of teams outside an established order that can then introduce something new into that order–that is, a kind of obsolescing innovation rather than improving innovation. For this, Bush wants more frontier knowledge, and people with the mandate to gather in that knowledge from remote and rare and unexpected places for a particular purpose–the John von Neumanns–and people and resources available for “exercise of initiative” to make something new, preferably a cure rather than a bomb, but perhaps just a new method of additive manufacturing.

So Bush proposes a National Research Foundation that would be responsible to Congress for its use of money–as Bush was during the war–and would fund research that would create new knowledge at the frontiers. Kraemer doesn’t point it out, but Bush’s National Research Foundation is remarkably similar to the Research Corporation, with the differences being that the Foundation was to be funded by public grants rather than patent royalties, and the Foundation was to be directed by a professional answering to a board of appointed citizens rather than to a board of industry representatives. Otherwise, the difference is one of scale–Bush’s organization has more money to spend, and expects to give grants for five years of support, not just one year, if that.

Bush’s concern is with the limits of status quo imagination. If as Godin puts it, innovation is “an introduced change in an established order,” then Bush’s concern is how science at its frontiers along with people trained in science and the application of science (to make things) can operate outside the established order to advance our well being and change the established order, not merely serve it or cater to it or disrupt it. For this, Bush argues for the availability of frontier science to allow for unexpected shifts among disciplines to address ideas and problems. For this, he needs von Neumanns–people who can motivate the shifts and manage coordinated teams to pull off something new. These activities often take the form of “skunk works”–stuff protected from the status quo, even by elements of the status quo.

For that, Bush asks for research that is pulled away from administrative demands–whether in the university or in a company or in government. That’s perhaps why university research is seen as “extramural,” that faculty asked to be “released from official duties” to conduct it, and it is set up “outside” normal activities (apart from the efforts of university patent policy drafters to make extramural research appear to be “in the scope and course of employment” when all of the markings of such research is that it is anything but in the scope and course of employment”). Get the university research outside of the establishment’s reasoning about research. Get industrial research outside of the establishment’s focus on products and earnings. Get government research outside the political demands of government. There’s plenty of research that can be within these centers of establishment status quo. It was not Bush’s objective to suppress that research. His agenda was to create a new space in which the von Neumanns could operate without having to beg for money, without having to couch their ideas in whatever call for proposals some government officer had decided was the worthy theme of the year’s available research funding. If it worked for military technology, why could it not also work for medicine? And if for medicine, why not for communications? And if for these, then for other areas as well.

We don’t have that government mechanism for innovation proposed by Vannevar Bush. Instead, we have programs built around a linear model of research that most everyone apparently denies is workable (except for university patent administrators, who are fixated on it because the rhetoric of university patent ownership is now built on the linear model). Instead, we have Bush described as the key proponent for the linear model (he wasn’t, he didn’t propone such a model) and what he did propone is nearly lost. What Bush was able to do during the war he imagined could also be done in peace time, in civilian contexts. It would take money, special leadership, and an organization that was both freedom-loving and accountable for its actions. Bush got an organization with money, but he could not get the leadership he imagined or the structure of accountability that kept the organization out of range of political appointees. So we have the National Science Foundation instead. Yes, the NSF plays a valuable role. But it is not the role that Bush imagined, not the methods that Bush found were successful, not what was possible by way of innovation built from frontier knowledge of science along with conventional knowledge along with the ability to build new things.

Kraemer characterizes Bush’s vision as one of “closed science” to be developed by “elites” pursuing platonic ideals of transcendent forms. I don’t see that in Bush’s writings, other than at the frontiers of science–at the unknown–unlike at the frontiers of geography, one has to be driven by something other than money, market, or reason. Bush expected that people would use frontier science in unexpected ways, bringing it to areas other than the one in which a finding was made. Such shifts would happen because people with special insight were given the freedom from other tasks to explore the taboos of their disciplines and beyond. Bush based his expectations on “confidence” among individuals rather than on “systems” that enforced proper research behaviors.

Bush wasn’t able to solve the problem of how to create within government an organization that dispensed public money based on special insight into broadening the frontiers of science as alternatives to the established order and yet was accountable to Congress and the public for what it spent. A linear model sounds right, sounds reasonable, sounds like something that the public can understand and support. An “unexpected” shift model run by people with special insights and ideas doesn’t, even if there’s next to no evidence that a linear model is productive and Bush had plenty of evidence that what he was trying to normalize for civilian research had been productive in a time of great need. “Cross-disciplinary” research is not close to what Bush proposed. “Multi-disciplinary teams” is only marginally better. These are still inventions of the status quo, to protect its interests in conventional science–all fun and interesting, of course, but pretty darned expectable. The productive organizations that are closer to Bush’s vision look more like Silicon Valley unicorns–companies that combine all sorts of talent, led by someone able to put it together in a distinct form, drawing from everything available. One might think that Steve Jobs had something in common with John von Neumann or with Vannevar Bush or Richard Feynman.

It’s not that we need more entrepreneurs or that venture capital is the answer–though it is true that venture capital does specialize these days in selling innovation to the established order, much as Bush promoted technological innovation to the military. It’s just that Bush had the chance to create a government-based organization that did this, without the financial ROI thinking that dominates venture capital. Perhaps we still have that chance, even leaving government funding the way it presently is, even with the awful smell of Bayh-Dole clouding university administration of research–perhaps we can create research environments that let people pursue a curiosity, and still create an accountability for the use of money that keeps the money well used–that doesn’t invite mediocrity or expediency or grandiose projects or harden the consciences of university administrators.

I’ve seen such projects at universities. They are rarely government funded. They operate as skunk works. They follow a line of inquiry that doesn’t suggest the proper career moves. They create opportunities. They take care of their people. They aim to do something worthwhile, but on their own terms. These programs scrounge for funding, and some don’t make it and others don’t do nearly as much as they could if they could be relieved of the menial tasks of scrounging or dealing with the discipline and taboos of university administration. Perhaps there remains a shadow economy of productive innovation research. It’s not what the government says it is funding and it is largely aloof from university “technology transfer” programs. It gains no advantages from Bayh-Dole and it doesn’t get better with lots and lots of money–just enough to free itself from research poverty.

What would happen if there as a government funding program that one didn’t have to apply for–not a “King Lear” program of the usual sort (“who loves me the most?”), but one in which the funding found the projects already under way and said–keep at it, we will help with the finances. No proposals, no begging, no doing whatever sounds most pleasing, and no involvement of university administrators to define the budget accounting. At one point during the Apollo program, I’m told, secretaries at NASA Ames kept tens of thousands of dollars in cash in their desks–if someone needed a part or an instrument, they just got a wad of money from a secretary and went and got it. That’s a different world, one based on confidence, perhaps (auditors would call it something different, no doubt). Not the world most administrators create and inhabit. Theirs is a world rooted in fears–mild fears, mostly, because they have not learned to fear the sufficiently great fears but rather the fear of non-compliance, fear of inconsistency, fear of a bruhaha on their watch, fear of the need to make a responsible choice based on judgment not pre-fabricated policy, fear of audit.

It may be that we can recover Vannevar Bush’s insights, and it may be that government programs can implement them. DARPA has at times shown a greater affiliation with Bush’s insights than has the NSF. Perhaps we are jaded by “frontiers” of science–we need the “frontiers of the frontiers” to get us interested. I once hosted a visit from an Army intelligence officer looking for frontier research. “What do you want to see?” I asked. “Show me what you think I’m looking for,” he replied, “because if I tell you what I think I want to see, you’ll show me just that, and I won’t learn anything.” So I showed him the out-there stuff–programs that assemble their own data base structure from field-supplied free-form annotations, and the like. At the end of the day, he said “Thanks! that blew away at least three years of planning–we will have to start afresh.” On the frontiers of the frontiers science, all one has is a new imagination of what is possible. That’s nowhere close to a linear model of innovation.


This entry was posted in History, Innovation, Policy, Technology Transfer. Bookmark the permalink.