The Seductions of Big Science

There’s a reason why Oppenheimer will focus on the Manhattan Project, and not the bopping quantum physics scene in Göttingen that gave a young Robert Oppenheimer his start. There’s a reason why our depictions of spaceflight focus on the Apollo Program, and not the innovations of the International Latex Corporation that made walking on the moon possible. 

Big Science is seductive. It exerts a draw on the imagination that goes beyond purely rational considerations. If you’re romantic about science, Big Science offers some of the most colorful characters, the most dramatic breakthroughs, the boldest public commitments. These projects are breathtaking societal moments, rare points when science broke from its usual staid, plodding, incremental mode and did something dramatically different. 

Big Science is also tangible in a way that a lot of scientific progress is not. You can point to Los Alamos in a perfectly midwit way and say “The Science Happens Here.” Big Science – or at least the dream of it – packages up progress in a way that is concrete, amenable to policy, something we can spend money on. 

I also suspect that, for many, there is a little daydream imagining what it would have been like to be part of something historic like Apollo 11. We should have the bravery to admit that sometimes what underlies an obsession with Big Science is a longing to create such opportunities for ourselves in the present. 

This Big Science imaginary haunts science and technology policy. The pantheon is well-known: Bell Labs, Xerox PARC, DARPA, Apollo, CERN, the Manhattan Project, Operation Warp Speed. We grasp for these templates when thinking about how the government should play a role in advancing progress.

We need a DARPA for X! An Apollo Program for Y! A Manhattan Project for Z! 

But this infatuation with Big Science as the pinnacle of government involvement in progress can also be corrosive. A love of Big Science can be an obsession about means that causes us to lose sight of the ends. 

For one, bigger isn’t always better. Matt Clancy and Arnaud Dyèvre have a nice piece out recently that looks at The Size of Firms and the Nature of Innovation. One of the interesting patterns that comes out of the data is the degree to which larger firms produce more incremental innovations and get less per R&D dollar spent. That is, we should think about the golden productive era of Bell Labs and Xerox PARC as an outlier, rather than the norm. 

We should also critically inquire whether Big Science is genuinely needed to produce the kinds of progress we need in the present. It was arguably geostrategic necessity around specific fields and technologies – say, space and nuclear – that required a highly specific configuration of big funding, giant physical infrastructure, and researcher concentration. Have national priorities (and technology itself) shifted to make such heavyweight configurations no longer necessary? 

Second, Big Science may not be how most progress happens. I’m a longtime fan of the work of Eric von Hippel, who has spent a career cataloging places where low cost, humble innovation at the fringes end up having a gigantic impact. Center-pivot irrigation, which has reshaped farming worldwide, began as a series of hacked together bicycles and hoses in the field. It may be mastery over this kind of innovation, rather than the operational art of Big Science, that makes the bigger difference in advancing progress. 

The point isn’t to argue whether or not Big Science or Small Science is the superior model for achieving scientific and technological progress. That’s an absurd question that flattens the debate into an irresolvable binary. 

Instead, we should discard a specific fixation on Big Science, and the transmutation of progress problems into ever more opportunities for Big Science. When all you have is a Manhattan Project, everything looks like a bomb. 

The macroscientific view is to zoom out and think about Big Science as being part of a vast and varied portfolio of institutions that advance science and technology. That includes the Big Science of massive corporate labs, secretive government programs, and hulking research universities. But it also includes the long-tail ecosystem of niche academic fields, independent scholars, FROs, startups, citizen research efforts, and more. In effect, this collective distribution of institutions is a series of bets on the approaches that will shift the scientific and technological frontier. 

The question that science strategy should confront is: does the existing portfolio of scientific institutions in a society produce the rate and type of progress that is desirable? We need policy levers that allow us to dial up and dial down the influence and number of various types of institutions as needed. This is not only because what is “desirable progress” may shift over time: Big Science may be the most efficacious tool to achieve our ends at one moment, the weirdo fringe scientist at another moment. 

We also need policy levers that can reallocate the institutional portfolio because science is a dynamic, interconnected system that requires balancing. Classes of institutions may begin as dynamic sites of exploration and progress, only to become sclerotic burdens over time. 

This approach would discard a fetishism around Big Science as a reflexive paradigm for government involvement in progress. This is a healthier vision, positioning government towards the job of continuously tailoring the institutional portfolio of science to meet the needs of the present.