Recently, Daniel Sarewitz, professor of science and society at Arizona State University, posted an essay on Saving Science at http://www.thenewatlantis.com/publications/saving-science . It is long, though I think it is worth the read. Others have offered their perspectives on his points, including https://andthentheresphysics.wordpress.com/2016/08/21/saving-science/ and Judith Curry at https://judithcurry.com/2016/08/22/dan-sarewitz-on-saving-science/
I think that Sarewitz raises some valid points of concern, but he makes some broad sweeping generalizations that I think are unsubstantiated and at times myopic. Although I cannot possibly address all of my concerns about his assertions here, I hit a few points that I think are most important.
Vannevar Bush stated
"Scientific progress on a broad front results from the free play of free intellects, working on subjects of their own choice, in the manner dictated by their curiosity for exploration of the unknown."
The crux of Sarewitz's argument is that financing the free play of free intellects is not productive and often yields results that turn out to be wrong. He advocates for managed science with planned outcomes in technology or applications, a concept that many define as engineering. He argues that when managed for a specific outcome in the presence of a pressing need, science (er, engineering) yields the greatest outcomes. Although I agree that many of the most transformative ultimate outcomes of science have followed such enterprise, history suggests that without the free play of free intellects, most such ventures could never have been conceived in the first place. I argue that today's science has too little rather than too much free play. Free play today is often overly constrained by imposed goals and social pressures, sometimes including too much drive to conform to pass peer review.
Yet, what modern technology or innovation of science did not have at some point at its roots the free play of free intellects?
The Manhattan project yielded the atomic bomb and nuclear energy technology. This project was clearly more engineering than science, managed for a specific outcome, following the ideal supported by Sarewitz. I argue that this project would not have had a context to begin without the works of free-playing predecessors, many of whom simply wanted to understand nature and could not have imagined what kind of applications would ultimately grow out of the understanding the physics of matter and radiation that they gave us. This point does not support Sarewitz's suggestion that all scientific ideas should have a clear trajectory to positive direct impacts on fulfilling human needs before we should consider them for funding. Fundamental research from the play of the intellects often needs further development in planned projects to yield such outcomes. Pondering on the results of the playful projects often yields the planned ones. Projects like Manhattan are not, for the most part, new ideas, but development of old ones, refining them to fulfill the needs of society.
Some people might see the outcomes of both free and managed science during the first half of the 20th century as wildly different from the outcomes of more recent science. Yet, we might perceive recent science as more wrong than science of the past because our recollection of the science of the past is biased toward the positive and well supported conclusions that we retain from the past. Science that was wrong then has since been supplanted or neglected and has thus been forgotten.
The science of today will in the same way as in the past filter into the future, leading to some grand leaps forward, while much of it that is ultimately deemed less relevant or wrong will eventually fade into the ether. I do not see it as problematic that many new ideas are wrong. Some ideas are right, and I argue that fact will ultimately make the scientific enterprise worth it.
Sarewitz is troubled by the constant debates and changes of prevailing views in evolving branches of science. He raises a nice example in the conflict in dietary science, which recently has undergone a refreshing transformation. I suggest that the process of arguing through the literature, discussing stark contrasts between different views in light of evidence, is not a flaw in science: It is science itself! Scientists need to work to educate the media and the general public that peer review, although perhaps the best preliminary sifting process we have, is only a first step in the process of scientific discussion that ultimately yields high confidence conclusions. Whether a field finds a new direction based on newly published science depends on how the original authors and other scientists interact with the new material, often over many years. The public and the media need to recognize that science is not simply a list of facts emerging from peer review. I give peer reviewed results deference over statements not subject to such review, but I don't see peer review as infallible. I see the process of open review and discussion after publication as far more important than the peer review itself.
Direct management to generate new and fundamental ideas to yield specific outcomes is difficult, because we do not yet know what those ideas will be or how they will pan out. We need new ideas, even if most of them eventually turn out to be wrong. Most ideas of the greatest scientists of the past were also wrong and were ultimately discarded. Although science is far from perfect, it is a winnowing process that gradually blows out chaff, however frustrating and sometimes drawn out the process might be.
The type of managed science or engineering that Sarewitz supports is also needed, but such projects need bases to work from. Free play of the intellects provides those bases. Our own collective actions and inactions serve as the filter that will yield the best of today's science to fulfill the needs of the future world.