The great impacts Houdini

ABSTRACT

I argue that the broader impacts conversation in research evaluation is designed to look like it addresses difficult questions about progress and the good life, whereas in fact it avoids them. In so doing, this discourse does not stay neutral on these questions. Rather, it supplies a default, unexamined answer. The use of normative anchors, or principles, in talk about Responsible Research and Innovation is laudable but inadequate. The problem is, though, that any adequate conversation would seem hopelessly antiquated if not hostile to the assumed goodness of technoscience.

KEYWORDS:

• Broader impacts
• research evaluation
• good life
• well-being

One way research and innovation can be irresponsible is by producing the wrong impacts. Thus, as René von Schomberg (2013) argues, Responsible Research and Innovation (RRI) would be that which produces the ‘right impacts.’ The unpredictability of research and innovation, especially given the central role of free markets, may make this a practical impossibility. The values plurality of modern society may make this a conceptual impossibility: we could never agree on what counts as a ‘right’ or ‘wrong’ impact.

Von Schomberg thinks we can, however, agree on certain ‘normative anchors,’ which amount to general principles that anyone can endorse. In this way, we can avoid two undesirable outcomes. First, by having explicit discussions guided by these anchors, we can avoid the current market proliferation of impacts that are ‘wrong’ from the standpoint of the anchors. Second, we can avoid entanglement in incommensurable philosophical discussions about the best kind of life and society. We can agree upon and selectively achieve the ‘right’ impacts while still peacefully disagreeing about the ‘good’ life.

I don’t think we can elide the philosophical questions like this, though. In supporting research and innovation, we are already promoting a certain vision of the good and in supporting any given particular project we are putting our thumbs on the scales determining which kinds of lives will be promoted and which will be marginalized. We cannot remain neutral on questions of the good; we can only answer them more or less explicitly.

To see this, consider the discourse surrounding broader impacts. They are often discussed in adversarial terms. The noble researcher guarding academic freedom from anti-intellectual, neoliberal jackals. Or the crusading public servant putting accountability into a system too long premised on blind trust.

But what if ‘broader impacts’ is a sleight of hand? The policymaker is the magician and the researcher is the lovely assistant. She steps into a box and he appears to impale her over and over again with long swords. In reality, though, she is unscathed. Nothing about science or its relationship to society is even touched in the process.

The whole thing has been an act and you have to look carefully, using the slow motion button on your remote control, to catch the crucial diversion. The scientist says, ‘What I do is vital to social progress even when I don’t intend any benefit.’ The policymaker says, ‘We cannot simply trust that your curiosity is a good way for us to spend our money.’ The scientist pouts as she is led into the box. The policymaker tests the blade of his metrics against his finger nails, ‘Well, we’ll see if you actually have any impact!’ The scientist gives a mock grimace as the sword goes in. What a show!

The diversion comes right away when the word ‘progress’ is mentioned and then promptly elided. Then the words ‘benefit’ and ‘good’ are used and quickly passed over in the same way. The final formulation rests on ‘impact,’ but the initial question was one about progress, that is, whether more knowledge, more science, is always a good thing. If the accountability culture or the broader impacts regime or RRI was something serious, and not just a show, then it would directly confront the questions: What is progress? Which is to say, what is a good impact?

That’s the central illusion: the assumption that any impact is a good impact. At the end of the sixth day, ‘God saw everything that he had made, and behold, it was very good’ (Gen. 1:31). It doesn’t matter that some of the animals eat other animals (so not everything is good from all perspectives), it only matters that stuff was created and it’s dynamic and complex and awe-inspiring.

Consider the mechanical tomato harvester as discussed by Langdon Winner (1980). In the 1940s, researchers at the University of California developed a marvelous machine that could greatly increase yield and accelerate tomato harvesting over the old hand-picking method. It also cut the cost of harvesting by about $6 per ton, which in turn meant a cheaper tomato for the consumer.

As Winner notes, though, the benefits were not equally distributed. With the introduction of the new technology, the number of tomato harvesters dropped from 4000 in the early 1960s to 600 in the early 1970s. Mechanization eliminated 32,000 jobs, massively transforming rural communities. Attorneys for California Rural Legal Assistance sued the university, arguing that they spent tax dollars on a project that benefited a handful of rich individuals to the detriment of family farms and rural California life more generally. The university argued that to accept these charges ‘would require elimination of all research with any potential practical application.’

Lots of research will have such mixed impacts, creating winners and losers. To delve into this issue would get hairy. Not only could researchers lose lawsuits, but society would have to confront deeply engrained myths about the unassailable goodness of knowledge. Maybe hand picking is better, after all. Yes, it’s harder and more expensive, but family farms and rural communities can survive with their traditions of slow, intentional, place-bound living. There is a certain way of life that has been displaced, but how could we take that seriously? What are we, anti-science, anti-progress?!

From what I can tell, the whole broader impacts conversation is painstakingly arranged to avoid this issue. I credit von Schomberg for at least taking a step toward it with his inclusion of ‘normative anchors.’ But these are so vague and manifold as to be useless in actually determining whether an impact is good or bad or right or wrong. For example, two of the anchors are the ‘promotion of scientific and technological advance’ and a competitive economy, which could act as a blanket greenlight for anything, including tomato harvesters. These would have to be balanced with other anchors such as solidarity and social justice. The conversation von Schomberg imagines falling out from these anchors would inevitably be premised on what kind of lives and society are best – big mechanized agribusiness or small, handcraft farms.

Von Schomberg’s normative anchors are at best something like principles in bioethics: they can start conversations about competing goods, but they cannot replace them. The danger is thinking that we are able to settle such substantive questions by trucking solely in principles. They turn into default, unexamined notion of progress and get baked into certain anodyne goals like ‘economic and social well-being’ (Fealing et al. 2011, 283). Then, metrics are built atop these goals, as if we know what they mean! Then, viola, the legerdemain is complete as we go about instrumentally linking the means of research to these wholly black-boxed ends. An enterprise that looks like it is about the examination of ends (what good is science?) is transformed into a flurry of means (how do we track cause and effect?).

What we don’t notice is how the swords of metrics, even if they are informed by a discussion of principles or normative anchors, bend when the policymaker thrusts them into the box. Indeed, these metrics (derived from goals like innovation, economic growth, health, sustainability, energy, and national security) are designed to curve perfectly around the shapely body of science. All the metrics are science-shaped swords. That is, they assume a certain vision of progress or the good life premised on modern technoscience. It is a vision that is supposed to give us, as Albert Borgmann (1984) puts it, a commodious life. Progress means being liberated from the toils of picking tomatoes and enriched with the bounties of perfect, genetically modified fruit made available without regard for season or place. Alternatives within this vision may be more or less explicitly discussed and von Schomberg is right to argue for more explicit discussion. But alternatives to this vision are simply not taken seriously.

Pick any of the principles and it is easy to imagine all sorts of bad or wrong impacts happening after they have been endorsed. Economic growth can cause environmental damage. Innovation can cause anxiety and burnout and erase cultures. High-tech healthcare can bankrupt families without any improvement in quality of life. Energy independence can create local sacrifice zones where extractive activities take place. National security can erode civil liberties. And more generally, technoscientific advance can imperil the planet and fuel a capitalist spiral into massive inequalities. In short, the vision that underwrites the goals that are built into metrics is ambiguous, but it is being taken at face value. Americans are no happier now than they were sixty years ago despite all the progress of mechanized tomatoes and so much more. Yet nothing in the broader impacts regime will reckon with this kind of unsettling reality. It has no sword that doesn’t swerve gracefully around hard questions.

A good deal of talk about RRI these days focuses on reams of un-reproducible and apparently invalid science. Daniel Sarewitz (2016), for example, argues that responsible research must be more tightly linked to practical problems and more often cashed out in technological solutions. We got our ‘cornucopia of miracles’ (Internet, jet aircraft, cell phones, lasers, satellites, GPS, nuclear and solar power, etc.) not through free curiosity but the heavy guiding hand of the government. Yet nowhere does Sarewitz question the goodness of these miracles. Perhaps in 50 years, this list of miracles will include self-driving cars, autonomous robotic nurses, genetically enhanced humans, and genetically novel organisms made in labs. Is it enough to ensure we bring about these ends efficiently or should we also question the ends themselves?

Everyone involved in the conversation assumes we know where we want to go (progress!), we just need to get there more efficiently and faster. The assumption is that science churns out a collection of goods that people are free to adopt or decline (see Briggle 2014). It’s the great buffet myth: that atomistic individuals can select from an ever-expanding economic pie of technoscientific goodies. Yet obviously, this is not how it works. The farm workers could not just ignore the tomato harvester. The hiker cannot just ignore the interstate highway cutting through his old mountain trail. We can’t still opt for a world without e-mail, atomic weapons, or climate change, whether we consider these miracles or not.

Any metric that defines an impact will carry a normative valence, and that valence bends toward the commodious life. The problem is that this bend is concealed behind the neutral-sounding goals of innovation or ‘economic and social well-being.’ Who could be against that?!

Well, maybe the family farmer.

Disclosure statement

No potential conflict of interest was reported by the author.

Note on contributor

Adam Briggle is an Associate Professor in the Department of Philosophy and Religion at the University of North Texas.

Additional information

Funding

This work was supported by the National Science Foundation [grant number 1445121].