I am delighted to continue our author interview series with this conversation with sociologist Janet Vertesi . The goal for this interview series is to connect with researchers who are publishing outstanding social science and humanities books relevant to AIP’s focus on empowering physical scientists to create a better world . Dr. Vertesi specializes in the sociology of science, knowledge, and technology. She has spent more than a decade working among NASA’s robotic spacecraft teams as an ethnographer. In this interview we discuss her most recent book: Shaping Science: Organizations, Decisions, and Culture on NASA’s Teams , which draws on her research at NASA to delve into how a scientific team’s social organization can impact the results they yield.

Trevor Owens: Early in the book you establish “organized science” as a central framework for how you understand scientific work. Could you briefly describe this concept and why it ended up being critical to how you understand the teams you were embedded with at NASA?

Janet Vertesi: Organized science explains the connection between the way we organize scientists, and the kinds of scientific outcomes their teams produce. What I found when I studied NASA teams exploring the solar system – at the same time, in the same institutions, in the same field, sometimes even the same people – is that people behaved very differently depending on which mission organization they were interacting with. So it wasn’t about “personalities” or the form of the robot or the kind of planet they were studying: it was about the social organization. Just as people who studied factories found that the way the work was organized was stamped indelibly on its outcomes, I discovered an isomorphism between social organization and scientific outcomes. Just in this case, the outcomes aren’t widgets or profits, they’re things like data, scientific questions, and peoples’ careers.

Organized science takes insights from the study of other kinds of organizations in the world and applies it to the sciences. First, science teams are organizations with their own structures, cultures, rituals, governance mechanisms, and decision-making processes, and we should study them using organizational tools. Second, the choice of these different organizational elements shapes the knowledge teams produce: what we know about what’s out there in the world, how we come to appreciate our instruments, what kind of data we collect and how we share it, and even the careers of men, women, and minorities in the field. Third, every time we study something “out there” using these scientific organizations, we re-encounter our own organization, producing a looping effect that it’s difficult to break out of.

TO: In the book you explain that the Mars rover mission made extensive use of PowerPoint in their planning meetings. In contrast, the mission to Saturn made extensive use of Excel. What do we learn about the teams from how they use these off-the-shelf office tools? Along with that, when did it become clear to you that this would be a way to illustrate differences in the respective teams’ values?

JV: At my first Saturn team meeting, I spent most of my time watching brilliant men and women with PhDs in the sciences and engineering coloring in boxes on spreadsheets! Each team so obviously gravitated toward predominant use of one of these tools in their planning and did so in a way completely consistent with their mission organization. But neither of them used the tools the way they’re supposed to be used! The matrix organization at Saturn loved the columns and rows of the spreadsheet because it visualized the “multiple hats” each of them wore as scientists belonged to both an instrument team and a science theme group. The colors, not numbers, helped visualize the patterns of agreement and negotiation across the mission as they settled on a local definition of “fair” resource allocation.

Meanwhile, on Mars, everyone was pulling in huge images from the rovers’ point of view, and marking them up with text in Comic Sans, arrows and boxes—not a single pre-set template with titles and bullet points in sight. This was also in line with their organization oriented around consensus: you start with the same point of view, and end up in a collectively agreed-upon place.

It was so obvious to me when I compared my experiences sitting in on one group’s meetings amid PowerPoint-propelled vistas, with being awash in red-yellow-green Excel boxes. These were not merely ritual practices but visual metaphors and structuring activities that reproduced each team’s local social order.

TO: You characterize the Saturn mission teams as valuing integration of different goals between sub-teams. In contrast, you describe the Mars mission staff as valuing team-wide consensus. Could you tell us a bit about what integration and consensus mean in this context? Along with that, how did you come to see these concepts as critical to understanding the work of the two missions?

JV: Assembling a single plan from so many people and their priorities is incredibly difficult, and it’s the main challenge of operating a spacecraft on another world. I call this “the bus problem”: you’re on a bus with 300 other people, and there’s only one steering wheel and one camera. Where do you go and what do you take pictures of? Each crew has to decide how they’re going to decide, and everyone follows those agreed upon rules.

Saturn explorers rarely used the phrase “consensus” to talk about this process. In fact, many were skeptical of consensus because it dampened or flattened pluralistic opinions. But on Mars, they talked about consensus constantly as the only possible way forward and the goal of their daily meetings. Meanwhile, on Mars, I never heard anyone describe assembling the robot’s plan of action as a question of “integration.” At Saturn, that was what it was uniformly called as people tried to assemble various needs and priorities into a plan their robot could execute.

As a sociologist, you get a sense of what concepts are critical in a community by following the key phrases people constantly repeat, that they use to describe and rationalize their activity. I look to how members themselves make sense of and describe their activities, so these concepts come directly from the teams as ways of articulating their governance structures and cultures. That’s why I spend so much time on them in the book.

TO: On page 151 you note that “Not only are there different norms of operation and interaction on these missions, but the kinds of science questions being asked and answered are different as well.” What are some key differences you observed between the kind of science and data that each mission produced?

JV: Saturn science is rich and deep, largely achieved through single instruments used over a long time. So for instance, we know a lot about changes to spokes in the rings over time, and seasonal variation in temperatures in the rings, as two separate investigations. But to the best of my knowledge, most people wouldn’t ask if, say, temperature variations impact spoke changes, or how best to observe or test that. When multi-instrumental questions arose about Saturn’s aurora, or Titan’s lakes, or Enceladus’ hotspots, it was hard to line up testing observations because neither the teams nor the spacecraft were optimized for that kind of thing.

Meanwhile, on Mars, you didn’t have “deep” information or broad coverage—with few exceptions, observations were collaborative across multiple instruments in the suite. Each science question required multiple instruments to solve, and observational targets were selected that maximized many instruments’ capabilities.

Over time, the Mars mission assembled enough of those kinds of observations that they could look in a more encyclopedic way at the planet through one instrumental lens; and similarly, at Saturn, over time they amassed data sets you could put together later on once it was all in the public domain. But it wasn’t part of how the datasets were crafted or optimized.

People ended up doing quite different science and being celebrated for very different science. Saturn elevated the kinds of scientists who went deep on one instrument, like radar or particle tracking, while the Martians celebrated multi-instrumental investigations as an achievement.

TO: Toward the end of the book, you make the case that an iterative loop forms between the structure of the mission teams, their instruments, the data they produce, and their resulting scientific findings. How do you think those kinds of reinforcing processes get started? Is it your sense that they change over time?

JV: Sociologists use the term “imprinting” to explain how the organizational, resource and leadership conditions during founding sticks with the organization its whole life. It gets embedded in the local institutional structure and culture, in the technologies used to do the work and to communicate, and it’s hard to break free. Looking at conditions at time of founding, at Saturn they started with a multi-national and multi-instrumental team inherited from Voyager days. The problems of getting everyone together were so complex that one mission leader tapped an economist – an expert in game theory – to set up some of the systems of exchange. That zero-sum style of thinking and trading stayed with the team through decades of work and interaction.

Meanwhile, the Mars mission was founded in a different crucible, coming out of the Faster, Better, Cheaper era with a new Mars Program for funding, a need to keep things lean, and a Silicon Valley next door obsessed with flat hierarchies, agile programming methods, charismatic founders, and garage-built scrappy tech companies.

Every time scientists returned to their team meetings or analyzed their data, they re-encountered their same organizational structure, culture, and practices, naturalized as good exploration, good science, and just what Mars or what Saturn is like. People didn’t even change but doubled down on those ways of doing things when the missions faced hardships or novelties: like the recessions of the 1990’s and 2010’s, or changes in leadership or discoveries. With few other countries or missions offering alternative views, it’s very hard to escape this epistemic monopoly.

TO: What do you hope that scientists and policy makers reading this book will take away from your research regarding how to design and structure teams of scientists and engineers?

JV: Two things. First, there’s no one best way to run your science team and organization! It’s just that different choices have different consequences for science, for scientists, for technologies, data, and knowledge. So instead of just blindly setting something up for political, fiscal, or technical expedience in the moment, I hope scientists and policy makers think carefully about their scientific goals up front, then craft a sociotechnical organization to match and enable those goals. These research infrastructures, once established, last for careers and lifetimes, and getting it right up front is essential. The last thing I want to see is scientists fighting their organizations and their instruments, banging their heads against walls we already know are there!

Second, the people I studied insisted that each team behaved this way or that because one was a rover and the other an orbiter – so they had to work in such and such a way. Since then, however, we’ve flown orbiters organized like the Mars team I studied, and rovers organized more like the Saturn team. My findings hold true: it’s the organizational form that matters, not the format of the spaceship. What this means is, we have more power than we think, when we set these teams and instruments up, over how people will ultimately work together. Our fates are not pre-determined by technology, but rather we can shape the futures we want to see.

TO: What kinds of reactions have you received to the book from planetary scientists and people working with NASA?

JV: The book has hit its mark in many ways! I’ve already worked intensively with several missions-in-formation and early-stage interdisciplinary centers to implement findings of the book and structure their organizations. My favorite thing about this is there is no one-size-fits-all solution – every case is unique! For example, the Europa Clipper team needed to operate more like the Mars rovers to achieve their synergistic science: after all, no single instrument can detect life or solve problems of habitability, all the instruments need to work together in concert. So I brought in a lot of lessons from Mars exploration to a community with more heritage in Saturn-style models.

At the same time, a team at APL building an Interstellar Probe to surpass Voyager and get to the interstellar medium had a different challenge: they’d all be dead by the time the mission got to its mark! The Saturn style of science and organizing is far better suited for extreme longevity than the Mars style, even if space physicists typically prefer to work more like the Mars scientists I observed. So I brought in a ton of lessons from the Saturn team.

Now that the two missions I wrote about are long over and their powerful legacies are cemented in the history books, my greatest pride is in seeing their sociological lessons implemented amid these new, next generation teams. These scientists and engineers are thinking seriously about teaming architectures as they assemble their collaborations.

TO: Has your thinking on any of the issues in the book further developed or evolved since it came out a few years ago? If so, how?

JV: People used to ask, “but why did the teams structure themselves this way?’ To answer, I gestured briefly to the 1990s, funding architectures, and budget cuts. I decided to take a closer look at those phenomena. So in the book I’m currently writing, I take the money element seriously as part of the social construction of technology and science and explain exactly how it is that missions (or any other large technical project) come to take the shapes they do. I draw on ethnographic work with Clipper while it was in formation throughout the financial crisis, furlough, shut downs, and political and economic uncertainty, and show how the pressures and solutions are the same as the ones that confronted historical touchstone missions in the 1970s and 1990s. Money matters for how and why we assemble these particular instruments and teams just so.

TO: For readers interested in the topics you explored in this book, are there any other recent science and technology studies books you would recommend on related issues?

JV: Yes, but they’re not about spaceflight! Two recent favorites are Ben Shestakofsky’s ethnography of a Silicon Valley AI startup, Behind the Startup, and Natalie Aviles’ An Ungovernable Foe, about the National Cancer Institute. Shestakofsky shows how venture capital shapes the entire organization and the way it divides up the work, such that the “AI” tasks are mostly done by ghost workers in the Philippines while the engineers in San Francisco interview more engineers to hit their VCs’ quarterly growth targets. Aviles shows how federal agencies like the NIH are not slow bureaucracies but sites of fast-moving, valuable innovation in research, with organizational innovations that produce drug and vaccine discoveries. Both are great studies of technical organizations that show how the sausage is actually made behind the scenes.